[*PG229]CONTROL OF HAZARDOUS AIR POLLUTION
Abstract: This article begins with an overview of ecosystem and human health impacts of hazardous air pollutants (HAPs) and is followed by a synopsis of the early efforts to control HAPs that were based on common law and the Clean Air Act (CAA) prior to 1990. In the CAA Amendments of 1990 Congress added a twenty-fold expansion of the statutory provisions aimed at the control of HAPs . In the decade that followed the Environmental Protection Agency (EPA) has generated thousands of pages of rules and guidance to implement the 1990 statutory changes. This article provides an analysis of these requirements and the program created to implement them. Several categories of HAPs are the subject of a more in depth analysis including asbestos, synthetic organic chemicals, and emissions from incineration. The article then examines the requirements aimed at preventing accidental catastrophic environmental releases of HAPs and the risk management plans that must be developed by about 70,000 facilities that handle regulated chemicals. The emergency planning and reporting requirements imposed in 1986 by the Emergency Planning and Community Right to Know Act (EPCRA) also are addressed.
In 1984, the Union Carbide toxic chemical release in Bhopal, India that killed about two thousand people focused attention on [*PG230]hazardous pollutants, particularly airborne hazardous pollutants.1 A subsequent chemical release from a West Virginia facility drew attention to the problem here in the United States.2 Congress responded by introducing a toxics reporting requirement in 1986 and increased implementation and regulation of hazardous air pollutants (HAPs) under the Clean Air Act (CAA) in the 1990 Amendments.3
In 1986, Title III of the Superfund Amendments and Reauthorization Act (SARA) created the free-standing Emergency Planning and Community Right-to-Know Act (EPCRA).4 Sections 11021–11023 of EPCRA mandate reporting to the Environmental Protection Agency (EPA) on the amount of toxic substances released into the air each year by industry. This toxic release inventory (TRI) must include releases of more than 320 chemicals included in the Committee Print Number 99–169 of the Senate Committee on Environment and Public Works.5 The statute also provides for EPA’s Administrator to add chemicals to the list, and any person may petition the Administrator to add or delete chemicals.6 Since July 1, 1988, industries with standard industrial classification (SIC) codes of 20 to 39 that have ten or more full time employees and that handle regulated chemicals above threshold amounts have to meet the annual TRI reporting requirements.7 The threshold amount is 25,000 lbs/year for 1989 and subsequent years.8
The first year of submissions showed HAPs were released in amounts estimated at 2.18 billion lbs/year.9 This did not include emissions from federal facilities or from facilities exempted because of their small size or less than threshold use of chemicals.10 It also did [*PG231]not include emissions to water or land that indirectly become air pollutants (e.g., evaporation). Nor did it include the release of hazardous materials not on the EPA list.11 The size of this toxic chemical emissions inventory had a significant effect on the public perception of the hazardous chemical problem and helped spur Congress to regulate these emissions.12
In 1997, 21,490 facilities filed 61,123 Form Rs13 as required by EPCRA.14 According to the TRI data, the total releases to land, air, and water were 2.58 billion pounds of the 643 chemicals subject to EPCRA reporting requirements.15 Air emissions comprised 62.9% of all on-site toxic releases.16 This reporting did not include releases from electric utilities, motor vehicles, or from emission sources with less than ten employees.17 The top states for total releases in 1997 were Texas with 261.7 million pounds, Louisiana with 186 million pounds, and Ohio with 158.7 million pounds.18 The hazardous chemical most released to the air in 1997 was methanol with air emissions of 194.4 million pounds.19 Ammonia ranked second with air releases of 156.1 million pounds.20 Air emissions represented over three-fourths of the total on-site and off-site releases for twelve of the top twenty chemicals.21 In addition to methanol and ammonia, these were toluene, xylene, n-hexane, chlorine, hydrochloric acid, methyl ethyl ketone, carbon disulfide, dischloromethane, styrene, and glycol ethers.22 Over 50,000 chemicals are in commerce in the United States, [*PG232]and approximately 14,000 of those are releases in quantities over 10,000 lbs/year.23 Fewer than 1000 of these chemicals have been properly evaluated for toxicity because of lack of resources by the regulating government agencies.24
EPA used three approaches in the early 1990s to obtain reductions in toxic releases. They were: (1) the Toxic Release Inventory required by EPCRA; (2) the voluntary 33/50 program to encourage a 33% reduction of releases of seventeen toxic chemicals by the end of 1992 and a 50% reduction by the end of 1995;25 and (3) the Early Reduction Program, under CAA section 112, to reward industry with a six-year extension for meeting CAA emission standards if the industry made a 90 to 95% reduction in toxic air emissions before the maximum achievable control technology (MACT) standards are proposed.26 The United States General Accounting Office made a study of the effectiveness of these efforts and concluded that there was little evidence that the 33/50 or Early Reduction programs had a meaningful benefit.27
According to TRI reports, during the period from 1988 to 1992, toxic releases to air, water, and land were reduced by 35%, or about 1.7 billion pounds.28 Over 50% of the reduction involved four chemicals—acetone, ammonium sulfate, chlorine, and hydrochloric acid. Also, over 40% of the reductions occurred in three states—Louisiana, Texas, and Virginia.29 Reductions, however, are reported by the companies making the releases, but they are not verified by EPA or the states,30 and EPA states that much of the reported TRI reduction is due to (1) changes in how chemicals are reported; (2) changes in how releases are estimated; and (3) decreases in production levels.31
Industrial sources in the United States subject to TRI reporting requirements emitted about 1.332 billion pounds of toxic air emissions into the atmosphere in 1997.32 Emissions from automobiles, pesticide use, municipal waste incineration, and other combustion processes not included in the TRI report were responsible for additional toxic air pollutant releases.33 Much of this pollution is eventually deposited on land or water.34
Some pollutants remain airborne and contribute to air pollution problems far from the pollution source. Other pollutants released into the air can be deposited to land and water through precipitation, or by settling directly out of the air into land or water. Eventually, a large portion of those pollutants deposited near water bodies or small tributaries will reach the water bodies via stormwater runoff or inflow from tributary streams.35
After deposition, some of these chemicals bioaccumulate in living organisms and can become more concentrated at higher levels in the food chain (a phenomenon known as biomagnification).36 Dichlorodiphenyltrichloroethane (DDT), for example, has bioaccumulated in the Great Lakes where it has an adverse effect on fish, birds, and other wildlife.37 Another problem involves “persistent” air toxics, such as mercury or lead, that can remain for a long time (sometimes forever) in the environment and can be transported long distances.38 Some toxic air pollutants are precursors that can lead to compounds such as ozone, acid aerosols, and carcinogenic hydrocarbons that form in the atmosphere over large areas of North America.39 One government study says that air deposition of lead and [*PG234]polychlorinated biphenyls (PCBs) account for 90% or more of those compounds entering Lake Superior.40
The 1990 CAA Amendments, in section 112(m), require the Administrator, in cooperation with the Under Secretary of Commerce for Oceans and Atmosphere, to study the extent of atmospheric deposition of HAPs to the Great Lakes, the Chesapeake Bay, Lake Champlain and coastal waters.41 Monitoring of these waters is required,42 and a report by the Administrator is to be used to promulgate further emission standards or control measures for HAPs (and their atmospheric transformation products) as may be necessary and appropriate.43 EPA has monitored some toxic compounds in the Great Lakes since the 1980s and in the 1990s began to measure deposition in the Chesapeake Bay.44
The first report required by section 112(m) was released in May 1994; the second report, “Deposition of Air Pollutants to the Great Waters,” was sent to Congress on July 1, 1997.45 It confirmed the conclusion of the first report that toxic air pollutants deposited in the waters identified in section 112(m) were having an adverse impact on their ecosystems. The fifteen pollutants of concern that are the focus of the reports are: cadmium and cadmium compounds; chlordane; DDT/DDE; dieldrin; hexachlorobenzene (HCB); a-hexachlorocyclohexane (a-HCH); lindane; y-hexachlorocyclohexane (y-HCH); lead and lead compounds; mercury and mercury compounds; (PCBs); polycyclic organic matter (POM); tetrachlorodibenzo-p-dioxin (TCDD; dioxins); tetrachlorodibenzofuran (TCDF; furans); toxaphene; and nitrogen compounds.46 The report, however, concluded that no additional [*PG235]legislation was needed because the existing section 112 provides adequate authority to deal with the problem.47
Stationary and mobile sources emit many different air pollutants to which large populations are exposed. Some air pollutants have toxic and/or carcinogenic effects following direct inhalation exposure (e.g., carbon monoxide and benzene).48 Others, such as lead and arsenic, reach humans by a variety of pathways including direct inhalation, inhalation of resuspended dust, ingestion of contaminated food products, ingestion of contaminated water, and skin contact with contaminated soil, water, or dust.49
Industrial and agricultural workers often are exposed to many toxic substances in the air in concentrations above those to which the general population is exposed. Such exposures can cause cancer or a wide range of non-cancer health effects.50
Building occupants may be exposed to radon and its decay products as well as to many airborne combustion products, including nitrogen dioxide and environmental tobacco smoke. Indoor exposures to toxic agents in consumer products (e.g., solvents, pesticides, formaldehyde) also can cause cancer and a range of non-cancer health effects.51 Due to the large population directly exposed to a number of agents, some of which are highly toxic, this problem poses relatively high human health risks.52
The known harm caused by these chemicals includes an estimated three thousand cases of fatal cancer each year as well as birth defects, lung disease, nervous system disorders, immune system disorders, endocrine system disorders, neurological problems, liver damage, and other health problems.53 In 1999, a preliminary study in southern California indicated that exposure to ten HAPs may be responsible for an additional 426 cancer cases per million exposed [*PG236]individuals.54 Very little is known about health problems, other than cancer, caused by air pollutants, because most testing has been for carcinogenic properties.55 In addition, EPA announced that it has “fair or better” data available only on twenty HAPs and that the health effects data for other HAPs is “spotty.”56
The Mickey Leland National Urban Air Toxics Research Center (Center) was created under the authority of CAA section 112(p).57 The statute requires that it be located in Harris County, Texas.58 One of its responsibilities is to direct a sound research program to better understand the risks posed to human health by toxic chemicals in the urban air. To do this, the Center researches the effects of the 189 materials Congress has identified in CAA section 112(b).59 The Center issued a strategic research plan in 1997 that focuses the Center’s activities on human health effects, community programs, and personal exposure.60 The plan is reviewed annually.61
On May 23, 1996, the Agency for Toxic Substances and Disease Registry published its methodology for calculating minimal risk levels.62 Minimal risk levels (MRLs) are screening tools to determine whether acute (1 to 14 days), intermediate (15 to 364 days), or chronic (365 days or more) exposure to a chemical poses a threat to human health and therefore merits investigation.63 Such evaluations are done for both inhalation exposure and oral exposure.64 As of [*PG237]February 2000, the agency has calculated 286 risk levels for 136 chemicals.65
Despite the comprehensive regulatory regime, a business is not necessarily protected from common-law state actions for nuisance, negligence, strict liability, or trespass for harm to an individual resulting from a hazardous air release. The United States Supreme Court once stated that “air pollution is . . . one of the most notorious types of public nuisance in modern experience.”66 The CAA specifically provides that nothing “shall restrict any right which any person (or class of persons) may have under any statute or common law to seek enforcement of any emission standard or limitation or to seek any other relief . . . .”67 While the CAA may preempt claims in federal courts, it does not stop actions brought under color of state law nor does the permit shield doctrine apply to estop assertion of a claim.68
HAPs have been subject to control since 1970 by CAA section 112,69 which provides for National Emission Standards for HAPs, known as NESHAPs. Prior to the 1990 CAA Amendments, the Administrator of EPA was required to publish a list of pollutants deemed hazardous.70 Within 180 days of publication, proposed regulations were to be promulgated and within the following 180 days, either an emission standard was to be set or the Administrator had to find that the listed pollutant was not a HAP.71 The Administrator was to set emission standards “at the level which in his judgment provides ample margin of safety to protect the public health from such hazardous air pollutant.”72
[*PG238] The statutory time frame was unrealistic. The data and testing methods were inadequate, and the risk analysis approach proved controversial at the time.73 It typically took from four to seven years from the time EPA identified a chemical as a candidate until a final regulation was issued.74 Thus EPA resisted using section 112. After twenty years, EPA had regulated only seven hazardous substances—asbestos, beryllium, mercury, radionuclides, inorganic arsenic, benzene, and vinyl chloride.75 EPA issued notices and health effects information on twenty-five other substances but never completed the regulatory process. For example, coke oven emissions were listed,76 but no standard under the pre-1990 CAA was ever promulgated.77 A notice of intent to list chromium or hexavalent chromium under section 112 was as far as EPA ever moved to regulate that chemical.78
Although a few hazardous chemicals were regulated, only a limited number of circumstances involving their use were covered by the regulations. For example, arsenic emissions were controlled at primary copper smelters, glass manufacturing plants, and arsenic plants.79 Vinyl chloride had similar restrictions on its regulatory applicability.80 Mercury emissions from electric power plants were exempt from the standards even though they contribute to the high mercury levels of fish in the Great Lakes.81
[*PG239] Radionuclide regulations control radon-222 emissions from a variety of sources.82 National standards, however, for radon for Department of Energy facilities were not promulgated until 1989,83 and radon emissions from phosphogypsum stacks were not regulated until 1992.84 Radionuclides from elemental phosphorus plants were regulated in 1989,85 but EPA changed the standard86 on December 19, 1991.87
In 1989 EPA regulated radon emissions from operating mill tailings88 and from uranium mill tailings disposal sites.89 Radon emissions from inactive sites are regulated under both the CAA and the Uranium Mill Tailings Radiation Control Act (UMTRCA).90 EPA proposed to amend the UMTRCA to include monitoring and closure requirements.91 It planned to make the UMTRCA the vehicle to regulate mill tailing disposal sites and to transfer the responsibility to implement the requirements to the Nuclear Regulatory Commision (NRC).92 There are approximately twenty NRC-licensed waste piles that are primarily located in the western states.93 The remaining waste is scattered in outdoor piles that range in size from a few acres to several hundred acres.94 The December, 1989 EPA rules require earth covers over the piles in order to limit radon emissions to an acceptable level, but the NRC claimed these rules duplicated their authority.95 EPA claimed that under the 1990 CAA Amendments it [*PG240]may transfer this authority to the NRC,96 and in September, 1992 appeared ready to give the NRC this responsibility.97 However, on September 25, 1992, the U.S. Court of Appeals for the D.C. Circuit ruled that EPA does not have authority to suspend emission standards for radionuclides from sources regulated by the NRC except with respect to the NRC-licensed medical facilities.98 The court rejected EPA’s argument that CAA section 112 permitted broad suspension of the radionuclide rules.
Although EPA regulated only a few HAPs, even the short list of regulated hazardous pollutants was the result of litigation that forced EPA to regulate. For example, the court in Sierra Club v. Gorsuch99 ordered EPA to regulate radionuclides.100 EPA’s failure to act later led to a contempt order.101 A similar case, New York v. Gorsuch,102 led to a court order requiring organic arsenic to be regulated.
[*PG241] The failure of EPA to seriously regulate HAPs led Congress in the CAA Amendments of 1977 to encourage EPA to act. The definition of “hazardous air pollutant” was expanded to mean “an air pollutant to which no ambient air quality standard is applicable and which in the judgment of the Administrator causes, or contributes to, air pollution which may reasonably be anticipated to result in an increase in mortality or an increase in serious irreversible or incapacitating reversible illness.”103 A new section 122104 was added to the CAA that required the Administrator to study the need for regulating radioactive pollutants, cadmium, arsenic, and polycyclic organic matter. If any of these substances, when released to the ambient air, would cause or contribute to air pollution that might reasonably be anticipated to endanger public health, they were required to be regulated.105 These CAA changes had little effect on EPA’s efforts to regulate HAPs.
The control of hazardous emissions under the 1977 CAA Amendments was not as stringent as the action being taken under the Clean Water Act (CWA). The CWA regulated only six toxic substances when the Natural Resources Defense Council (NRDC) sued EPA in 1975.106 In June, 1976 a consent decree was approved requiring EPA to regulate sixty-five pollutants and classes of pollutants. In the 1977 CWA, the consent decree was codified in sections 301(b)(2)(A) and 307(a)(1) that required a readjusted list of 129 toxic water pollutants to be regulated. The consent decree, however, continued to be enforceable even if its provisions were beyond the scope of the 1977 CWA Amendments.107
EPA’s reticence to use CAA sections 112 and 122 was due not only to the short time period provided for promulgating regulations, but also was due to the agency’s interpretation that the statute did not allow cost and technology availability to be considered in setting standards.108 Section 112 appeared to require that, once a hazardous pollutant was listed, a no-risk standard must be set regardless of [*PG242]cost.109 In situations where there was scientific uncertainty as to the threshold of exposure below which there is no risk, section 112 seemed to require a zero-emissions standard.110 Thus, for carcinogens or substances with no known safe level, EPA believed no emissions could be allowed.111 Therefore, the use of section 112 might result in regulations so harsh that major segments of the economy would be adversely impacted.112 In addition, the time allowed for compliance after regulations were promulgated was unrealistically short because industry was required to implement section 112’s requirements in two years or less.113 Thus industry could face penalties for failing to comply with requirements that were impossible to meet.114
In 1976 EPA set a final emission standard for vinyl chloride, a gaseous synthetic chemical used to manufacture polyvinyl chloride plastics.115 It is a human carcinogen known to cause liver cancer and had been implicated as a cause of cancer in other organs, such as the brain and the lungs.116 The Environmental Defense Fund (EDF) sued EPA claiming that cost and technology factors had been considered in deciding to require a 95% emission reduction.117 A settlement agreement required EPA to make its vinyl chloride standards stricter.118 EPA proposed new regulations in June, 1977,119 but took no final action during the next seven years.120 In 1985 EPA withdrew the proposed standards121 and proposed in their place the 1976 [*PG243]regulations with minor revisions.122 The NRDC then sued to challenge the rescission.
In NRDC v. EPA,123 the U.S. Court of Appeals for the D.C. Circuit, en banc, held that EPA must set section 112 standards using a two-prong procedure.124 EPA reasoned that, due to the scientific uncertainty surrounding safe levels of exposure to vinyl chloride, it was authorized under section 112 to consider economic and technological factors in establishing emission levels.125 Based on this reasoning, EPA advocated emission levels set in accordance with the best available control technology.126 EPA argued that consideration of health factors alone in this situation would require complete prohibition of the carcinogen because a safe level could not be conclusively determined.127
The court rejected the NRDC position that section 112 demands a zero emission standard when a no-risk threshold cannot be determined.128 “Safe” does not mean “risk-free” according to the D.C. Circuit.129 While deferring to EPA’s construction of the statute and agreeing that economic and technological factors are permissible considerations within the Administrator’s discretion under section 112, the court held that the clear intentions of Congress relating to safety must be satisfied prior to any consideration of economics and technology.130 The D.C. Circuit established a two-step process requiring the Administrator to first make a determination of safety, exclusive of other factors.131 Only after safety is assured may the Administrator consider economic and technological factors to “diminish as much of the statistically determined risk as possible by setting the standard at the lowest feasible level.”132 This decision made the legal validity of other section 112 standards suspect because some [*PG244]standards were set after considering costs. On September 14, 1989, EPA promulgated a new standard for benzene.133 This and one of the [*PG245]radionuclide standards134 were the only standards established after the NRDC case and prior to the 1990 CAA Amendments.135
In 1989, EPA established its risk policy by deciding that the life-time excess risk of cancer, or maximum individual risk (MIR), to any given toxic should not be greater than one in 10,000 (1 x 10–4).136 In the same year, EPA promulgated its final rule for radionuclides, that protects the public from a MIR risk of 1 x 10–4 but also regulates to protect 90% of the people within eighty kilometers of a source to risk levels of 1 x 10–6 137 using the two-step approach required by the NRDC v. EPA case.138
While EPA was reluctant to regulate hazardous emissions under section 112, some regulation occurred using other sections of the CAA. Under the basic state implementation plan (SIP) program, both particulate matter and hydrocarbons, which frequently are hazardous pollutants, are regulated.139 CAA section 111, regulating new and modified sources, imposes technology-based controls to reduce hazardous emissions.140 For example, fluorides from phosphate and aluminum plants, sulfuric acid mist from acid plants, and sulfur compounds from pulp mills are controlled through the new source performance standards (NSPS) process. NSPS for wood stoves were promulgated specifically to regulate toxics.141 CAA subchapter II, controlling mobile sources, also results in reductions in HAPs,142 as does CAA section 211, which authorizes the regulation of fuels and fuel additives.143
[*PG246] In 1985, EPA announced a policy of shifting the control of hazardous emissions to the states.144 This approach pressed for more control over small area sources, such as gasoline stations, dry cleaners, and small combustion sources. A few states, such as New Jersey and California, moved to aggressively control hazardous emissions.145 These states regulated a total of 708 different HAPs. Because the CAA allowed states to regulate air toxics for which national standards did not exist, however, state programs were not consistent and differed in their degree of stringency.146 Thus, public exposure to toxic air pollution could vary from state to state.147
EPA’s limited regulation of air toxics was especially poor when contrasted with the federal regulation promulgated to control air pollutants in the workplace. The Occupational Safety and Health Administration (OSHA) set standards for almost 428 substances harmful when released into the air.148 Most of the substances on the EPCRA list are included in the regulation by OSHA. These regulations, however, were vacated and remanded by the U.S. Court of Appeals for the Eleventh Circuit because OSHA did not show that existing limits presented a significant health risk or that the new standards were economically or technologically feasible.149
In 1990, a new technology-based regulatory program was created to control HAPs.150 The new section, however, 112(q) adds a “savings provision” that preserves the existing NESHAP in 40 C.F.R. part 61 until they are amended.151 NESHAPs in part 61 must be reviewed and, if appropriate, revised to comply with the standard-setting requirements of the new section 112 within ten years of November 15, 1990.152 If standards that were promulgated and judicially challenged [*PG247]before November 15, 1990 are remanded to EPA, the Agency has the discretion to apply either the requirements of the new section 112 or the former section 112.153 The effect of section 112(q) is to continue the existing provisions of part 61 for sources affected by pre-1990 NESHAP.154
The almost complete failure of section 112’s NESHAPs to control HAP emissions led to a different approach in the 1990 CAA Amendments.155 The old chemical-by-chemical risk-based approach could have led to a “no emissions standard” for many chemicals because any emission arguably would create some risk.156 The old section 112 imposed standard setting requirements that were so strict that EPA tried not to use it, resulting in just seven regulated HAPs.157 The pre-1990 CAA section 112 was one and a half pages of text,158 but the 1990 amendments created a new CAA section 112 that fills twenty-eight pages.159 The 1990 approach to the regulation of HAPs is similar to the program created by section 307 of the CWA that is predicated on requiring the best demonstrated technology used by sources in an industrial category.160 To accomplish this, the new CAA section 112 in 1990 became much more complex, establishing a two-tiered regulatory approach in which the first component consists of developing technology-based standards (the “maximum achievable control technology” standards, or MACTs) and then a second, risk-based approach is to be used to assess the success of the MACT standards.161
“Major source” is defined broadly to include groups of stationary sources in a contiguous area that are under common control.162 It is a stationary source with the potential to emit ten tons per year (tpy) of any HAP or twenty-five tpy of any combination of HAPs. For radionuclides, however, EPA can establish a different criteria.163 Note that “stationary source” is defined at section 111(a)(3) as any building, structure, facility, or installation which emits or may emit any air pollutant.164 The Administrator may establish a lesser quantity of pollutant emissions for the definition of a major source based on various characteristics of the pollutants being emitted (including potency, persistence, potential for bioaccumulation, or other relevant factors).165
An “area source” is any stationary source of HAPs that is not a major source (e.g., a dry cleaner is an area source if its emissions are below the trigger amount).166 Motor vehicles are not included.167 A “[n]ew source” is a stationary source constructed or reconstructed after regulations are proposed.168 A “hazardous air pollutant” is any air pollutant listed under section 112(b).169 An “adverse environmental effect” is any significant and widespread adverse effect, which may reasonably be anticipated, to wildlife, aquatic life, or other natural resources.170 It seems to expand the “effects on welfare” defined in section 302.
Section 112(b)(1) originally listed 189 HAPs; the list is used to determine the chemicals to be regulated within industrial source categories. This list evolved from lists of pollutants found in: the list of substances regulated under the toxic release inventory program of [*PG249]section 313 of the EPCRA;171 high-priority environmental contaminants listed pursuant to the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) section 104(i) by the Agency for Toxic Substances and Disease Registry (ATSDR);172 and pollutants in the National Air Toxic Information Clearinghouse (NATICH) database for July 1986 that had an ambient concentration established by at least one state.173 While the list is based on the lists of chemicals found in other statutes, the current list of 188 hazardous pollutants under the CAA is the second smallest compared to the other major statutes, with the CWA listing 148, OSHA listing 450, RCRA listing 502, and EPCRA listing almost 600.174 The list resulted largely from a preference by Congress for an EPA-prepared list rather than one derived from the legislative process.175 Chemicals regulated under CAA subchapter VI dealing with stratospheric ozone are not regulated under section 112.
Section 112(b)(2) allows EPA to add pollutants that present a threat of adverse human health effects or adverse environmental effects due to ambient concentration, bioaccumulation, or deposition.176 Any person can petition EPA to add to the list, and the Administrator must grant or deny the petition with a written explanation.177 EPA is currently gathering exposure data and health effect information in a two-year process while it considers listing hydrogen sulfide as a hazardous air pollutant under the CAA.178 There is also a procedure for deleting a substance from the list, which in 1996 resulted in the delisting of an HAP.179 The Administrator has eighteen months to act on a petition for delisting, and as of this [*PG250]writing there are two petitions pending to delist methyl ethyl ketone180 and methanol.181 EPA periodically updates the lists.182
Section 112(b)(4) allows the Administrator to use any authority available to obtain information on health or environmental effects of a substance.183 Thus, for example, EPA could use provisions such as the CAA section 114, CERCLA section 104(i), or several provisions in the Toxic Substances Control Act (TSCA).
EPA was to publish, within one year of enactment of the 1990 CAA Amendments, a list of categories of major and area sources emitting one or more listed HAPs.184 Categories of area sources may be listed subject to the additional requirements of section 112(c)(3), if the Agency finds a threat of adverse effects to human health or the environment (by such sources individually or in the aggregate) that warrants regulation under section 112.185
The term “stationary source” has the same meaning under section 112 as it has under section 111(a), which is any building, structure, facility, or installation which emits or may emit any air pollutant.186 However, EPA may aggregate emissions from a site even if they are emitted from sources with different standard industrial classifications (SIC codes).187 Thus, the definition of “source” may differ under section 112 from the definition applicable under section 111.188 Section 112 applies to all stationary sources emitting HAPs, thus this section covers a building, structure, facility, or installation that emits HAPs.189 Whether such source is considered “major” will [*PG251]depend upon the size and configuration of the source itself or of the larger source of which it is a part.190
A “category” of sources is a group of sources having some common features, suggesting that they should be regulated in the same way and on the same schedule.191 Thus, for example, industrial process cooling towers would be considered a source category.192 Each cooling tower emitting more than the amount of HAPs provided in section 112(a) that qualifies a source as a major source, or each cooling tower located within a larger source emitting that amount of HAPs, would be subject to MACT for major sources.193
On July 16, 1992, EPA published an initial list of categories of major and area sources of HAPs, as required under section 112(c)(1) of the 1990 CAA.194 The preliminary draft list was compiled from a number of databases, including the National Emissions Data System (NEDS) and the Toxic Release Inventory System (TRIS).195 These were augmented by a review of the available scientific literature. For example, categories of the synthetic organic chemical manufacturing industry (SOCMI) were identified from literature describing SOCMI reactants and products.196 A SOCMI category was listed if it either manufactured a chemical on the list of HAPs or if it used one or more of the listed HAPs to produce another chemical.197 Published production and consumption data for organic chemicals also were used to identify organic chemical end-user processes emitting HAPs.198 This list does not include the sources subject to the requirements under section 112(c)(6) that requires EPA to identify and control sources responsible for 90% of the emissions of certain chemicals,199 nor does it include all categories of major and area sources meeting the listing criteria in section 112(c)(1) because of [*PG252]uncertainties in the available databases concerning sources and emissions of HAPs.200 The initial list contained 174 categories—166 major and eight area sources.201 On December 3, 1993, EPA published a schedule for promulgation of the MACT standards following the ten-year guideline found in section 112(e)(1),202 and by November 15, 2000, EPA is to have completed issuing emission standards for the listed source categories.203
Under section 112(c)(5) and (9), the Agency may add or delete a category from the list.204 An addition may occur when the Administrator finds that a category meets the criteria, at which point the emissions standards shall be promulgated by November 15, 2000, or two years after the date the category is listed, whichever is later.205 A deletion may occur based on a petition of any person or on the Administrator’s own motion, upon a determination that: (1) in the case of sources that emit HAPs that may result in cancer, no source in the category (or group of sources in the case of area sources) emits HAPs in quantities that may cause a lifetime cancer risk greater than one in a million to the most exposed individual; or (2) in the case of sources that emit HAPs that may result in non-cancer adverse health effects or adverse environmental effects, emissions from sources in the category (or group of sources in the case of area sources) do not exceed a level adequate to protect public health with an ample margin of safety, and no adverse environmental effects will result.206 The Agency is to grant or deny a petition to delete a category within one year after the petition is filed.207
Section 112(c)(1) requires EPA to periodically amend the list in response to public comment or new information, and no less often than every eight years.208 The listing of categories is therefore an ongoing process. Since the initial list and schedule, there have been [*PG253]several revisions.209 On June 4, 1996, EPA revised the initial list of industrial categories subject to MACT.210 Note that the list did not use the SIC code approach that is used under the PSD, NSR, NSPS, and operating permit programs. EPA’s approach was approved by the D.C. Circuit Court of Appeals in National Mining Association v. EPA.211 The revision formally added seven major source categories, one area source category, and deleted five major source categories, and one area source category.212 More revisions were made on July 18, 1996213 and on February 12, 1998.214 On April 10, 1998, EPA promulgated additional source categories subject to the section 112(c)(6) requirements to reduce specified pollutants by 90%.215 A revised emissions standards promulgation schedule was published on May 17, 1999.216 On November 18, 1999, a revised schedule and category list was published.217
Various other Agency actions may trigger the need to list, revise, or delete categories.
(1) The Agency may list categories of area sources pursuant to a number of authorities in section 112.218
(2) The Administrator may delete categories of sources on its own motion or on petition.219
[*PG254](3) The Agency may revise the list to delineate the applicability of “case-by-case” emission standards pursuant to section 112(j), in instances where the Agency desires to delimit the affected sources because it may not establish a Federal emission standard by the deadline in the regulatory schedule for standards.220
(4) The Agency may revise the regulatory schedule for standards associated with a listed source category, heeding the limitations in section 112.221
(5) After investigation EPA may revise the regulation if it has been determined that the titles or descriptions of some categories of sources can be amended to clarify the applicability of the ensuing emission standard, or a category may be included under a different, more appropriate industry group.222
Section 112(c)(3) required that the Agency list, within five years, sufficient categories of area sources223 to account for 90% of the aggregate emissions of each of thirty or more HAPs determined to present the greatest threat to public health in the largest number of urban areas.224 Section 112(c)(6) required the listing within five years, of categories of sources assuring that the sources of at least 90% of the aggregate emissions of each of seven specific pollutants that are subject to emission standards under section 112(d) are controlled by November 15, 2000.225
Section 112(e)(4) states that, notwithstanding section 307 of the Act, no action of the Administrator listing a source category or subcategory under section 112(c) shall be a final Agency action subject to judicial review, except that any such action may be reviewed under section 307 when the Administrator issues emission standards for such pollutant or category.226
[*PG255] On July 21, 1995, the United States Court of Appeals for the D.C. Circuit ruled on a petition for review of an EPA order concerning three aspects of EPA’s definition of “major source” as applied to the CAA section 112 hazardous air pollutant program.227 The issues were:
(1) whether EPA may require all hazardous emissions from a plant site to be aggregated even if they are emitted from sources with different industrial classifications (SIC codes);228
(2) whether EPA may require “fugitive emissions” to be counted in a source’s aggregate emissions to determine whether it is a major source;229 and
(3) whether EPA may limit a source’s ability to reduce its “potential to emit” by allowing only “federally enforceable” emission controls and limitations to be used.230
EPA was challenged under the provisions of CAA section 307(d)(9), which states that EPA rules that are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,” or in excess of EPA’s “statutory jurisdiction, authority, or limitations,” must be set aside.231 Procedural errors will invalidate a rule only “if the errors were so serious and related to matters of such central relevance to the rule that there is a substantial likelihood that the rule would have been significantly changed if such errors had not been made.”232
The court dealt with the first question by noting that section 112(a)(1) says nothing about combining emissions only from sources within the same source category or SIC code.233 The court went on to say: “[i]ndeed, one could infer from § 112(a)(1) a congressional intent, in the context of hazardous air pollution regulation, to override Alabama Power.”234 The court further noted that “major source” was defined under the Title V operating permit program and the ozone nonattainment area provisions to mean a “group of sources located within a contiguous area and under common control.”235 This approach has been interpreted by EPA to mean a group of sources [*PG256]with a common industrial grouping—that is, with the same two-digit SIC code.236 But the court also held that EPA was not bound to any single definition of “major source.”237 The court held the Agency’s construction was reasonable, and it upheld EPA’s interpretation of “major source.”238 Addressing the second issue, concerning fugitive emissions, the Court of Appeals for the D.C. Circuit supported EPA’s interpretation: fugitive emissions may be counted in a source’s aggregate emissions.239
On the third question concerning limiting “potential to emit” to federally enforceable measures, the court granted the petition for review and noted, “EPA has not explained why it is essential that a control be included within a SIP.”240 The court went on to criticize EPA, stating that the CAA Amendments do not “suggest that Congress necessarily intended for state emissions controls to be disregarded in determining whether a source is classified as ‘major’ or ‘area’ under that national standard.”241 The court also noted, “EPA has not explained why the criteria for federal approval and the consequences of that approval are related to ensuring the practical effectiveness of state controls such that the set of controls considered under section 112 should be limited in that fashion.”242 Therefore, the petition for review was granted. The National Mining Association and the American Petroleum Institute sought a writ of mandamus on November 13, 1995, to force EPA to vacate the potential to emit definition per this decision.243 The Court of Appeals for the D.C. Circuit declined, stating that “[s]ince the opinion and judgment did not vacate EPA’s rule, the motion actually seeks a modification of the judgment and, as such, is not timely.”244
For each category of sources, EPA must establish the MACT.245 It is applicable to both new and existing sources.246 The standards must [*PG257]consider cost, non-air quality impacts on the environment and human health, and energy requirements.247 However, section 112 defines a minimum level of control or “floor.”248 Costs and other impacts are only used to evaluate options more stringent than the floor defined by the statute.249 This is based on section 112(d)(4) which states: “[w]ith respect to pollutants for which a health threshold has been established, the Administrator may consider such threshold level, with an ample margin of safety, when establishing emission standards under this subsection.”250
For new sources, MACT must be at least as stringent as the average emissions achieved by the best controlled sources in the same category.251 For existing sources, MACT may not be less stringent than the average emission limitation of the best performing 12% of existing units.252 Where there are few sources, the average of the best performing five sources is to be used.253 Hazardous pollutants are to be controlled by the reduction in emissions through process changes, substitution of materials or other modifications, as well as through the use of traditional pollution control devices and work practices.254
EPA does not develop control requirements for each of the listed substances.255 Instead, EPA identifies industrial categories that emit air toxics. EPA was required to publish a list of industrial categories and subcategories by November 15, 1991 for major sources and area sources.256 All major sources in each designated category must be regulated.257 In addition, CAA section 112(c)(6) requires seven specific pollutants from sources accounting for a least 90% of the aggregate emissions for each pollutant to be listed by November 15, 1995, and to be subject to standards by November 15, 2000.258 They are alkaylated lead compounds, polycyclic organic matter, hex-chloro[*PG258]benzene, mercury, polychlorinated biphenyls, 2, 3, 7, 8-tetrachloro-dibenzo-p furans, and 2, 3, 7, 8-tetrachlorodibenzo-p-dioxin.259
The statute provides a ten-year schedule for promulgating MACT standards for 174 source categories, with not less than forty categories to be regulated by November 15, 1990; coke oven batteries were to be regulated by December 31, 1992; 25% of categories to be regulated by November 15, 1994; an additional 25% to be regulated by November 15, 1997; and all categories to be regulated by November 15, 2000.260 EPA regulated all of the forty-seven source categories in the two- and four-year groups by 1997.261 After a standard is promulgated, a source that is subject to the standard has three years to comply, although a one-year extension is authorized.262 Thus, the first sources subject to the new law had to comply in late 1995 or 1996.263 If EPA does not act in a timely manner to implement a hazardous air pollution control program, then “hammer” provisions apply that require an emission limitation to be imposed on a case-by-case basis.264 The requirements are to be equivalent to the limitations that would apply to the source if emission standards had been promulgated in a timely manner.265
The basic approach in the new section 112 is to use an industrial category MACT-based program.266 However, the law does allow for more stringent requirements to be imposed to protect public health with an ample margin of safety.267 EPA is to establish standards for carcinogens that present a cancer risk greater than one in a million.268 Eight years after promulgation of MACT-based standards for a category of sources, EPA must assess the residual risk remaining after the application of MACT and, if necessary, revise the relevant standard on a health basis in accordance with the standard-setting criteria in effect before section 112 was amended.269 The National [*PG259]Academy of Sciences is responsible for evaluating the techniques relating to this risk assessment methodology.270
Area sources, which include all stationary sources of HAPs that are not major sources,271 that are responsible for 90% of the hazardous emissions of the thirty HAPs that present the greatest threat to public health in urban areas must be regulated by November 15, 2000.272 These sources may be regulated by requiring generally available control technology or management practices (GACT).273 The statute also provides for the control of HAPs from urban area sources in section 112(k)(3).
As of April 28, 2000, the final MACT standard schedules are represented in Appendix 1.274 In promulgating standards for all forty-seven source categories in the two-Year and four-Year groups, EPA estimates a reduction in air toxics of close to 1 million tpy and a reduction in criteria pollutants of approximately 1.8 million tpy at an annualized cost of approximately $672 million.275
As of June 5, 2000, there were six proposed MACT standards, represented in Appendix 2.276 As of June 6, 2000, twenty-six MACT standards were expected to be finalized in 2001 and another twenty-six finalized in 2002.277 EPA expects that by 2005, the MACT standards will reduce toxic air emissions by 1.5 million tpy.278
If EPA misses the deadline for promulgating the MACT standards, sources are still required to obtain an “equivalent emission limitation” by permit within eighteen months of the statutory deadline.279 May 15, 1999, was the statutory deadline for [*PG260]promulgation of the seven-year MACT standards.280 That deadline was extended to December 15, 1999.281 Therefore, if the applicable MACT standard is not finalized by June 15, 2001, sources must then apply for an “equivalent emission limitation” per CAA section 112(j).282 While EPA has often missed the statutory deadline for rules, it has “virtually always” issued the standards within the following eighteen months.283
Parts 60 and 61 of title 40 of the C.F.R. contain national standards required by sections 111 and 112 of the CAA before it was amended in 1990.284 Part 61 contains the NESHAP requirements imposed by the pre-1990 CAA. Because of the savings provision of section 112, these parts continued to apply but with some modifications to comply with the 1990 CAA Amendments.285 To deal with the new section 112, general regulatory provisions, located in subpart A of part 63 of title 40 of the C.F.R., codify procedures and criteria used to implement emission standards for stationary sources that emit (or have the potential to emit) one or more of the 188 HAPs listed in the Act.286 The standards applicable to specific source categories also are codified in 40 C.F.R. part 63.
On March 16, 1994, EPA promulgated a final rule covering “General Provisions” applicable to HAPs.287 HAPs are to be established for stationary source categories that emit one or more of the 188 regulated HAPs.288 A list of sources emitting listed HAPs was published on July 16, 1992.289 Each source category is to be regulated by specific standards codified in 40 C.F.R. part 63, and by the General [*PG261]Provisions that apply to all regulated sources.290 If there is a conflict between the specific requirements and the General Provisions, the specific requirements apply.291 The General Provisions cover applicability determinations (including new versus existing and area versus major sources), compliance extensions, and requests to use alternative compliance measures, compliance dates, operation and maintenance requirements, methods for determining compliance, and procedures for testing, monitoring, reporting, and recordkeeping.292 The General Provisions do not apply to regulations developed pursuant to section 112(r), concerning accidental releases, unless specified in the section 112(r) regulations.293
Different requirements may be established for major and area sources of the same source category.294 The determination of whether a facility is a major source is based on the facility’s “potential to emit” HAPs “considering controls.”295 For a physical or operational limitation on HAP emissions to limit a facility’s “potential to emit,” it must be federally enforceable.296 Thus, limits on HAP emissions, including the use of air pollution control devices, should be incorporated into an applicable permit or into the SIP to be a legal limitation.297 The major source versus area source determination is made by considering all HAP emissions from sources within a contiguous area that are under common control.298 As previously discussed, EPA has interpreted section 112 to apply to entire contiguous adjacent sites without being subdivided according to industrial SIC codes; thus, the definition for “major source” in part 63 may not be the same as the definition used in other CAA [*PG262]regulations.299 Major sources and area sources are subject to section 112 if they are an “affected source.”300 EPA has defined affected source in part 63 differently from the definition used in subchapter IV of the CAA and the related regulations.301 Under the General Provisions, “affected source” refers to a collection of process, equipment, or groups of equipment defined in standards under part 63 (including case-by-case MACT standards or equivalent emission limitations).302
Until EPA promulgates part 63 requirements for a source, section 112(g) addresses the modification, construction, and reconstruction of major sources after the effective date of the operating permit program.303 Section 112(j) provides for states to establish emission limits in operating permits if EPA fails to promulgate standards for a category of sources on a schedule set out in section 112(e).304 Under these provisions, states may be required to make case-by-case MACT determinations.305 The issue of the relationship of the General Provisions to sections 112(g) and (j) was left to future rulemaking.306 Once an operating permit program is established in a state, the “hammer” provisions of section 112(j) apply.307 If EPA fails to promulgate a standard for a source category as required by section 112(e)(1) and (3), then the permit shall contain emission limitations, on a case-by-case basis, that are “equivalent to the limitation that would apply to such source if an emission standard had been promulgated in a timely manner . . . .”308
The General Provisions are intended to implement the preconstruction review requirements of section 112(i)(1), but not the review requirements of section 112(g).309 However, EPA generally agrees that compliance with an applicable MACT standard based on section 112(d) or (h) would constitute compliance with (g).310 A case-by-case MACT determination under section 112(j) is only necessary [*PG263]when no applicable emission limitation has been established by EPA.311 The rules governing the case-by-case process are found at 40 C.F.R. section 63.43.312 On May 10, 1996, EPA issued a direct final rule governing the establishment of equivalent emission limitations pursuant to section 112(j).313
CAA section 112 recognizes that EPA could not immediately issue MACT standards for all industries and provides for emission standards to be promulgated for the initial list of sources as late as November 15, 2000.314 Historically, regulations concerning construction, reconstruction, and modification of sources subject to either NSPS or NESHAPs have been regulated at 40 C.F.R. parts 60 and 61, subpart A. CAA section 112(g)(2) prohibits any person from constructing or reconstructing a major source of HAPs unless they meet a MACT standard.315 If no applicable emission limitation exists then such determinations are made on a case-by-case basis.316 If emission standards are not feasible, then either a design, equipment, work practice, or operational standard, or combination of these approaches may be used.317 The operating permit required by subchapter V is to contain appropriate emission limitations for HAPs.318 EPA promulgated a proposed section 112(g) rule on April 1, 1994,319 and on March 20, 1996, EPA released a draft final rule.320 The final section [*PG264]112(g) rule was promulgated on December 27, 1996.321 It is narrower in scope than the 1994 proposed rule and requires MACT controls only for new facilities that are major sources for toxic air pollutants and for new or reconstructed major emitting production units at existing facilities.322 For sources for which national MACT standards have not been issued, states continue to impose MACT standards on a case-by-case basis.323 The Agency decided not to issue regulations at that time for modified sources because the section 112(g) program is primarily a transitional program designed to operate until all categories of major sources of HAPs are subject to MACT standards issued under section 112(d).324 Such sources would be subject to existing state air toxics programs, if applicable.325
The effective date of the program under section 112(g)(2)(B) in any state or local jurisdiction was the date on which a permitting authority places its implementing program into effect or eighteen months after December 26, 1996, whichever is earlier.326 Implementing programs were to be in effect by June 29, 1998.327 If at that time there was no working implementation program, either the permitting authority could issue a MACT determination with EPA’s concurrence or EPA regional administrator would issue a 112(g) determination for up to one year.328 On June 30, 1999, EPA extended the time period allowed for the regional administrator to render case-by-case MACT determinations to thirty months because some permitting authorities did not have implementing programs by June 29, 1999, nor the authority to issue MACT determinations.329
The program excludes steam generating units from coverage.330 Since only new and reconstructed major sources will be subject to section 112(g), other sources may avoid section 112(g) requirements by using design or emission controls that keep potential emissions [*PG265]below the ten tpy (or other threshold).331 Netting is not allowed.332 Alternatively, sources may avoid section 112(g) by obtaining a preconstruction minor NSR permit from the state.333
Major affected sources, or sources that become major affected sources due to reconstruction, must undergo preconstruction review.334 Affected sources that are nonmajor or unaffected nonmajor sources that reconstruct must only submit a notification.335 To construct a new major affected source or reconstruct a major source so that the source becomes a major affected source requires approval in advance from the Administrator.336 The details concerning the requirements of the application are found at 40 C.F.R. section 63.5(d).337 Reconstruction requirements also are based on the policy expressed in the December 16, 1975 Federal Register’s New Source Performance Standard requirements.338 After an affected source completes construction or reconstruction it must comply with the relevant emission standards.339 There also are requirements for startup, shutdown, and malfunction plans that can be used to avoid noncompliance penalties.340 Startup, shutdown, and malfunction reports are also required.341
The section 112 program is implemented through the operating permit program developed under subchapter V of the CAA.342 EPA’s regulations to control HAPs in 40 C.F.R. part 63 must coordinate the requirements with the operating permit program under subchapter V of the CAA.343 The operating permit program clarifies and makes a source’s air pollution requirements more enforceable by consolidating its CAA obligations in one document. On July 21, 1992, EPA promulgated regulations requiring states to develop permit programs for stationary sources including HAP sources.344 Thus the part 70 permit regulations of July 21, 1992,345 and the General Provisions applicable to all HAP sources, as well as part 63, form the basis for permit conditions.346
The definition of a “regulated air pollutant” under 40 C.F.R. section 70.2 for operating permit purposes includes any pollutant subject to a standard promulgated under section 112.347 An explanatory memo by Lydia Wegman in 1993 notes that “if a pollutant is regulated for one source category by a standard or other requirement, then the pollutant is considered a regulated air pollutant for all source categories.”348 The one exception to this rule is case-by-case MACT determinations based on section 112(g)(2).349
Sources must file periodic reports showing their compliance status.350 Regulations concerning a federally-implemented permit program, to be used in states that fail to obtain EPA’s approval of the state permit program, are found at 40 C.F.R. part 71. EPA believes it is appropriate for all sources regulated under part 63 to undergo the permitting process, but non-major sources may be subject to deferral of the subchapter V requirements.351 When a source is required to obtain an operating permit, the general provisions of 40 C.F.R. [*PG267]part 63 will be implemented through the permit program.352 When HAPs also are regulated under a SIP provision, as for example, VOC emissions, additional coordination is needed.353
On January 19, 1996, a memorandum by John S. Seitz, director of EPA’s Office of Air Quality, Planning, and Standards, was released concerning contingency plans for sources of HAPs.354 Sources of HAPs must prepare and implement startup, shutdown, and malfunction plans under 40 C.F.R. section 63.6(e)(3)(1).355 These contingency plans do not have to be included in the operating permit. Sources must state, however, that they are subject to the requirement to maintain startup, shutdown, and malfunction plans.356 The permitting authorities may require a source to present a copy of its plan for review and the plan must be kept at the source’s location.357 It is not necessary, however, for the plan to be included in the permit.358
The General Provisions include requirements applicable to all affected HAP sources, while subparts of part 63 have additional requirements applicable to specific source categories.359 Major sources also are subject to the enhanced monitoring program as proposed on October 22, 1993.360 These requirements will be incorporated into the rules being promulgated under part 63.361 This is to include Continuous Emission Monitoring Systems (CEMS)362 when required.363
[*PG268] The regulations also contain both general performance364 testing requirements as well as testing required under the individual categories developed under part 63.365 Each affected source must prepare a site-specific test plan. EPA has created a guideline document, Prepar-tion and Review of Site-Specific Test Plans, to assist in the preparation of test plans.366
The owner or operator of an affected source is required to maintain all records required by the General Provisions for at least five years following the date of occurrence, measurement, or maintenance.367 If an owner or operator believes the source is not subject to a relevant standard under part 63, a record must be kept of this applicability determination.368
In May 2000, EPA issued a proposed rule on imposing new reporting requirements for HAPs.369 Each major source must obtain an operating permit which necessarily requires a HAP inventory.370 In addition, the Urban Toxics Air Program mandates reductions by accounting for all emission control measures, and a HAPs inventory will be crucial to the development of that program.371 For HAPs reporting, EPA estimated the consolidated reporting rule will cost each state seven hundred hours annually.372 Industry will be affected by an estimated three hours annually of work and the rule will have an industry-wide cost of $844,000.373
However, monitoring and measurement requirements for HAPs are still to be resolved.374 “Of 188 HAPs, 134 have applicable [measurement] methods. [Forty-three] have likely methods. [Eleven] have potential or no methods. [Thirty-one] of the 33 high-priority urban air toxics have applicable methods.”375
The 1990 CAA Amendments added several provisions designed to encourage the use of risk assessment to control air pollutants. Under section 112(o), the National Academy of Sciences (NAS) was to review the risk assessment methodology used by EPA to determine carcinogenic risk associated with HAPs.376 The NAS was to recommend improvements in methodology including methods for assessing the adverse health effects of risks other than cancer.377 A report was to be submitted to the Risk Assessment and Management Commission and to congressional committees by May 15, 1993.378 Section 303 of the 1990 CAA Amendments established a Risk Assessment and Management Commission to investigate the policy implications and appropriate use of risk assessment and risk management under various Federal laws.379 The Commission was to be composed of ten members.380 The President appointed three members, the majority and minority leadership of Congress appointed six members, and the president of the NAS appointed one member.381
Under section 112(f), the Administrator of EPA was to report to Congress by November 15, 1996 on the methods of calculating the health risks remaining from sources subject to MACT regulation under section 112, the significance of the remaining risk, and recommendations regarding such risk.382 Under section 112(g), EPA must rank the hazardous risk of the regulated air pollutants if offsets by equally or more HAPs are to be allowed.383
The National Research Council fulfilled the NAS mandate in 1994 with the publication of Science and Judgment in Risk Assessment.384 The Congressional Commission on Risk Assessment and Risk Management began meeting in 1994, and in 1997 released its report, Framework for Environmental Health Risk Management.385 This report [*PG270]“introduces a unique Risk Management Framework to guide investments of valuable public-sector and private-sector resources in researching, assessing, characterizing, and reducing risk.”386 Volume One of the report discusses the Commission’s comprehensive Risk Management Framework.387 Volume Two is a large comprehensive risk management analysis that addresses many technical and policy issues related to health and environmental risk-based decisions. It includes recommendations for the management of residual risks that are to be regulated under the CAA section 112(f).388
In 1998, EPA issued Guidelines for Ecological Risk Assessment.389 The guidelines provide a three-phase approach to risk assessment390 and is the “basis of the residual risk approach to ecological risk assessment.”391 For human health risk assessments, EPA will rely heavily on their Integrated Risk Information System (IRIS), which is an electronic database containing information on human health effects from chemical exposure.392 Risk assessment guidelines for mutagenicity, carcinogenicity, chemical mixtures, developmental toxicity, exposure assessment, risk characterization, reproductive toxicity, probabilistic analysis, and neurotoxicity have been issued by EPA.393 The “EPA plans to establish data source hierarchies for each type of toxicity information to be used in residual risk assessments . . . .”394
EPA will model risk assessment of HAPs using the approach found in the 1989 benzene NESHAP.395 This is a two-tiered approach [*PG271]following the 1987 NRDC v. EPA case decided en banc by the Court of Appeals for the D.C. Circuit prior to the 1990 Amendments that discussed what is “safe” under CAA section 112.396 This approach was endorsed by Congress in the 1990 Amendments as illustrated by the Conference Report.
In the first step of this analysis, the Administrator must determine a safe and acceptable level of risk considering only health factors. In the second step, the Administrator may consider cost, feasibility and other relevant factors in addition to health in order to set a standard to provide an “ample margin of safety.” This approach is required under the decision in . . . NRDC v. EPA . . . and is set forth in the rulemaking on emissions standards for benzene . . . .397
EPA, therefore, “strives to provide maximum feasible protection against risks to health from HAPs”398 by limiting the maximum individual risk to 1 in 10,000 and the individual lifetime risk to 1 in 1 million.399 Despite IRIS and other data sources, EPA is not prepared to conduct full residual risk analysis due to a paucity of data on human health, ecological, and exposure data.400
On June 26, 1996, EPA issued a proposed chemical test rule under the Toxic Substances Control Act (TSCA) to assist in implementing CAA section 112.401 The proposed rule was to be the first in a series of rules used to generate health effects data to support [*PG272]emissions standards for twenty-one HAPs.402 EPA, however, repeatedly extended the comment period for the proposed chemical test rule.403 On April 21, 1998, the agency promulgated a notice of another extension of the comment period.404 The agency expects to issue a final rule in January 2001.405
The CAA Amendments create problems for those companies that reduce emissions voluntarily.406 Voluntary reductions may not count toward mandatory reductions required in the future by the CAA.407 The CAA section 112(i)(5) provides for EPA (or a state) to issue a permit under subchapter V that allows an existing source that reduces hazardous air pollution emissions by 90% to avoid the emissions reductions required by section 112(d) for six years after the compliance date that would otherwise be applicable if the 90% reduction is achieved before an applicable section 112(d) standard is proposed.408 However, states may require reductions in excess of those specified in section 112(i)(5).409
The CAA’s Early Reduction Program requires HAPs to be reduced by 90% to postpone for six years the deadline for MACT compliance.410 If a source emits high risk pollutants as specified in the regulations,411 the non-high risk pollutants can be used to offset only a fraction of the amount of required reduction of a high risk pollutant.412 Final regulations to implement section 112(i)(5) were originally promulgated on December 29, 1992413 and have been [*PG273]subject to numerous changes. They are codified at 40 C.F.R. part 63, subpart D.
To determine what pollutants are high risk EPA used three criteria: chlorinated dioxins and furans; all listed HAPs that are known human carcinogens; and listed pollutants that are toxic at or below threshold ambient concentrations.414 The toxicity is based on an estimated maximum exposure to a pollutant that can be tolerated by a human exposed continuously for seventy years without having any adverse health effects. This is called a reference concentration (RfC) and is extrapolated from animal studies. High risk HAPs are generally those with an RfC of 0.05 ug/m3, based on a two-year rat study, that are included in EPA’s Integrated Risk Information System.415 EPA determines human exposure by assuming a release of ten tpy of HAPs. Using their air dispersion model, this release will produce an ambient concentration of 5.02 ug/m3. All HAPs with an RfC that is one-tenth or more of this assumed concentration (>.502 ug/m3) are deemed high-risk HAPs.
This regulation was challenged successfully for methylene diphenyl diisocyanate (MDI). In an opinion by Judge Ginsburg, the U.S. Court of Appeals for the D.C. Circuit held that designating MDI as a high-risk pollutant based on a generic air dispersion model that bore no relationship to the properties of the chemical and that was based on an unwarranted assumption that MDI would be gaseous at temperatures at which it is a solid was an arbitrary and capricious action.416
Less than 10% of the 8000 to 13,000 major sources subject to MACT standards agreed to participate in the Early Reductions Program. Only forty applications for the six-year extensions had been filed by late 1994. Industry declined to participate because of the costs and difficulties in developing the needed historical base-year emissions data.417 There also was uncertainty concerning how state standards would relate to federal MACT standards, which could [*PG274]require industries that add controls to add more or different controls, or both, at a later date. Finally, the six-year extension offered little benefit if EPA’s regulations were delayed or if the standards provided adequate time to comply.418 Nevertheless, EPA continued the program, but shifted its implementation to its regional offices.419 In 1997 there were about twenty-five early reduction permits pending and about six permits had been issued.420 With the passage of time, this program has become largely irrelevant.
Section 112(c)(3) requires sources representing 90% of the area sources with emissions of the thirty hazardous pollutants presenting the greatest risk to public health to be regulated by November 15, 2000. Area sources also may be subject to alternative standards pursuant to section 112(d)(5).421 Mobile sources of hazardous emissions are to be regulated pursuant to section 202(l).422 Health-based standards promulgated under section 112 prior to November 15, 1990 remain in force until modified.423 Within eight years of promulgation of a standard for major or area sources the Administrator may establish additional requirements pursuant to the “residual risk” provisions in section 112(f) to protect the public health with an ample margin of safety.424 It is EPA’s expressed intention to deal with thirty-three HAPs, including thirty HAPs that are emitted from area sources, by developing an integrated program dealing with its responsibilities under sections 112(c)(3), (k), and 202(l).425 This program also will deal with a substantial portion of the HAPs being considered for regulation under section 112(d).426
Section 112(k)(3) tasks EPA with identifying not less than thirty HAPs from urban area sources, identifying the sources that represent 90% of the emissions of these pollutants and reducing the incidence [*PG275]of cancer attributable to emissions from these sources by 75%.427 Urban air is of particular concern because of the higher concentration of population and pollution in a close geographic proximity and is feasible in part because many urban areas share the same toxics profile.428
EPA promulgated a draft urban air toxics strategy on September 14, 1998, that introduced an integrated strategy to deal with HAPs.429 On July 19, 1999 EPA promulgated its Integrated Urban Air Toxics Strategy (Strategy).430 There are three risk reduction goals within the Strategy:
1. ensure a 75% reduction in cancer incidence from HAPs;431
2. substantially decrease non-cancer health risks from area sources in urban areas;432 and
3. target and address disproportionate risk in urban areas from all sources.433
The Strategy consists of four basic components:
1. regulatory tools and programming activities for source-specific and sector-based standard-setting on federal, state, and local levels;434
2. local and community efforts focusing on cumulative risks within urban areas;435
3. modeling and monitoring for urban areas under the National Air Toxics Assessments program;436 and
4. communication and education of the public.437
Thirty-three HAPs are identified in the Strategy.438 These HAPs were selected based on 1990 baseline data, and emissions of some of [*PG276]the selected HAPs have been dramatically decreased since that time due to technology-based standards.439 It is therefore imperative to develop a monitoring system and acquire comprehensive data before the implementation of any regulatory structure.440
While CAA section 112(c)(3) requires the listing of area sources representing 90% of the emissions of the thirty hazardous pollutants presenting the greatest risk to public health, the current list does not.
The current list of area source categories doesn’t include categories representing 90[%] of the emissions of each of the 30 area source HAPs. The current list meets the 90[%] or greater requirement for 11 of the 30 area source HAPs. For 10 other HAPs, the list accounts for at least 80[%] of the emissions, and for ethylene dichloride the list accounts for approximately 78[%] of the emissions. . . . The remaining HAPs on the list represent less than 75[%] of the emissions . . . . 441
It is important to note that the Strategy does not provide a regulatory regime, but seeks to integrate and expand current regulatory schemes to deal with the air toxics in urban areas after assessing the cumulative risks of HAPs exposure from various sources.442 It is a strategy, not a rule.443 EPA expects to pursue a three-tiered approach to implement the Strategy.444 The first tier will consist of MACT standards, the second tier of source category specific GACT standards, and the third tier will be a flexible GACT process.445
In meeting the Strategy’s goals, [EPA will] consider reductions in HAPs resulting, not only from actions under [EPAs] overall air toxics program . . . and measures resulting from programs to attain the national ambient air quality standards for particulate matter and ozone . . . , but also [*PG277]from State, local and Tribal measures. Further, [EPA will] consider cumulative risks presented by exposures to emissions of HAPs from sources in the aggregate.446
The resounding lack of success of regulation of air toxics prior to the 1990 CAA Amendments led many state and local governments to create their own programs to deal with the problem.447 The programs vary, but they have focused on addressing air toxics through ambient air levels, control technology standards, and risk assessment.448 In recognition of these state and local (S/L) programs, Congress created section 112(l) of the CAA to allow for the state and local programs to be implemented rather than other applicable section 112 standards.449 The difference between the “State/local residual risk assessment approach and the strategy set forth in sections 112(d) and 112(f) of the CAA is one of timing. While the CAA envisions control of HAPs from major sources as a two-step process . . . many State and local agencies consider these simultaneously.”450 In a 1995 survey, 60% of state and local air toxic programs were risk-based, and half of those addressed both new and existing sources.451
Each state may develop and submit to the Administrator for approval a program to implement and enforce the emission standards and other requirements for air pollutants that are subject to CAA section 112 or to implement the accidental release provisions of section 112(r).452 A guidance document was required to be published by the Administrator by November 15, 1991 to explain how the states should develop programs for submittal.453 After a program is submitted by a state, the Administrator has 180 days to approve or disapprove the program based on criteria in section 112(l)(5).454 If a proposed program is rejected then a state may revise and resubmit it. If a state with an approved program does not administer or enforce it [*PG278]in accordance with EPA’s published guidance then EPA can withdraw approval of the program.455 In states with delegation authority, EPA can still enforce applicable emission standards or requirements.456
On November 26, 1993, EPA promulgated regulations to provide guidance to states seeking EPA’s approval of state programs in lieu of federally promulgated HAP standards under section 112.457 This final rule contained guidance concerning rules or programs that can be implemented or enforced by states in place of certain section 112 rules.458 States may obtain partial or complete delegation; such delegation results in state programs that are substitutes for applicable federal requirements within the state, but these programs are federally enforceable.459 The November 1993 rule was subsequently amended to clarify the rule, reduce administrative burdens, and provide more flexibility to the states.460
The federal regulations governing EPA’s approval of state rules or programs under section 112(l) are located at 40 C.F.R. section 63.90. Section 112(l) and the related regulations allow state and local governments to receive EPA delegation in three ways: (1) states may substitute a state rule that is no less stringent for an EPA approved industry-specific rule;