1594 ENVIRONMENT REPORTER EPA'S STANDARD-SETTING FOR TOXIC POLLUTANTS December 1983 re re ee A. Introduction The Clean Air Act requires the Environmental Protection Agency to establish national emission standards for hazard- ous air pollutants. Section 112 of the Act requires that EPA first “list” a pollutant as hazardous, and then set emission standards for industrial plants emitting the listed pollutant. By the end of 1980, EPA had listed seven pollutants as hazardous and set or proposed plant emissions standards for five of them. EPA recently proposed standards for plants emitting the remaining two pollutants — inorganic arsenic and radionuclides.’ This paper considers the central policy issues raised by EPA's standard-setting approach in these proposed rules. EPA's preamble to its proposed rules for inorganic arsenic and radionuclides outlines the following three-step approach for establishing standards for hazardous air Ie pollutants:? 3 UP oe 7m, a = ‘aA * BT PP eet pc Toro Seo ae on ot mh Om - si} (D in . -' . — Categories of pollution sources are. classified accord. ing to whether they pose a “significant risk” to public health. In making such a determination, EPA considers that a source category poses a “significant risk”if there is a Strong likelihood that it emits a carcinogen and that individuals or the general population receive significant exposure to the substance emitted by the source category. — A source category judged by EPA as posing significant public health risks is then evaluated to determine the current level of control and the level of control constituting Best Available Technology (BAT)for plants or facilities in the source category. EPA’s determination of BAT takes into account such factors as the potential for improved control, the economic effects of additional contro} Tequirements on the source category, and the age and temaining useful life of the facilities. -— EPA determines whether the public health risks posed by the residual emissions of a source category would be unreasonable after installation of BAT control. In making this determination, EPA considers the likely additional reduction in public health risks, the economic effects, and other effects of reguiatory alternative that ‘are more stringent than the selected BAT requirements. " ‘See FR 15076 and 48 FR 33112. "See 48 FR 15076 and 48 FR 33112. EPA also has outlined this three-tiered standard-setting process in a draft staff paper describing a process for evaluating and controlling toxic po!lutants. U.S. Environmental Protection Agency, Proposed Process for Evaluation and Control of Toxic Pollutants, External Staff Draft, chooses to use — and not use — risk assessment information.’ The critical step in this process is EPA’s BAT ap- proach to standard-setting. This approach explicitly ex- cludes consideration of the likely public health effectiveness of BAT-leve) controis for source categories posing ‘“‘significant” risks. At other points in its decision process, where EPA does consider risk assessment information, it does so in & way that imparts a large and inappropriate conservative bias to the ultimate regulatory decision. In this paper we begin by considering the regulatory policy that results from the standard-setting approach used . in EPA's proposed rules (Section B). We then discuss EPA's use of a BAT-approach and the possible modification of that ‘ approach described in EPA’s proposed rule for limiting inorganic arsenic emissions from low arsenic feed copper smelters (Section C). Finally, we discuss the role of risk assessment information at other junctures in EPA’s standard-setting process, and the relative weighting given to reductions in individual risk as opposed to population risks from exposure to these substances (Section D). , B. The Effectiveness of EPA's Regulatory Requirements The objective of the Section 112 hazardous air pollutant standards is protection of public health, so it is important to assess the effectiveness of EPA’s standard-setting decisions in terms of the likely public health gains. We have developed such information on the public health effectiveness of EPA’s proposed BAT standards using available data for sources emitting radionuclides and inorganic arsenic. (See Tabie 1.) In the case of EPA’s proposed rule for sources emitting inorganic arsenic, for example, the public health gains per million dollars expenditure range from 2 expected cancers avoided per million dollars of expenditure for the high arsenic-feed copper smelter at Tacoma, Washington,to Jess than 0.001 expected cancer avoided per million dollars of expenditure for some of the other copper smelters and glass manufacturing plants regulated under the proposed rule. The effectiveness of EPA's proposed rules in terms of public health gains varies across individual plants by a factor of 2,000. To place in perspective an effectiveness of 0.001 expected cancer avoided per million dollars of expenditure, it would require an expenditure of one billion dollars to avoid a single expected case of cancer. ‘Neither the statute nor the legislative history specifically address the role of risk assessment information in standard-setting or spells out the nature of the requirements to be applied under the “ample margin of safety” language for pollutants that may present health risks at any level of exposure. In the absence of specific statutory language, EPA's practice has been to rely on a technology-based approach in regulating these pollutants. oo vf, wate ne rasa AOE AOET OST Tegra March 23, 1983. ‘We think there are important shortcomingsin this process arising in large measure from the way in which EPA 1-1: 1-13-84 Published by THE BUREAU OF NATIONAL AFFAIRS, INC., Washington, D.C. 20037