25 ELR 10657 | Environmental Law Reporter | copyright © 1995 | All rights reserved

Regulation of Radiological and Chemical Carcinogens: Current Steps Toward Risk Harmonization

Editors' Summary: Until recently, the regulation of chemical carcinogens and the regulation of radiological carcinogens developed independently. Different governmental agencies operating under different statutory directives were responsible for addressing the dangers from these carcinogens. As a result, different policies and practices were developed. This Article explores these differences and the record on resolving them. It first examines the history of federal regulation of chemical and radiological carcinogens and summarizes EPA's approach to risk assessments for them. It then analyzes the traditional risk management approach for radiation, as exemplified by the recommendations of the International Commission on Radiological Protection (ICRP), and compares this approach to the approaches EPA has taken. Finally, the Article compares the results obtained under these approaches. It concludes that EPA's radiation standards, which in many cases were derived under policies applicable to chemical carcinogens, are, for the most part, consistent with the ICRP's recommendations.

[25 ELR 10657]

Until recently, regulation of the hazards of radiation and chemical carcinogens developed independently.¹ This occurred primarily because protection from exposure to these two categories of carcinogens was the responsibility of different governmental agencies operating under separate legislative mandates. Not surprisingly, this division of responsibilities resulted in the development of different policies and practices for protection from exposure to these carcinogens.

This Article explores the record on resolution of these differences, as exemplified by the radiation standards established by the U.S. Environmental Protection Agency (EPA) during its first 25 years of existence, under a variety of statutes that address both chemical and radiological carcinogens. It begins by briefly examining the history of federal regulation of chemical and radiological carcinogens and summarizing EPA's approach to risk assessment for these carcinogens. It then analyzes the traditional risk management approach for radiation, as exemplified by the recommendations of the International Commission on Radiological Protection (ICRP), and compares this to the approaches taken by EPA under the Atomic Energy Act (AEA);² the Clean Air Act (CAA);³ the Safe Drinking Water Act (SDWA);⁴ the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA);⁵ and the Indoor Radon Abatement Act (IRAA).⁶ Finally, the Article compares the results obtained under these various risk assessment and management approaches. It concludes that EPA's radiation standards, which were, in many cases, derived under policies applicable to chemical carcinogens, are, for the most part, consistent with the recommendations of the ICRP.

[25 ELR 10658]

BACKGROUND

The federal policies and practices that address exposure to chemical and radiological carcinogens have roots that extend to the early decades of the 20th century, when the hazards to health from these materials first gained governmental attention. Federal regulation of chemical carcinogens began through the regulation of pesticides under the Federal Insecticide Act of 1910, which was replaced in 1947 by the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA).⁷ Regulatory authority under FIFRA was originally vested in the U.S. Department of Agriculture (USDA), which could withdraw a pesticide's registration if it posed a threat of "unreasonable adverse effects on the environment," defined as "any unreasonable risk to man or the environment, taking into account the economic, social, and environmental costs and benefits of the use of any pesticide."⁸

The regulation of pesticides increased in 1954 when the Food and Drug Administration (FDA) was given the authority under the Federal Food, Drug, and Cosmetic Act (FFDCA)⁹ to set tolerances in food and animal feeds "to the extent necessary to protect the public health" taking into account the need for an "adequate, wholesome, and economical food supply."¹⁰ In 1958, the so-called Delaney Clause amended the FFDCA to prohibit any pesticide residue in processed foods "if it is found . . . to induce cancer when ingested by man or animal, or if it is found, after tests which are appropriate for the evaluation of the safety of food additives, to induce cancer in man or animal."¹¹ Present-day policy for the control of chemical carcinogens in the environment has evolved from this highly protective position, primarily under a series of broad environmental statutes, to a more relaxed view that permits relatively small residual risks at levels characterized as "safe" or "acceptable."

Federal regulation of radiation began in 1946, with enactment of the AEA. The AEA established the Atomic Energy Commission (AEC), granting it exclusive regulatory authority in the United States over most man-made and certain naturally occurring radioactive materials.¹² The most significant sources of radiation exposure not regulated under the AEA are indoor radon (which arises from radium, a naturally occurring radionuclide in soil) and x rays used in medicine and industry.¹³ Collectively, the non-AEA materials are known as naturally occurring and accelerator-produced radioactive materials (NARM). NARM and x rays traditionally have been regulated, in varying degrees, by the states. Although there is no federal regulation of indoor radon, EPA has issued guidance under the IRAA for remediation of radon in homes.

The regulatory authority of the AEC included the establishment of standards to "protect health or to minimize danger to life or property,"¹⁴ without further statutory direction on how the standards were to be developed. Initially, the AEC based its standards on recommendations of the National Council on Radiation Protection and Measurements (NCRP), a private group of U.S. experts founded in the 1920s. Beginning in the mid-1950s, the AEC and its current successor agencies, the Nuclear Regulatory Commission (NRC) and the U.S. Department of Energy (DOE),¹⁵ increasingly relied on a more broadly based group, the ICRP, whose recommendations have shaped the regulation of exposure to radiation worldwide.¹⁶ The ICRP and the NCRP set forth similar principles; these have evolved gradually over a period of almost 70 years and express what may be characterized as the "traditional" approach to protection from the hazards of radiation.

The AEC's responsibility for radiation changed in 1970, when President Nixon created EPA through Reorganization Plan No. 3.¹⁷ The plan transferred the AEC's authority to establish environmental protection standards for radioactive materials under the AEA to EPA.¹⁸ The plan also transferred the USDA's regulatory authority for pesticides under FIFRA and the FDA's authority to set pesticide tolerances for raw agricultural commodities and processed foods under the FFDCA.¹⁹ In addition, through the enactment and amendment of a series of environmental statutes—most notably the CAA, the SDWA, and CERCLA—EPA became responsible for regulating both radiological and chemical carcinogens under the same statutes and legislative requirements.²⁰ Like FIFRA and the FFDCA, these environmental [25 ELR 10659] statutes typically contain more extensive legislative direction than the AEA for the development of standards for the protection of human health from carcinogens.

As a result of these changes, EPA has been faced with the challenge of establishing standards for radiation that are consistent with its standards for chemical carcinogens. EPA's response to this challenge has aroused concern by some in the radiation community that EPA's standards for radiation may not be consistent with the "traditional" approach to radiation protection, e.g., as exemplified by the current recommendations of the ICRP.²¹ Specifically, it is claimed that EPA's environmental standards for radiological carcinogens, which are generally set at lifetime risk levels on the order of 10<-4>, are not consistent with the overall ICRP limit for radiation exposure of members of the public, which corresponds to a lifetime risk of approximately 3 x 10<-3>. Further, some have argued that EPA should consider cancer risks from background sources of radiation²² as a basis for the selection of its radiation standards.²³

In the following sections, this Article explores EPA's approach to limiting risks from radiological and chemical carcinogens through an examination of the consistency of EPA's standards for radiation with the traditional approach to radiation protection. The Article makes a similar comparison of EPA's treatment of indoor radon.²⁴ The Article begins with a summary of the Agency's approach to risk assessment for chemical and radiological carcinogens. Risk assessment is generally defined as the process that attempts to make objective use of the available factual base, coupled with scientific judgments, to characterize the health risks associated with human exposure to a hazardous material.²⁵ The Article then examines the current risk management approaches recommended by the ICRP and those of EPA under the AEA and related statutes, and under the CAA, the SDWA, CERCLA, and the IRAA. Risk management is here characterized as the set of principles or policies used to establish a regulatory standard. It takes into account the risk assessment and additional nonscientific factors, such as statutory directives and judicial mandates, cost, and technological limitations and feasibility, as well as social concerns.

EPA Risk Assessment for Chemical and Radiological Carcinogens

For potentially carcinogenic chemicals, risk assessment usually involves four steps: Hazard identification, dose-response assessment, exposure assessment, and risk characterization. In the first step, hazard identification, the qualitative likelihood that the substance poses a risk of cancer is determined based on information from human, animal, biological, chemical, or physical studies. EPA describes the overall weight of this evidence in terms of five categories ranging from "known human carcinogens" through "not classifiable" to "noncarcinogens."²⁶ Next, the dose-response assessment, i.e., the quantitative relation between the magnitude of exposure and the probability of developing cancer, is developed, to the extent that the data permit.

The third step, exposure assessment, involves a "real world" assessment of human exposure to the carcinogen.²⁷ Data considered include the magnitude of the source of the carcinogen; its transport mechanisms in environmental media, food, and drinking water; and the amount of air breathed, water and food consumed, and information on the lifetime activity patterns of relevant populations. The fourth and final component of risk assessment, risk characterization, combines the results of the dose-response and exposure assessments. It also includes a summary of the strengths and weaknesses of the assessment, including its scientific uncertainties. The risk characterization is intended to provide decisionmakers with the essential information concerning magnitude and distribution of the risk from a particular source or class of sources of the carcinogen that they need in order to make risk management (standard-setting) decisions.

EPA applies equivalent risk-assessment principles and procedures to radiation from radionuclides. For example, on the basis of extensive human studies EPA categorizes radiation as a known human carcinogen. Also, for radiation, as for most chemical carcinogens, EPA assumes a nonthreshold, linear dose-response relationship.²⁸ In other words, EPA generally assumes that there is a degree of risk associated with any exposure to a carcinogen and that this risk varies linearly with the level of exposure. Risk assessments for radiation and chemicals exhibit some significant differences, [25 ELR 10660] however, due to differences in the data available. For radiation, there are several large, long-term human studies of effects on health that cover a broad range of exposure under a variety of situations. These human studies include (1) the Hiroshima and Nagasaki atomic bomb survivors; (2) occupationally exposed underground miners in the United States, eastern Europe, Sweden, and Canada; (3) U.S. radium dial painters; and (4) several groups of patients exposed to large doses of radiation for medical purposes.²⁹ The data from these studies can generally be applied to all exposures to radionuclides.³⁰

There are currently about three dozen chemicals for which enough human data exist to classify them as known human carcinogens, and this human data is sufficiently detailed to quantify the dose-effect response for several of these chemicals (arsenic, benzene, chromium VI, and nickel subsulfide). For most chemicals, however, the lack of human data requires reliance on findings of cancer incidence through exposure of laboratory animals to relatively large amounts of a given chemical. Further, in contrast to the situation for radionuclides, findings on carcinogenicity for one chemical are not generally applicable to others unless chemical similarity and structure-activity relationships can be relied on.³¹

When chemical risk assessments are based on the inference of human carcinogenicity from animal studies, there are necessarily greater uncertainties involved in the assessment of risk than is the case for radiation. In addition, even in those cases where human evidence is available, the database for a given chemical is generally far more limited than that for radiation. Finally, for quantitative estimates of the risk from chemical carcinogens, EPA generally uses a linear multistage model from which toxicity values (unit risks) are calculated as plausible upper limits.³² As such, risk estimates for many chemical carcinogens are properly characterized as "upper bound" estimates (e.g., estimates of the 95 percent confidence level). In contrast, because of the availability of extensive human data from long-term epidemiological studies, unit risks for radionuclides can be expressed as statistical "best estimates" of cancer mortality. As a result of the generally more conservative nature of risk assessments for chemicals, the actual degree of protection at a given level of "assessed" risk will tend to be much greater for chemicals than it is for radiation.

The "Traditional" Approach to Radiation Protection

The principal source of nongovernmental guidance on protection from the hazards of radiation is the ICRP.³³ This body, which draws its membership from the world's leaders in radiation protection policy and in scientific aspects of risk assessment for radiation, has operated continuously since its formation in 1928. It provides the principal external source of guidance for radiation protection policies or regulations of all major international bodies, most notably the International Atomic Energy Agency, the World Health Organization, and a variety of U.N. organizations, as well as for the national radiation protection programs of most countries, including the United States. The ICRP views the primary aim of radiological protection to be to "provide the appropriate standard for protection for man without unduly limiting the beneficial practices giving rise to radiation exposure."³⁴ Its recommendations for protection of members of the public from radiation distinguish two broad categories of exposure situations, which it designates as "practices" and "intervention."

ICRP Recommendations for Public Exposure From Practices

The ICRP defines "practices" as human activities that increase overall human exposure to radiation by introducing whole new blocks of sources, pathways, and individuals, or by modifying the pathways to human beings from existing sources. Emphasis is on prior control of the source, which can generally be planned for before commencing the practice. The ICRP system of protection for practices is based on three general principles, which must be satisfied independently.

The first principle is that no practice that involves radiation exposure should be adopted unless it produces sufficient benefit to exposed individuals or to society to offset the radiation detriment³⁵ it produces. This principle is called the "justification of a practice."³⁶ The determination that a particular activity involving radiation exposure of the general public is justified is usually not made in the United States by those regulatory agencies directly responsible for radiation protection decisions. This balancing decision is a complex process involving congressional, executive, and judicial inputs. One vehicle for assisting such decisionmaking is the National Environmental Policy Act (NEPA),³⁷ [25 ELR 10661] which requires federal agencies to assess and consider the environmental impacts of any major federal action. Another vehicle is the regulatory analysis process federal executive agencies pursue in issuing standards.³⁸

The second principle, called the "optimization of protection," is that the radiation detriment (usually expressed in terms of the collective exposure of populations) from individual sources should be maintained "as low as is reasonably achievable," taking economic and social factors into account. In its most straightforward application, optimization consists of minimizing the exposure of populations through balancing the benefits of control against their costs. The ICRP also recognizes the use of multivariate analysis for optimization; this is an analytical procedure in which the various factors involved in a judgment can be given different weights to reflect the decisionmaker's priorities. The "main aim" of optimization for public exposure, according to the ICRP, should be to "develop practical restrictions on the sources of exposure, e.g. in the form of restrictions on the release of radioactive waste to the environment."³⁹

The final principle, the "limitation of individual risk," is that the maximum risk to individuals resulting from the combined exposure to all relevant practices should be subjected to an upper limit. The ICRP chose this "individual" limit with the view that any exposure above the limit would result in risks from practices that would be deemed "unreasonable" in normal circumstances. The ICRP further recognizes an important distinction between limitation of the maximum exposure of an individual from all sources combined and the control of each source of exposure. This distinction is made because an individual may be subject to exposure from several sources, and also because it may not be equitable, in the ICRP's view, for an individual to be subject to exposure up to the limit from a single source, since the benefits of a specific source do not necessarily flow to the individual most highly exposed to that source.

For these reasons, the ICRP recommends that upper bounds be established for "optimization" of exposure from individual practices in the form of "constraints" that are below the limit on maximum exposure of individuals.⁴⁰ That is, the ICRP recommends that the range of doses available for satisfying the principle of optimization be restricted to values below a dose constraint, chosen to be a fraction of the limit on individual exposure applicable to all sources of exposure combined.

In its more recent recommendations, the ICRP recommends an individual limit of 1 millisievert (mSv), or 100 millirem (mrem), per year from all relevant practices combined, after considering the acceptability of the risks (of fatal and nonfatal cancer and genetic effects) in general, as well as in relation to variations in background radiation, which averages about 1 mSv, excluding indoor radon.⁴¹ This limit corresponds to a lifetime risk of 3 x 10<-3>. Importantly, although the ICRP found that exposure from background was not "unacceptable," it noted that background radiation is not harmless and that even small radiation doses may produce deleterious effects. Further, the ICRP found that although the component of public exposure to natural sources is by far the largest, "this provides no justification for reducing the attention paid to smaller, but more readily controlled, exposures to artificial sources."⁴²

ICRP Recommendations for Intervention in Public Exposure

The ICRP defines intervention situations as human activities that act to decrease human exposure to preexisting sources of radiation that do not result from the controlled operations of practices. Typical interventions include the response to major accidents (e.g., through evacuation or the restriction of food); imposition of requirements for remediation of contamination that is due to historical practices (i.e., those not addressed by a current system of control for practices); and measures to reduce exposure due to preexisting natural sources, such as indoor radon in existing structures. The system of protection through intervention is based on two principles, similar to their counterparts for practices, which are applied, however, to the intervention itself instead of to the source of exposure.

The first of these principles, "justification of intervention," requires that the intervention itself should do more good than harm—the reduction in the radiation detriment should be enough to justify the harm, including the cost, of the intervention. Under the second principle, "optimization of intervention," the form, scale, and duration of the intervention is chosen so as to obtain the maximum net benefit.⁴³ The application of intervention is not subject to a principle requiring limitation of individual risk, as the control of practices is, because in some circumstances it would be possible to do more harm than good if predetermined limits were established, especially in cases where intervention involves severe measures, such as involuntary evacuation or the restriction of foods.

In perhaps the most important application of intervention principles, the ICRP recognized the complex problems involved in controlling exposure to indoor radon, which is by far the single largest source of exposure to radiation. It recommended that the choice of an action level for radon should depend not only on the level of risk to individuals, but also on the likely scale of action required and its economic implications for communities and individual homeowners.⁴⁴ The ICRP also stated that the choice of a particular national action level may be that which defines a significant, but not unmanageable, number of houses in need of remedial work.⁴⁵

EPA Standards for Radionuclides and Chemical Carcinogens

Radiation Protection Standards Under the AEA

Reorganization Plan No. 3 of 1970 transferred to EPA the authority of the former AEC to establish "generally applicable [25 ELR 10662] environmental standards for the protection of the general environment from radioactive material."⁴⁶ Pursuant to this general authority and other radiation-related statutes, EPA has promulgated standards that apply to the nuclear fuel cycle, the cleanup and disposal of wastes from the processing of uranium and thorium ores, and the storage and disposal of high-level and transuranic radioactive wastes. As noted earlier, the AEA does not specify conditions governing the establishment of standards. The Agency was therefore free, with a few exceptions noted later, to choose the risk management policy it used in establishing each of these standards.

Uranium Fuel Cycle Standards. EPA's uranium fuel cycle (UFC) standards, promulgated in 1977, apply to the radioactive discharges from nuclear power plants and most other facilities associated with the production of electrical power through the use of nuclear energy.⁴⁷ The risk management approach used in these, the first radiation standards EPA established, included four basic considerations: (1) the total radiation dose to populations; (2) the maximum dose to individuals; (3) the risk of health effects attributable to these doses, including the future risks arising from the release of long-lived radionuclides to the environment; and (4) the effectiveness and costs of the technology available to mitigate these risks through effluent controls.⁴⁸

The Agency included among these criteria the cost effectiveness of controls in order to strike a balance between the need to reduce health risks and the need for electrical power.⁴⁹ The resulting UFC standards contain two components. The primary standard is expressed in terms of annual limits on individual doses to members of the public. However, as a result of the analysis of total radiation dose to populations, coupled with an examination of the costs and efficacy of controls, the standards contain additional restraints on risks to current as well as future populations⁵⁰ by providing limits on the quantities of certain long-lived radioactive materials released to the general environment.⁵¹

In setting the UFC standards, EPA did not articulate or target a specific numerical risk range or relate the standards to background radiation. The Agency did note, however, that the standards were below (by a factor of 20), and were intended to supplement, the then current federal guidance (and ICRP limit) on annual dose (500 mrem/year, which was then estimated to correspond to a 6 x 10<-3> lifetime risk).⁵² It also commented that background exposure was essentially different from man-made exposure, in that background exposure was unavoidable.⁵³ Overall, the standards represented the lowest radiation levels at which the Agency determined that the costs of control were justified by the reduction in total health effects in populations.⁵⁴ At the time they were established, the resulting standards represented estimated maximum lifetime individual risks of 3 x 10<-4> or less, depending on the type of release and facility involved.⁵⁵ Although the individual risks permitted by these standards are now estimated to be slightly higher than the risks posed under most of EPA's other radiation standards, it should be noted that the UFC standards limit the combined risks from all radionuclides released from most components of the entire nuclear fuel cycle and address the combined risks through all environmental media.⁵⁶

Disposal and Cleanup Standards for Uranium and Thorium Mill Tailings. Congress enacted the Uranium Mill Tailings Radiation Control Act (UMTRCA)⁵⁷ as an amendment to the AEA to provide for the "stabilization, disposal, and control in a safe and environmentally sound manner" of uranium and thorium mill tailings to "prevent or minimize radon diffusion into the environment and to prevent or minimize other environmental hazards from such tailings."⁵⁸ In establishing standards, UMTRCA required EPA to "consider the risk to the public health, safety, and the environment, the environmental and economic costs of applying such standards, and such other factors as the Administrator determines to be appropriate."⁵⁹ In addition, it specifically required that these standards be "consistent" with the requirements of the Resource Conservation and Recovery Act (RCRA)⁶⁰ to the "maximum extent practicable."⁶¹ In 1983, EPA promulgated standards that apply to the 24 inactive, closed uranium processing sites that are managed by DOE (Title I sites) and the 26 uranium sites licensed by the NRC (Title II sites).⁶²

Although EPA evaluated both maximum individual risk and cumulative effects on health in populations, as in the case of the UFC standards EPA did not articulate or target a predetermined numerical individual risk level or range and did not base its conclusions on comparisons to the risks from background radiation.⁶³ Instead, the Agency's objective was to set standards that "take into account the tradeoffs between costs and benefits in a way that assures adequate protection of the public health, safety, and the environment; that can be implemented using presently available [control and survey] techniques . . .; and that are reasonable in terms of overall costs and benefits."⁶⁴ The most important hazardous constituent of uranium mill tailings is radium, which produces radon gas.⁶⁵ EPA concluded that the most reasonable course of action was to mitigate the health and environmental hazards of radium through requiring use of the best available technology (BAT). That is, the Agency found that the achievable degree of protection from radon was limited by the availability of controls that were feasible to implement, and, further, that use of BAT yielded an acceptable level of risk reduction.

The Title I and II standards each address two situations—disposal and cleanup. The disposal standards deal with the long-term control of radium and hazardous chemicals in uranium and thorium mill tailings piles.⁶⁶ In light of the long half-life of radium (1,600 years), EPA's primary objective for the disposal standards was to isolate and stabilize the piles to prevent the release of radon, misuse of tailings by humans, and dispersal by natural forces for the longest feasible period of time, which the Agency concluded was 1,000 years. The standard was also designed to reduce gamma radiation from the tailings. The standards generally limit emissions of radon from the piles to a lifetime individual risk level of about 10<-4>. This is consistent with the CAA standards enacted subsequently for the Title II sites.⁶⁷ The standards also require that both radioactive and chemical contaminants in ground and surface water conform to the Agency's requirements established under RCRA; these specify, inter alia, that underground and surface sources of drinking water must satisfy the drinking water standards established under the SDWA.⁶⁸

The cleanup standards address soil and water contaminated by tailings. Primary objectives were to reduce elevated indoor radon levels in existing buildings where tailings had been used as fill, to avoid high levels of indoor radon in new buildings, to reduce gamma radiation exposure from windblown surface contamination, and to clean up contaminated groundwater. The soil cleanup standards were set at levels that were high enough to allow the presence of radium tailings to be distinguished from background levels of radium, given the practical limits of the available field detection technology at the time.⁶⁹ The risk levels associated with these soil cleanup standards under actual conditions are difficult to quantify, but lie in the range of 10<-3> to 10<-4> lifetime risk. The indoor radon decay product standard was set at a level consistent with the current Agency recommendations for remediation of indoor radon. Cleanup requirements for groundwater are the same as those discussed for disposal.

Importantly, the Agency noted that the standards were not meant to establish precedents for other situations or regulations involving similar environmental objectives but different economic and/or technological circumstances.⁷⁰ The standards were only meant to address the cleanup of the uranium and thorium mill tailings sites, most of which are in sparsely populated, predominantly semiarid areas of the West with high background levels of radium.

Disposable Standards for High-Level and Transuranic Radioactive Wastes. EPA's high-level and transuranic radioactive waste disposal standards⁷¹ were developed pursuant to its general AEA authority and the Nuclear Waste Policy Act,⁷² which directed EPA to "promulgate generally applicable environmental standards for protection of the general environment from offsite releases from radioactive material in [high-level and transuranic waste] repositories."⁷³ The standards were first promulgated in 1985⁷⁴ and the individual and groundwater protection requirements were later repromulgated in 1993.⁷⁵

[25 ELR 10664]

EPA's risk management approach to the disposal standards included several considerations: (1) the expected technological capabilities of disposal technologies and repositories to isolate the wastes; (2) consistency with other related environmental standards for radiation exposure; and (3) the acceptability of the residual risks in comparison with risks that would have existed in the absence of the practices that produced the wastes.

The primary standards are the containment requirements. These define the required degree of isolation of wastes from the accessible environment by limiting the total quantities of specified radionuclides released over a period of 10,000 years. These release limits and this time frame were chosen to reflect the technological limits of the best currently available disposal technology. The Agency also evaluated health risks to future generations and projected that disposal in compliance with the containment requirements would cause no more than 1,000 premature cancer deaths over the entire 10,000-year period from the future disposal of all existing high-level wastes and most of the future wastes expected to be produced by all currently operating reactors—an average of 0.1 fatality per year. As part of its judgment of the acceptability of this level of impact on future generations, the Agency noted that this risk was comparable to the risk that those generations would have faced if the natural uranium ore used to create the wastes had never been mined.⁷⁶

While the containment requirements limit the effects of these wastes on entire populations, the supplementary individual and groundwater protection requirements limit the maximum risk to individuals in the vicinity of the repositories. The individual radiation protection requirement⁷⁷ corresponds to a lifetime cancer risk level of approximately 5 x 10<-4>. Although this risk is slightly higher than the risks associated with many other Agency standards, the standard was found to be acceptable for the following reasons. First, there is a very small number of potential sites involved and these are located in relatively remote areas. As a result, only a small number of persons could be exposed to the maximum allowed individual risk. Second, the standard considers the combined exposures delivered through all environmental media from all radionuclides. Finally, the limit is an appropriate apportionment, for the practice of high-level waste disposal, of the ICRP recommendation for maximum exposure of individuals.⁷⁸ As in the case of the UMTRCA standards, the standard contains specific groundwater protection requirements: Disposal of radioactive waste in repositories may not cause levels of radioactivity in underground sources of drinking water that exceed the radionuclide maximum contaminant levels (MCLs) established under the SDWA.⁷⁹

Air Emission Standards Under the Clean Air Act of 1977

Section 112 of the Clean Air Act of 1977 required EPA to set national emission standards for hazardous air pollutants (NESHAPs) that "protect the public health" with an "ample margin of safety."⁸⁰ In Natural Resources Defense Council v. U.S. Environmental Protection Agency (the so-called Vinyl Chloride case),⁸¹ a federal court interpreted this section to require the Agency to set the NESHAPs on the basis of a two-step process: First, to determine a "safe" or "acceptable" level of risk considering only the health risk, and then to set an emission standard no higher than this level of risk that provides "an ample margin of safety" in which costs, feasibility, the degree of uncertainty in the risk estimates, and other relevant factors may be considered.⁸² Although the court declined to determine what risk level is "acceptable" or to set out a method for its determination, the court found that "safe" does not require the elimination of all risk and directed the Administrator to determine a "safe" or "acceptable" level of risk based on a judgment of "what risks are acceptable in the world in which we live."⁸³

To aid in this inquiry, EPA compiled a Survey of Societal Risks.⁸⁴ Individual lifetime risks identified in the survey range from 10<-1> to 10<-7>. EPA concluded that no single factor could be identified that defined acceptability under all circumstances, and that the acceptability of a risk depends on a variety of factors, including certainty and severity of the risk; reversibility of the health effect; familiarity of the risk; whether the risk is voluntarily assumed (or the individual receives a direct benefit for accepting the risk voluntarily); benefits of the activity; and the risk and advantages of any alternatives. EPA selected a limit of 10<-4> on maximum individual risk (MIR) to an individual over a lifetime, finding this level to be within the survey range and to provide health protection at a risk level lower than many other risks common "in the world in which we live."⁸⁵

In the survey, EPA found that average lifetime risks from exposure to background radiation range from 10<-3>, excluding indoor radon, to as high as 10<-2> when it is included. Although EPA recognized that the amount of radiation most people receive from industrial sources is typically lower than that from background radiation, the Agency concluded that doses from man-made sources "can still be significant" and that a "source does not present an acceptable risk simply by being less than natural background."⁸⁶ Therefore, comparison of the MIR to largely uncontrollable background risks from radiation was not considered to be relevant. This is because EPA decided, consistent with its earlier rulemakings under the AEA, that it was the incremental risk above background from a particular regulated activity that was important in establishing the MIR.⁸⁷

[25 ELR 10665]

The second step in the selection of a standard is the determination of an "ample margin of safety." This step considers additional health information, such as the total impact on health in the population and the uncertainty in the risk estimate, along with other factors such as cost and technological feasibility. In carrying out this step, EPA also took as a specific objective protection of the greatest number of persons at a lifetime risk level no greater than 10<-6>.

This risk management policy was applied to benzene⁸⁸ and radionuclide⁸⁹ sources in 1989.⁹⁰ EPA's standards for benzene and radionuclide sources were chosen to ensure that the risk from a source did not exceed an MIR⁹¹ of 10<-4> and that the risk did not exceed 10<-6> for the greatest number of persons reasonably achievable through the application of emission controls.⁹² In establishing the radionuclide standards, the Agency analyzed the control costs and savings in health effects projected in exposed populations for a range of different levels of control, and found in all cases that satisfying the MIR also satisfied the "ample margin of safety" criterion.⁹³ As in the case of the radiation standards established under the AEA, if a source emits more than one radionuclide these CAA standards apply to the combined exposure from all radionuclides to ensure that the specified objectives are met for the combined risks from all radioactive emissions from the source.⁹⁴

The SDWA Standards

The SDWA requires EPA to publish MCL goals (MCLGs) for drinking water contaminants that "may have any adverse effect on the health of persons and [that are] known or anticipated to occur in public water systems."⁹⁵ The SDWA requires the MCLGs to be set "at the level at which no known or anticipated adverse effects on the health of persons occur and which allows for an adequate margin of safety."⁹⁶

At the same time that EPA publishes an MCLG, which is an unenforceable health goal, it must also promulgate a national primary drinking water regulation (NPDWR) for a contaminant that either (1) establishes an MCL or (2) requires use of a specified treatment technique if it is not "economically or technologically feasible to ascertain the level of a contaminant."⁹⁷ An MCL must be set as close to the MCLG as is "feasible." The SDWA defines feasible as "with the use of the best technology, treatment techniques and other means which the Administrator finds, after examination for efficacy under field conditions and not solely under laboratory conditions, are available (taking cost into consideration)."⁹⁸

Contaminants for which strong evidence of human carcinogenicity exists are classified as "Category I." For these contaminants, EPA assumes that there is no threshold below which exposure poses no risk and therefore sets the MCLGs at zero. MCLs are then determined based on a number of factors: (1) the capability of analytical laboratories to measure the contaminant reliably and consistently; (2) the availability, performance, and cost of treatment technologies for removing the contaminant; and (3) the health risk associated with the contaminant levels under consideration.⁹⁹

To consider measurability, a quantity called the practical quantitation level (PQL) is determined. The PQL reflects the contaminant concentration that can be measured by good laboratories under normal operating conditions within specified limits of precision and accuracy. In determining treatment technologies, the Agency identifies those having the highest removal efficiencies, that are compatible with other water treatment processes, and that are affordable for use by regional and large metropolitan public water systems. In some cases, the Agency also evaluates the total national costs for each contaminant in light of the number of systems that would have to install treatment in order to comply with the MCL.¹⁰⁰

Finally, the Agency considers the health risks associated with various contaminant levels to ensure that the MCL will adequately protect public health. EPA has historically used an individual lifetime reference risk range for carcinogens in drinking water of 10<-4> to 10<-6>.¹⁰¹ Where sufficient occurrence data is available, EPA also evaluates the number of cancer cases avoided in relevant populations as a function of the level of the MCL.¹⁰²

Generally, the Agency has used this approach for both [25 ELR 10666] chemical carcinogens and radionuclides.¹⁰³ Chemical carcinogens and radionuclides are Category I contaminants with MCLGs of zero¹⁰⁴ and MCLs generally fall within the range 10<-4> to 10<-6> lifetime risk.¹⁰⁵ In 1977, the Agency set MCLs for radium-226 and radium-228,¹⁰⁶ and in 1991 proposed MCLs for uranium and radon.¹⁰⁷ The Agency also has established a single MCL for the broad category of all of the less prevalent man-made radionuclides.¹⁰⁸ However, the radionuclide MCLs focus, for the most part, on individual naturally occurring contaminants that commonly occur in public drinking water systems, and do not generally take into account co-occurrence of several radionuclides.¹⁰⁹ Similarly, EPA set separate standards for individual organic and inorganic chemical contaminants and for microbiological contaminants.¹¹⁰

Cleanup Standards Under CERCLA

Under CERCLA, remedial actions for hazardous substances¹¹¹ must (1) protect human health and the environment¹¹² and (2) satisfy standards or other requirements of federal or state laws that are found to be "applicable" or "relevant and appropriate under the circumstances" (ARARs).¹¹³ EPA's approach to cleanups is set out in the National Contingency Plan (NCP).¹¹⁴ First, the Agency conducts a baseline risk assessment of a site to determine whether cleanup action is warranted.¹¹⁵ Generally, cleanup action is deemed to be warranted if this assessment indicates that cumulative risk from a site to an individual exceeds a lifetime¹¹⁶ cancer risk of approximately 10<-4>.¹¹⁷ However, the existence of ARARs that define lower acceptable risk levels for specific substances also may lead to a decision that remedial action is warranted.¹¹⁸

If remedial action is found to be warranted, the second stage is to establish site-specific preliminary remediation goals (PRGs) that are "protective of human health and the environment."¹¹⁹ For known or suspected carcinogens, acceptable PRGs are concentration levels that satisfy ARARs or represent an individual lifetime risk between 10<-4> and 10<-6>.¹²⁰

Technical factors, including detection or quantification limits for contaminants, technical limits to remediation, the ability to monitor and control movement of contaminants, and background levels of contaminants, can enter into the determination of where within the risk range the cleanup goal for a given contaminant at a site will be established.¹²¹ Although background levels have been used as a basis for cleanup goals,¹²² cleanups are not always required to achieve background levels. In some cases, ambient background levels are not necessarily "protective of human health and the environment," such as in urban or industrial areas. In other cases, cleaning up to background levels may not be necessary to protect human health, because the background level for a particular contaminant may be close to zero, as in pristine areas.¹²³

The risk level selected within the risk range does not generally depend on the size of the population, because the Agency has decided that the limitation of individual risk should be consistent across all CERCLA sites.¹²⁴ Also, since it is often difficult to determine the size of future exposed populations at specific sites, CERCLA implementation decisions [25 ELR 10667] have not generally taken population size into account. However, the size of the currently exposed population can affect the Agency's decision to place a site on the national priorities list (NPL), which makes it eligible for Superfund-financed remedial action.¹²⁵

After selecting the PRGs, the Agency develops alternative plans to address the risks identified in the baseline risk assessment and records the final remedy selection in a record of decision (ROD).¹²⁶ Under the NCP, the criteria for evaluation of different remedial alternatives at a site are (1) overall protection of human health and the environment; (2) compliance with ARARs; (3) long-term effectiveness and permanence; (4) reduction of toxicity, mobility, or volume through treatment; (5) short-term effectiveness; (6) implementability; (7) cost; (8) state acceptance; and (9) community acceptance.¹²⁷ EPA generally uses the same approach for radionuclides as for chemicals in its cleanup actions under CERCLA.¹²⁸

The 1992 Citizens guide to Radon

The "national long-term goal" of the IRAA is that "air within buildings in the United States should be as free of radon as the ambient air outside of buildings."¹²⁹ Importantly, in order to make "continuous progress toward the long-term goal," the IRAA does not require EPA to promulgate standards. Instead, the IRAA uses a nonregulatory approach that directs EPA to publish A Citizen's Guide to Radon that contains a series of action levels indicating the health risks associated with different levels of radon exposure and also provides information regarding the cost and technological feasibility of reducing radon concentrations within new and existing buildings.¹³⁰

In the 1992 Citizens Guide, EPA chose a primary "action level" of 4 picocuries per liter (pCi/L) as the point above which mitigation is always advised to reduce indoor radon in existing buildings.¹³¹ This level corresponds to an "average" lifetime individual risk of approximately 10<-2> for the general population (combining the much greater risks from smokers with those for nonsmokers).¹³² Importantly, the Agency does not consider this to be a "safe" or "acceptable" level of exposure and emphasizes in the Guide that since significant health risks exist below the action level, mitigation is also valuable at lower levels.

EPA considered a number of alternative action levels, but ultimately recommended 4 pCi/L based on research showing that most homes can be readily mitigated to levels below 4 pCi/L. Available technology is progressively less able to reduce radon levels to lower levels, although an estimated 70 to 80 percent of homes with elevated radon levels would be able to achieve an action level of 2 pCi/L (which corresponds to an average lifetime risk of 7 x 10<-3>).¹³³ The choice of a 4 pCi/L level is also supported by research showing that it is more difficult to accurately measure radon at lower levels. Based on these considerations, EPA recommends 4 pCi/L as the level at which action is always indicated, but advises homeowners that they should also consider mitigating radon in homes that have confirmed levels between 2 and 4 pCi/L.

Comparing ICRP Recommendations With EPA's Radiation Standards

Comparing the risk management policies that EPA used to establish its radiation standards with the traditional approach to radiological protection, as exemplified by the recommendations of the ICRP, requires care. Simply observing that EPA's radiation standards, which usually fall near lifetime risk levels of approximately 10<-4>, are inconsistent with the ICRP recommendation for maximum dose to an individual, which corresponds to a lifetime risk level of 3 x 10<-3>, is both misleading and inappropriate. EPA's standards apply to individual sources or classes of sources, whereas the ICRP limit is intended to serve as an upper bound on the maximum dose to any individual, now and in the future, from all sources combined. Also, the ICRP recommends that regulatory limits on doses from sources be based on a "constrained optimization," i.e., that optimization of the control of radiation exposure from individual sources be carried out within source-related dose constraints that fall below the ICRP's overall limit. This Article therefore examines, in light of the preceding review of the risk management policies that underlie EPA's radiation standards, the degree to which EPA's radiation standards are consistent with the ICRP's recommendations for the regulation of sources: Specifically, the two ICRP recommendations that standards be based on optimization of the control of sources, and that this optimization be bounded by source-related dose constraints lower than the dose limit for individuals.

The first recommendation, optimization of the control of sources, consists, essentially, of analyzing the options for reduction of health effects (e.g., number of cancer cases) in populations in relation to the cost of control to achieve this reduction, and then requiring the optimal result through a standard. The Agency explicitly carried out such analyses as part of its risk management rationale in establishing standards for the UFC and for uranium and thorium mill tailings.¹³⁴ In the case of the UFC standards, this not only provided the basis for the primary standard for individual dose, but resulted in additional requirements for the control of long-lived effluents.

Optimizing reduction of total health effects in populations in relation to the cost of control of potential releases from the disposal of high-level and transuranic wastes [25 ELR 10668] posed a unique problem. Since the hazards from these wastes persist for many thousands of years, it is difficult to quantify either the anticipated impacts on populations or the efficacy of control. For this reason, the Agency opted to require the best containment reasonably achievable. EPA also compared the risks from disposal with the risks that would have been present if the ores (that eventually enabled the practices that produced the wastes) had never been mined, to reach a judgment on optimization. In this case, the optimization was more akin to a multivariate than to a cost/benefit analysis.

In the second step of the CAA standard-setting process, EPA considered further reductions in health effects in populations as a function of costs, as well as other factors, and found that its individual risk goal of 10<-4> also satisfied its goal to protect the greatest number of persons possible to a risk of not more than 10<-6>. That is, the Agency, among other tests, analyzed and considered the cost effectiveness of further reductions below its risk goal, but concluded that none are justified. In the proposed radionuclide drinking water MCLs, the Agency evaluated the number of cancer cases avoided in relevant populations and used "limited" consideration of the cost effectiveness of requiring additional increments in technology or contaminant control.¹³⁵ However, the determining criterion under the SDWA is individual risk and the main cost consideration is whether a technology is affordable to municipal water systems. Similarly, CERCLA PRGs focus primarily on individual risk, although optimization can enter into CERCLA cleanup decisions through ARARs that were developed to consider population risks in relation to costs, and cost is one of nine factors in selecting between remedial alternatives. In addition, the size of the exposed population is a factor the Agency considers in placing a site on the NPL for expedited action.

The second recommendation requires that the optimization of source-related standards be bounded by source-related dose constraints that are below the total dose limit for individuals. Although EPA did not use predetermined source-related constraints on individual dose in establishing its UFC and UMTRCA standards, in both cases it examined the results of the risk management analyses it carried out (essentially a cost/benefit analysis) to assure that the resulting source standards were below existing limits on exposure of individuals. That is, although source-specific constraints were not applied as a prior constraint on these analyses, EPA assured itself that the resulting source standard fell below the then current individual dose limits. EPA carried this further in establishing its standards for high-level and transuranic wastes, and noted that the dose limits for these wastes satisfied the ICRP's recommendation to establish dose constraints for particular activities.¹³⁶

In the standards established under the CAA and SDWA, and the PRGs adopted for use under CERCLA, the Agency used a predetermined risk goal of 10<-4> to 10<-6> lifetime risk. Although this risk goal is lower than required to minimally satisfy the ICRP objectives for source-related dose constraints, it nonetheless does not satisfy those objectives, and in practice most of EPA's radionuclide standards for sources have fallen on the high side, i.e., 10<-4>, of this risk goal. Thus, although EPA has not always used a predetermined constraint on individual risk, and has seldom explicitly related use of such a constraint to the ICRP recommendations, the resulting standards have invariably satisfied the ICRP recommendation that standards governing individual sources be established at levels that are below the ICRP limit on individual dose.

Therefore, with few exceptions, EPA, although not usually setting out to conform to the ICRP recommendations, does, in effect, manage risk consistently with the ICRP's recommendations for practices. EPA's recently proposed guidance to federal agencies on formulating regulations and conducting programs to protect the general public from exposure to ionizing radiation underscores this conclusion.¹³⁷ The guidance recognizes that in the last 30 years our knowledge of the effects of radiation has become more definitive and the extent of our use of radiation has grown. Before 1960, major uses of radiation were limited, and the primary concern of radiation protection was to ensure that doses to those few individuals who were affected did not exceed limits on dose from all sources combined. Since then, the numbers and types of man-made sources of radiation have increased and public concern about environmental contaminants of all kinds has begun to influence regulatory decisionmaking. A primary result of this concern has been that attention is now focused on the control of specific sources of exposure, and legislative requirements are far more specific as to how effective this control must be.¹³⁸

As a result, the guidance not only recommends a radiation protection guide of 100 mrem per year that serves as an upper bound on doses from all sources, but also recommends that limits for individual sources be restricted to a fraction of this overall dose limit, consistent with the recommendations of the ICRP. The guidance sites EPA's standards under the AEA, the CAA, and the SDWA as examples of acceptable application of the latter criterion.¹³⁹

EPA is also consistent with the ICRP in its consideration of exposure due to background sources of radiation. The ICRP chose an overall limit of 100 mrem after considering the average dose from background and recognizing that the public receives a greater dose from background than from artificial sources. However, it also recognized that exposure to background is unavoidable and therefore does not justify reducing the attention paid to smaller, but more readily controlled artificial sources of radiation exposure. In recommending 100 mrem, EPA also considered risks from background exposure as providing a perspective for its proposed limit. However, background exposure did not provide a justification for the proposed limit, since background represents an uncontrollable source of risk while the proposed limit is focussed on the incremental, voluntary exposures from man-made activities. The Agency took the same position in establishing the UFC and UMTRCA standards [25 ELR 10669] under the AEA and the radionuclide standards under the CAA.¹⁴⁰

EPA's risk management policy for establishing cleanup standards (PRGs) under CERCLA corresponds in some respects to the ICRP's treatment of practices, in that under CERCLA a risk-limiting criterion is applied and it can be argued that the multiplicity of factors considered—e.g., practicality of measurement, efficacy and practicability of remedial actions, cost, and community acceptance—constitutes a multivariate approach to optimization. However, there is an important difference. The ICRP treats some major classes of cleanup situations as interventions (all those that cannot be considered as practices, such as continuing exposure to historical contamination and contamination resulting from an accident) and therefore it does not impose any risk constraint. For these kinds of situations the ICRP recommendations can, therefore, lead to a considerably higher level of residual risk than would be permitted under CERCLA.¹⁴¹

Finally, EPA and the ICRP clearly address control of the hazards posed by indoor radon using essentially identical risk management approaches. Both organizations agree that these risks should not be treated in the same manner as other risks due to the greater and more complex social and economic impacts involved, as well as the technological limitations of both testing for and mitigation of radon. Pursuant to its statutory mandate, EPA follows a nonregulatory approach, using action levels to determine when remediation is recommended. The ICRP treats control of exposure to indoor radon as an intervention, and also recommends the use of action levels. It suggests a numerical action level similar to (although marginally higher than) EPA's action level.

Conclusion

Although EPA has not developed a formal statement that articulates an overall framework under which radiological versus chemical risks are addressed, it has been an objective of the Agency to achieve general consistency in its standards for radiation and chemical carcinogens. The standards developed under the AEA, the CAA, the SDWA, and CERCLA generally limit lifetime individual risks from both radiological and chemical carcinogens to no higher than 10<-4> despite the Agency's different statutory mandates, social and economic concerns, and other factors. Further, although there are significant differences in the human epidemiological data available for radiation and chemicals, EPA develops standards for these two sets of carcinogens using the same risk-assessment principles and procedures. Nonetheless, it is noteworthy that because of the much smaller database for chemical carcinogens, risk assessments for chemicals must generally be more conservative than those for radiological carcinogens.

The Agency's radiation standards for sources are also generally consistent with the recommendations of the ICRP for control of practices, i.e., activities using radiation sources. This paradigm specifies that limits on individual sources should be constrained to values below its overall limit on individual exposure, which corresponds to a lifetime risk for 3 x 10<-3>. EPA's approach of setting source- and media-specific limits has resulted in levels that make it highly likely that an individual exposed to a number of sources will satisfy the overall ICRP limit. Further, the Agency's nonregulatory approach to indoor radon involving the use of action levels is consistent with the recommendations of the ICRP for intervention. There are, however, significant differences in the approach to cleanup—the Agency's standards almost invariably conform to a risk limit, whereas important categories of contamination, e.g., those due to accidents and historical events, are not subject to individual risk limits under the ICRP paradigm.

EPA's standards treat background contamination from radionuclides the same way they treat background contamination from chemical carcinogens which, it is commonly not recognized, can, like background radiation, also present cumulative lifetime risks on the order of 10<-3>.¹⁴² EPA's treatment of background contamination in the promulgation of standards is consistent with the recommendations of the ICRP. The standards seek to reduce the incremental risks from radionuclides and chemical carcinogens independently of whether this results in limits that may be small compared to background exposures. However, background contamination from both chemicals and radionuclides and other technical factors, including detection limits and technical limits to remediation, can influence standards and the ability of the Agency to achieve its risk goals.

EPA handles the tradeoffs between reduction of total health effects in exposed populations and the costs of that control in a variety of ways, not all of which are consistent. This is an area in which a more consistent approach could lead to more consistent and defensible outcomes. At one extreme, in EPA's UFC standards, the tradeoff was evaluated explicitly and directly. At the other extreme, under CERCLA, EPA does not generally consider population risks, although the size of the exposed population can influence the decision whether to place a site on the NPL.

Finally, there are also differences in the manner in which the standards address risks from multiple contaminants. The traditional radiation approach usually involves limiting the total dose from a source and therefore considers the combined exposure from all radionuclides.¹⁴³ EPA's radionuclide standards promulgated under the AEA most directly follow this approach, because they consider the risks from all radionuclides of concern released to all environmental media. Similarly, EPA's [25 ELR 10670] cleanup standards for radionuclides under CERCLA consider the total risk from all radionuclides released to the environment. Although they deal only with air emissions, EPA's standards for radionuclides under the CAA also control the total dose from all radionuclides emitted from a facility. The drinking water standards established under the SDWA for radionuclides are an exception. Although those for man-made radionuclides are expressed in terms of a combined dose from all such radionuclides, the standards for naturally occurring radionuclides are generally set on a radionuclide-by-radionuclide basis and therefore potentially allow a greater combined risk where multiple contaminants are present.

Similarly, EPA's standards for chemicals usually focus on individual chemicals. For example, EPA's standards under both SDWA and the CAA are set on a chemical-by-chemical basis and do not consider aggregate risks from multiple contaminants. In contrast, cleanup standards for chemicals under CERCLA address the release of all hazardous chemicals from a site into the environment and therefore consider risks from multiple contaminants. In summary, although there are some important exceptions, EPA's radionuclide standards tend to limit the combined effect of all radionuclides, whereas its standards for chemical carcinogens do not—they are established for individual chemicals.

David Overy is an attorney advisor in the Federal Facilities Enforcement Office of the U.S. Environmental Protection Agency (EPA). At the time this Article was being prepared, he was attorney advisor and special assistant to the Director, Office of Radiation and Indoor, EPA. Allan Richardson is Associate Director for Radiation Guidance, Radiation Protection Division, Office of Radiation and Indoor Air, EPA.

This article represents the views of the authors in their individual capacities and does not necessarily reflectthe position of EPA or the U.S. government. The authors thank those who reviewed and commented on this paper, including Margo Oge, Gordon Burley, Bob McGaughy, George Wyeth, Chuck French, John Davidson, Stuart Walker, Robert Fegley, Tim Gill, and Judy Tracy.

1. For histories of nuclear and chemical regulation, see SUSTAINABLE ENVIRONMENTAL LAW: INTEGRATING NATURAL RESOURCE AND POLLUTION ABATEMENT LAW FROM RESOURCES TO RECOVERY, chs. 13, 17 (Celia Campbell-Mohn et al. eds., 1993).

2. 42 U.S.C. §§ 2011-2296.

3. Id. §§ 7401-7671q, ELR STAT. CAA §§ 101-618.

4. Id. §§ 300f to 300j-26, ELR STAT. SDWA §§ 1401-1465.

5. Id. §§ 9601-9675, ELR STAT. CERCLA §§ 101-405.

6. 15 U.S.C. §§ 2661-2671.

7. 7 U.S.C. §§ 136-136y, ELR STAT. FIFRA §§ 2-31.

8. Id. § 136(bb), ELR STAT. FIFRA § 2(bb).

9. 21 U.S.C. §§ 301-392.

10. Id. § 346a(b); see Joseph V. Rodricks et al., Significant Risk Decisions in Federal Regulatory Agencies, 7 REG. TOXICOLOGY & PHARMACOLOGY 307, 308-10 (1987).

11. 21 U.S.C. § 348(c)(3)(A); see NATIONAL RESEARCH COUNCIL, NATIONAL ACADEMY OF SCIENCES, REGULATING PESTICIDES IN FOOD: THE DELANEY PARADOX (1987); Regulation of Pesticides in Food: Addressing the Delaney Paradox, Policy Statement, 53 Fed. Reg. 41104 (Oct. 19, 1988).

12. These are designated as "source, special nuclear, and byproduct" materials. Source material is natural uranium and thorium or their ores. Special nuclear material includes uranium enriched in its fissionable isotopes (U-233 or U-235) and plutonium. Byproduct material is radioactive material produced by fission of special nuclear material or wastes produced by the processing of source material ores. 42 U.S.C. § 2014(e), (z), (aa).

13. On average, radon accounts for approximately two-thirds of U.S. exposure to radiation, exposure to medical x rays is about five times smaller, and exposure to AEA materials is much less than 1 percent of these combined.A third category of minor significance also not covered by the AEA includes artificial radionuclides created in accelerators; these are principally used in medicine.

14. 42 U.S.C. § 2201(b).

15. Congress abolished the AEC in 1974 through the Energy Reorganization Act. Id. §§ 5801-5891. The Act transferred the AEC's nuclear weapons and promotional functions to the Energy Research and Development Administration (ERDA) and created the NRC to take over the AEC's commercial regulatory function. In 1977, ERDA's functions were transferred to the newly created DOE by the Department of Energy Organization Act. Id. §§ 7151-7353.

16. See, e.g., ICRP, 1990 RECOMMENDATIONS OF THE INTERNATIONAL COMMISSION OF RADIOLOGICAL PROTECTION, ICRP PUBLICATION 60 (1991) [hereinafter ICRP PUBLICATION 60].

17. 35 Fed. Reg. 15623 (Oct. 6, 1970) (codified at 5 U.S.C. app. 1) [hereinafter the Plan].

18. The Plan, supra note 17, § 2(a)(6). The Plan also abolished the Federal Radiation Council and transferred its functions to EPA. The council served to assure adequate and consistent regulation of radiation by federal agencies through presidentially approved "Federal radiation protection guidance." Id. § 2(a)(7).

19. Id. § 2(a)(4), (8).

20. EPA also regulates chemical carcinogens under a number of other environmental statutes, including the Toxic Substances Control Act, 15 U.S.C. §§ 2601-2671, ELR STAT. TSCA §§ 2-412; the Solid Waste Disposal Act as amended by the Resource Conservation and Recovery Act (RCRA), 42 U.S.C. §§ 6901-6992k, ELR STAT. RCRA §§ 1001-11012; and the Federal Water Pollution Control Act (FWPCA), 33 U.S.C. §§ 1251-1387, ELR STAT. FWPCA §§ 101-607. These statutes, however, exclude from their coverage source, special nuclear, and byproduct materials. See 15 U.S.C. § 2602(2)(B)(iv), ELR STAT. TSCA § 3(2)(B)(iv); 42 U.S.C. § 6903(5), (27), ELR STAT. RCRA § 1004(5), (27); and Train v. Colorado Pub. Interest Research Group, 426 U.S. 1, 6 ELR 20549 (1976). There are, however, standards under the FWPCA that apply to naturally occurring radioactive materials discharged from mining activities. 40 C.F.R. §§ 440.30-.34 (1994). In addition, RCRA applies to the hazardous portion of radioactive mixed wastes. 42 U.S.C. § 6903(41), ELR STAT. RCRA § 1004(41). For a discussion of EPA's approach to standard-setting procedures for chemicals under these statutes, see Alon Rosenthal et al., Legislating Acceptable Cancer Risk From Exposure to Toxic Chemicals, 19 ECOLOGY L.Q. 269-362 (1992).

21. This concern was most notably expressed by EPA's Science Advisory Board. See U.S. EPA, HARMONIZING CHEMICAL AND RADIATION RISK-REDUCTION STRATEGIES—A SCIENCE ADVISORY BOARD COMMENTARY, EPA-SAB-RAC-COM-92-007 (May 18, 1992).

22. One may distinguish two types of background levels for chemicals and radionuclides. Naturally occurring levels are those due to naturally occurring materials already present in the environment that have not been disturbed by humans, e.g., arsenic and radium in soil, or cosmic radiation. Anthropogenic levels are those due to contaminants in the environment or naturally occurring materials relocated as a result of human activities, e.g., chemical contaminants from automobiles and radioactive fallout from atom bomb testing, or radioactive mining wastes.

23. See Stephen L. Brown, Harmonizing Chemical and Radiation Risk Management, 26 ENVTL. SCI. TECH. 2336-38 (1992); David L. Kocher & F. Owen Hoffman, Regulating Environmental Carcinogens: Where Do We Draw the Line? 25 ENVTL. SCI. TECH. 1986-89 (1991).

24. This Article does not examine worker protection standards. EPA issued new recommendations for protection of workers from radiation that were approved by President Reagan on January 27, 1987. 52 Fed. Reg. 2822 (Jan. 27, 1987).

25. See NATIONAL RESEARCH COUNCIL, RISK ASSESSMENT IN THE FEDERAL GOVERNMENT: MANAGING THE PROCESS (1983).

26. Guidelines for Carcinogen Risk Assessment, 51 Fed. Reg. 33992 (Sept. 24, 1986). In addition to cancer risks, EPA also considers genetic, teratogenic or developmental, reproductive, and other toxicologic effects in its risk assessments.

27. See EPA Guidelines for Exposure Assessment, 57 Fed. Reg. 22888 (May 29, 1992).

28. This assumption is made for chemical carcinogens unless there is convincing evidence that such an approach is inappropriate for a particular agent. See 51 Fed. Reg. 33992, 33997-98 (Sept. 24, 1986).

29. For recent reviews of the health effects of radiation, see NATIONAL ACADEMY OF SCIENCES, COMMITTEE ON THE BIOLOGICAL EFFECTS OF RADIATION, HEALTH EFFECTS OF EXPOSURE TO LOW LEVELS OF IONIZING RADIATION (BEIR V) (1990); UNITED NATIONS SCIENTIFIC COMMITTEE ON THE EFFECTS OF ATOMIC RADIATION, SOURCES, EFFECTS, AND RISKS OF IONIZING RADIATION (1988).

30. All radionuclides affect humans in the same manner, through radioactive decay that transfers energy to human tissue and, presumably, eventually causes cancer through damage to DNA or RNA. Although approximately 2,300 nuclides have been identified, most of which are radioactive, there are only three principal modes of radioactive decay: Alpha, beta, and gamma radiation. These three modes of decay involve only two substantially different patterns of energy deposition in tissue, designated as high and low linear energy transfer (or LET). Information on risks from each of these two types of energy deposition can be used to estimate risks associated with specific radionuclides. See THE NATIONAL COUNCIL ON RADIATION PROTECTION AND MEASUREMENTS, COMPARATIVE CARCINOGENICITY OF IONIZING RADIATION AND CHEMICALS, REPORT NO. 96 (1989).

31. Chemicals vary greatly in molecular structure, in their uptake and distribution in the body; in their metabolism, toxification, and detoxification; and in their mechanisms of action and the stage in the process at which they assert carcinogenic effects. Therefore, knowledge of the carcinogenicity of one chemical usually cannot be transferred to another. Id. In some cases, however, chemical similarity and structure-activity relationships are well enough known for scientists to make cross-chemical inferences.

32. 51 Fed. Reg. 33992, 33998 (Sept. 24, 1986).

33. Governmental guidance in the United States is provided by EPA under its federal guidance authority. See supra note 18. This guidance draws heavily on recommendations of the ICRP.

34. ICRP PUBLICATION 60, supra note 16, para. 100.

35. Radiation detriment is defined as the weighted sum of fatal and nonfatal cancers, leukemias, and deleterious genetic and other effects on health in an exposed population. It can be analyzed in terms of the collective exposure because the ICRP assumes, as does EPA, that the relationship between dose and effect is linear at environmental dose levels.

36. ICRP PUBLICATION 60, supra note 16, para. 112.

37. 42 U.S.C. §§ 4321-4347, ELR STAT. NEPA §§ 2-209.

38. Exec. Order No. 12866, 58 Fed. Reg. 51735 (Oct. 4, 1993).

39. ICRP PUBLICATION 60, supra note 16, para. 187.

40. Id. para. 121.

41. Id. paras. 190-191.

42. Id. para. 140.

43. Id. paras. 211-214.

44. Id. para. 217.

45. Id. The ICRP recommends that the choice of an action level for annual effective dose from indoor radon should normally fall in the 3 to 10 mSv range. This corresponds to a lifetime risk of approximately 1.5 to 4.3 x 10<-2>. ICRP, PROTECTION AGAINST RADON-222 AT HOME AND AT WORK, ICRP PUBLICATION 65 (1993).

46. See supra note 18.

47. 40 C.F.R. § 190.02(b) (1994).

48. 40 Fed. Reg. 23420 (codified at 40 C.F.R. pt. 190) (proposed May 29, 1975).

49. Id. at 23421.

50. Population risk refers to an estimate of the collective harm in the population or population segment being addressed and is often expressed as the projected number of health effect cases, e.g., cancer deaths, in the population of interest. It serves the same function as the ICRP radiation detriment.

51. This part of the standard limits the total quantity of radioactive materials released to the general environment from the entire uranium fuel cycle, per gigawatt-year of electrical energy produced, to 50,000 curies of krypton-85, 5 millicuries of iodine-129, and 0.5 millicuries combined of plutonium-239 and other alpha-emitting transuranic radionuclides with half-lives greater than one year. 40 C.F.R. § 190.10(b) (1994).

52. EPA commented that the standards provided "more explicit . . . protection . . . from radioactive effluents from the uranium fuel cycle. . ." and that it was not then "proposing revisions in existing Federal [guidance] because of its belief that a detailed examination of each major activity contributing to public radiation exposure is required before revision of this general guidance should be considered." See supra note 48.

53. U.S. EPA, ENVIRONMENTAL RADIATION PROTECTION REQUIREMENTS FOR NORMAL OPERATION OF ACTIVITIES IN THE URANIUM FUEL CYCLE, FINAL ENVIRONMENTAL STATEMENT, vol. 1, at 23 (1976).

54. 40 Fed. Reg. 23421 (May 29, 1975).

55. These risks are now estimated to be higher. The individual limits are annual dose equivalents of 25 mrem to the whole body (a currently estimated lifetime risk of 9 x 10<-4>, 75 mrem to the thyroid (currently estimated as a risk of 1 x 10<-4>, and 25 mrem to any other organ (currently estimated as a risk of approximately 10<-4>). 40 C.F.R. § 190.10(a) (1994). This "whole body/specific organ" approach has since been replaced by the effective dose equivalent (EDE) methodology. The EDE is a risk-weighted sum of the doses to individual organs. The dose to each organ is weighted according to the risk of cancer in that organ (or, in the case of the gonads, the risk of serious genetic effects). These weighted organ doses are then added together to form the EDE. The EDE, which is simpler to implement and is more closely related to risk, is a regulatory concept recommended by the ICRP. The EDE approach is used in EPA's Radiation Protection Guidance to Federal Agencies for Occupational Exposure, has been proposed in EPA's recommendations for Federal Radiation Protection Guidance for Exposure of the General Public, and has been used in all of EPA's rulemakings for radiation standards subsequent to the UFC standards. Radiation Protection Guidance to Federal Agencies for Occupational Exposure, 52 Fed. Reg. 2822 (Jan. 27, 1987); Federal Radiation Protection Guidance for Exposure of the General Public, 59 Fed. Reg. 66414 (Dec. 23, 1994).

56. 40 Fed. Reg. 23421 (May 29, 1975).

57. 42 U.S.C. §§ 7901-7942.

58. Id. § 7901.

59. Id. § 2022(a).

60. Id. §§ 6901-6992k, ELR STAT. RCRA §§ 1001-11012.

61. Id. § 2022(a).

62. 40 C.F.R. pt. 192 (1994).

63. U.S. EPA, FINAL ENVIRONMENTAL IMPACT STATEMENT FOR THE CONTROL OF BYPRODUCT MATERIALS FROM URANIUM ORE PROCESSING (40 C.F.R. pt. 192), Vol. II, at A.3 (1983).

64. 48 Fed. Reg. 590 (Jan. 5, 1983).

65. Other potentially hazardous constituents of the tailings include arsenic, selenium, molybdenum, and uranium. Id. at 592.

66. 40 C.F.R. § 192.02 (1994).

67. See id. pt. 61, subpt. T (1993). These CAA standards apply to the Title II sites in addition to, and address certain limitations of, the UMTRCA standards. For example, the UMTRCA standards did not establish a timetable for implementation. However, both EPA's Title II UMTRCA standard and NRC's regulations implementing those standards at NRC facilities subsequently were amended to address these limitations so as to allow EPA to rescind application of its CAA standards at subpart T for NRC facilities. 58 Fed. Reg. 60340 (Nov. 15, 1993).

68. 42 U.S.C. §§ 300f to 300j-26.

69. The standard calls for a radium soil content no greater than 5 picocuries per gram (pCi/g) above background at the soil surface, which, for the highly unlikely scenario of full-time residence outdoors on contaminated soil, corresponds to a risk of 1.2 x 10<-3>. The subsurface radium standard was set at 15 pCi/g, the limit of readily performed field measurements for subsurface surveys. It is intended to be used as a practicable criterion for locating buried tailings, which have an average concentration of about 300 pCi/g radium.

70. 48 Fed. Reg. 590, 592 (Jan. 5, 1983).

71. 40 C.F.R. pt. 191 (1994).

72. 42 U.S.C. §§ 10101-10270.

73. Id. § 10141(a).

74. 50 Fed. Reg. 38066 (Sept. 19, 1985).

75. 58 Fed. Reg. 66398 (Dec. 20, 1993). In 1987, the U.S. Court of Appeals for the First Circuit vacated and remanded the 1985 disposal standards, finding that the Agency had failed to adequately explain internal discrepancies (the portions of the standards applicable to individuals and groundwater applied for 1,000 years, whereas the containment requirements applied for 10,000 years) and the apparent failure of the standards to conform to certain requirements of the SDWA. See Natural Resources Defense Council v. U.S. Environmental Protection Agency, 824 F.2d 1258, 18 ELR 20088 (1st Cir. 1987). The 1992 Waste Isolation Pilot Plant Land Withdrawal Act reinstated all aspects of the disposal standards except for the individual and groundwater protection requirements. Pub. L. No. 102-579, 106 Stat. 4777, 4786 (1991).

76. 50 Fed. Reg. 38066, 38069, 38071 (Sept. 19, 1985).

77. 40 C.F.R. § 191.15 (1994).

78. 58 Fed. Reg. 66398, 66402 (Dec. 20, 1993).

79. 40 C.F.R. § 191.24 (1994).

80. 42 U.S.C. § 7412(b)(1)(B) (1977). Section 301 of the 1990 CAA Amendments replaced this health-based approach with a requirement for technology-based standards. Pub. L. No. 101-549, § 301, 104 Stat. 2399, 2539 (1990) (amending CAA § 112(d)). However, § 301 of the Amendments requires the Administrator to examine the residual risks that would remain after implementation of technology-based standards and to return to health-based standards, if necessary, to protect public health with "an ample margin of safety." Id., 104 Stat. at 2543 (amending CAA § 112(f)).

81. 824 F.2d 1146, 17 ELR 21032 (D.C. Cir. 1987).

82. Id. at 1164-66, 17 ELR at 21043 (emphasis added).

83. Id. at 1165, 17 ELR at 21043.

84. 54 Fed. Reg. 9612, 9621-22 (Mar. 7, 1989).

85. Id. at 51654, 51657 (Dec. 15, 1989).

86. Id. at 51658-59.

87. Id. at 51659, 51688.

88. Id. at 38044 (Sept. 14, 1989). The benzene standards are based on cancer incidence.

89. Id. at 51654 (Dec. 15, 1989). The radionuclide standards are based on cancer mortality. For radiation exposure, on average, the level of cancer incidence is 50 percent higher than that for mortality.

90. See 40 C.F.R. pt. 61 (1994). Benzene and radionuclides were listed by rule as hazardous air pollutants (HAPs) under § 112(b)(1)(A) of the Clean Air Act of 1977. 42 Fed. Reg. 29332 (June 8, 1977) (benzene); 44 Fed. Reg. 76739 (Dec. 27, 1979) (radionuclides). This section required the Administrator to list any pollutant that may result in an increase in mortality or an increase in serious, irreversible, or incapacitating reversible illness. The CAA Amendments of 1990 subsequently listed benzene and radionuclides (including radon) as HAPs at CAA § 112(b)(1). Pub. L. No. 101-549, § 301, 104 Stat. 2399, 2533, 2535 (1990).

91. In some cases, the MIR for certain sources of benzene and radionuclides was calculated to be somewhat above 1 x 10<-4>. However, these risk estimates were, in some cases, based on conservative exposure assessments, e.g., 24 hours per day outdoor exposure extending over a 70-year period. See 54 Fed. Reg. 51654, 51660 (Dec. 15, 1989).

92. The controls required to satisfy the standard were found to protect over 90 percent of persons within 80 kilometers of the source to lifetime risks of no more than 10<-6>. Id. at 51655.

93. In reaching this finding, the Agency carried out a truncated analysis of population risk (to 80 kilometers), but observed that a more extended analysis would not have affected the conclusions for the radionuclides considered. Although this is true, the analysis failed to identify and address the risks associated with releases of carbon-14, a very long-lived radionuclide later identified by EPA as a significant source of population risk in its most recent rulemaking for high-level waste, a problem the Agency had first noted, but lacked the necessary data to act on, when it established its standards for the uranium fuel cycle in 1975. 42 Fed. Reg. 2858, 2859 (Jan. 13, 1977).

94. See 40 C.F.R. pt. 61, subpt. H (1994).

95. 42 U.S.C. § 300g-1(b)(3)(A), ELR STAT. SDWA § 1412(b)(3)(A).

96. Id. § 300g-1(b)(4), ELR STAT. SDWA § 1412(b)(4).

97. Id. §§ 300f(1), 300g-1(a)(3), 300g-1(b)(7)(A), ELR STAT. SDWA §§ 1401(1), 1412(a)(3), 1412(b)(7)(A).

98. Id. § 300g-1(b)(5), ELR STAT. SDWA § 1412(b)(5).

99. See, e.g., 57 Fed. Reg. 31776, 31797 (July 17, 1992).

100. Id.

101. Id. at 31798.

102. See, e.g., 56 Fed. Reg. 33050, 33080, 33101, 33113 (to be codified at 40 C.F.R. pt. 141) (proposed July 18, 1991) (proposed rule for radionuclides). But see 57 Fed. Reg. 31776, 31797 (July 17, 1992) (insufficient occurrence data available for certain chemical carcinogens).

103. However, like the CAA NESHAPs, the standards for radionuclides are based on cancer mortality while the standards for chemicals are based on cancer incidence. As noted earlier, for radiogenic cancers, incidence averages a factor of 1.5 higher than mortality. See supra note 89. This is usually not a significant difference in comparison to the factor of 2-3 uncertainty in the underlying risk estimates.

104. 40 C.F.R. §§ 141.50-.52 (1994) (MCLGs for organic, inorganic, and microbiological contaminants); 56 Fed. Reg. 33050, 33126 (to be codified at 40 C.F.R. pt. 141) (proposed July 18, 1991) (proposed MCLGs for radionuclides). Although the proposed MCLG for uranium is zero based on its potential carcinogenicity in humans, the proposed MCL is based primarily on its kidney toxicity, which may occur at levels below the cancer risk level. Id. at 33068, 33076-77, 33100.

105. The MCLs for vinyl chloride, dioxin, and ethylene dibromide and the proposed MCL for radon and radium-226 allow risks somewhat greater than 1 x 10<-4>. However, the risk estimates are based on the conservative assumption of ingestion of 2 liters of drinking water per day contaminated at the level of the MCL over a lifetime of 70 years.

106. 40 C.F.R. § 141.15(a) (1994). These standards did not address radon or uranium, because the Agency did not have sufficient health and occurrence data at the time on radon and uranium.

107. 56 Fed. Reg. 33050, 33126 (July 18, 1991).

108. 40 C.F.R. §§ 141.15(b), 141.16 (1994).

109. In the proposed rule, EPA solicited comments on this issue and possible approaches to ensure that overall risks do not rise above the 10<-4> level. See 56 Fed. Reg. 33062, 33101 (to be codified at 40 C.F.R. pt. 141) (proposed July 18, 1991).

110. See, e.g., 57 Fed. Reg. 31766 (July 17, 1992) (synthetic organic and inorganic chemicals).

111. 42 U.S.C. § 9601(14), ELR STAT. CERCLA § 101(14).

112. Id. § 9621(d)(1), ELR STAT. CERCLA § 121(d)(1).

113. Id. § 9621(d)(2)(A), ELR STAT. CERCLA § 121(d)(2)(A). An alternative that does not meet an ARAR under federal environmental or state environmental or facility siting law may be selected under certain circumstances. See 40 C.F.R. § 300.430(f)(1)(ii)(C) (1994).

114. 40 C.F.R. pt. 300 (1994).

115. Id. § 300.430(d); see U.S. EPA, THE RISK ASSESSMENT GUIDE FOR SUPERFUND: VOLUME I, HUMAN HEALTH EVALUATION MANUAL - (PART A) INTERIM FINAL, EPA/540/1-89/002 (Dec. 1989).

116. For these assessments under CERCLA, EPA assumes that an individual is exposed to a contaminated site for 30 years.

117. Risks of less than 10<-4> warrant cleanup in certain situations depending on the presence of additional risk of noncarcinogenic effects or an adverse environmental impact. See Role of the Baseline Risk Assessment in Superfund Remedy Selection Decisions, Memorandum of Don R. Clay, Assistant Administrator, Office of Solid Waste and Emergency Response, EPA, to Directors of Regional Divisions (Apr. 22, 1991) [hereinafter Memorandum of Don R. Clay].

118. In situations involving contamination of groundwater, for example, MCLs are generally used to gauge whether remedial action is warranted, and many of these specify risk levels lower than 10<-4>.

119. 40 C.F.R. § 300.430(e)(2)(i) (1994); see RISK ASSESSMENT GUIDANCE FOR SUPERFUND: VOLUME I - HUMAN HEALTH EVALUATION MANUAL (PART B, DEVELOPMENT OF RISK-BASED PRELIMINARY REMEDIATION GOALS) INTERIM (Dec. 1991).

120. The upper boundary of this risk range is not a "discrete line" at 1 x 10<-4>. A risk estimate on the order of 10<-4> may be considered adequately protective if justified by site-specific conditions, including the nature and extent of contamination and associated risks. EPA uses 10<-6> as a point of departure when ARARs are not available or are not sufficiently protective because of greater overall risks due to the presence of multiple contaminants or cross-media contamination. See Memorandum of Don R. Clay, supra note 117, at 4.

121. See 55 Fed. Reg. 8666, 8717 (Mar. 8, 1990).

122. See U.S. EPA, METHODS FOR EVALUATING THE ATTAINMENT OF CLEANUP STANDARDS: VOLUME 1: SOILS AND SOLID MEDIA, EPA/230/02-042 (1989); U.S. EPA, DETERMINING SOIL RESPONSE ACTION LEVELS BASED ON POTENTIAL CONTAMINANT MIGRATION IN GROUNDWATER: A COMPENDIUM OF EXAMPLES, EPA/540/2-89/057, at 56, 60 (1989).

123. See 55 Fed. Reg. 8666, 8717-18 (Mar. 8, 1990).

124. Id. at 8718-19.

125. 42 U.S.C. § 9605(g)(2)(B), ELR Stat. CERCLA § 105(g)(2)(B) (consideration of exposure to "human population" in placing sites on the NPL).

126. 40 C.F.R. § 300.430(f) (1994).

127. Id. § 300.430(e)(9)(iii)(A)-(I); see also U.S. EPA, RISK ASSESSMENT GUIDANCE FOR SUPERFUND: VOLUME 1-HUMAN HEALTH EVALUATION MANUAL (PART C, RISK EVALUATION OF REMEDIAL ALTERNATIVES), EPA/540/R-92/004, INTERIM (Dec. 1991).

128. EPA considers cancer incidence for both chemicals and radionuclides in its site-specific risk assessments under CERCLA.

129. 15 U.S.C. § 2661. Individually measured ambient radon concentrations in outside air range from 0.1 picocuries per liter (pCi/L) to 1.2 pCi/L.

130. Id. § 2663.

131. U.S. EPA, A CITIZEN'S GUIDE TO RADON, 402-K92-001 (May 1992).

132. The lifetime risk at the action level for persons who have never smoked is 1.6 x 10<-3>; the risk for current smokers is 3.0 x 10<-2>.

133. See U.S. EPA, TECHNICAL, SUPPORT DOCUMENT FOR THE 1992 CITIZEN'S GUIDE TO RADON, EPA 400-R-011 (May 1992).

134. Except for the groundwater provisions, which were required to conform to the Agency's RCRA regulations under UMTRCA.

135. 51 Fed. Reg. 33050, 33080, 33081 (July 18, 1991).

136. EPA made this point in promulgating the individual protection requirements of the 1993 high-level radioactive waste disposal rule. The Agency stated that its 15 mrem limit is an appropriate and acceptable fraction of the 100 mrem ICRP recommendation. 58 Fed. Reg. 66398, 66402 (Dec. 20, 1993).

137. Federal Radiation Protection Guidance for Exposure of the General Public, 59 Fed. Reg. 66414 (to be codified at 3 C.F.R.) (proposed Dec. 23, 1994).

138. Id. at 66416.

139. Id. at 66420-24.

140. In the CAA rulemaking, EPA further found that an artificial source of radiation does not present an acceptable risk simply by being less than natural background.

141. However, in cases where the ICRP principles for intervention lead to higher risk levels because the cleanup costs are high, a similar result is possible under CERCLA when sufficient funds are not available in the fund established for federal funding of CERCLA cleanups. 42 U.S.C. § 9621(d)(4)(F), ELR STAT. CERCLA § 121(d)(4)(F).

142. See, e.g., Curtis C. Travis & Sheri T. Hester, Background Exposure to Chemicals: What is the Risk?, 10 RISK ANALYSIS 463 (1990). Chemicals from which there can be significant background exposures include, for example, arsenic, chloroform, benzene, and dioxins.

143. See Curtis C. Travis et al., Is Ionizing Radiation Regulated More Stringently Than Chemical Carcinogens?, 56 HEALTH PHYSICS 527 (1989).

25 ELR 10657 | Environmental Law Reporter | copyright © 1995 | All rights reserved