Master The D.C. Circuit (en banc) upheld EPA's two-step interpretation of Clean Air Act §112's "ample margin of safety" standard for hazardous air pollutants while remanding portions of the benzene NESHAP for inadequate justification. with this comprehensive case brief.
Natural Resources Defense Council v. EPA is a foundational administrative and environmental law decision interpreting the pre-1990 Clean Air Act's hazardous air pollutant program. Confronted with EPA's benzene National Emission Standard for Hazardous Air Pollutants (NESHAP), the D.C. Circuit sat en banc to resolve competing claims by environmental petitioners and industry about what it means to set a standard that provides an "ample margin of safety to protect the public health." The court endorsed EPA's two-step risk-based framework—first identifying an "acceptable risk" and then adding a further buffer as an "ample margin of safety"—while policing the agency's reasoning and record support for specific regulatory choices and exemptions.
This opinion is widely cited for articulating how agencies may incorporate risk assessment, feasibility, and limited cost considerations when a health-based statute demands protection with a margin of safety but does not specify a numeric risk level. It also illustrates Chevron deference, the de minimis doctrine, and the requirement that agencies articulate a rational connection between facts found and choices made. Even though Congress revamped §112 in 1990, this case continues to inform residual-risk rulemakings and the broader law of risk regulation.
824 F.2d 1146 (D.C. Cir. 1987) (en banc)
Benzene, a known human carcinogen, was listed by EPA as a hazardous air pollutant under §112 of the Clean Air Act, which required standards that protect public health with an "ample margin of safety." In 1984, EPA promulgated a benzene NESHAP targeting emissions from several industrial source categories (including fugitive equipment leaks and certain production processes). To implement §112's health-protection mandate, EPA adopted a two-step risk policy: (1) identify an "acceptable risk" level using risk assessment tools (including the maximum individual lifetime risk, or MIR, metric) and (2) add an "ample margin of safety" by tightening controls further, taking into account uncertainties, the distribution and magnitude of risks, numbers of people exposed, and the technological and economic feasibility of additional reductions. In the benzene rule, EPA treated an MIR on the order of 1 in 10,000 as within the range of acceptable risk (subject to context) and then adjusted standards to provide additional protection. The rule also contained certain categorical distinctions and limited exemptions EPA characterized as de minimis. Environmental petitioners (including NRDC) argued EPA unlawfully tolerated excessive cancer risk and improperly factored costs, while industry petitioners contended that the statute did not compel such stringent health-based controls and that EPA's risk policy was arbitrary. Multiple petitions for review were consolidated before the D.C. Circuit sitting en banc.
Does Clean Air Act §112 permit EPA's two-step interpretation of the "ample margin of safety" mandate—first determining an acceptable level of risk and then adding a margin of safety informed by feasibility and other factors—and were EPA's specific benzene NESHAP choices (including risk thresholds, subcategorization, and de minimis exemptions) reasonable and supported by the administrative record?
Under Chevron, when a statute is ambiguous, an agency's reasonable interpretation controls. Section 112 requires EPA to set standards that protect public health with an "ample margin of safety," which the court held does not compel a zero-risk standard or pure cost-benefit balancing. EPA may (1) identify a level of "acceptable risk" grounded primarily in health protection and scientific risk assessment (e.g., the MIR and related risk metrics), and then (2) provide an additional "ample margin of safety" that accounts for uncertainties, the distribution and magnitude of risk, and technological and economic feasibility considerations to the extent consistent with health protection. EPA may adopt narrow de minimis exemptions where regulation would produce trivial or no health benefits, but only if those exemptions are consistent with the statute's purpose and supported by a reasoned explanation and record evidence. Agency action must reflect a rational connection between the facts found and the choices made.
The court upheld EPA's two-step "acceptable risk" and "ample margin of safety" interpretation of §112 as a permissible construction under Chevron. It approved EPA's use of MIR and related risk metrics and permitted consideration of feasibility and limited cost factors at the margin-of-safety step (but not as a substitute for health protection). However, the court remanded portions of the benzene NESHAP where EPA failed adequately to justify specific thresholds, subcategorizations, and de minimis exemptions, and where the record did not sufficiently support the agency's choices or consideration of risk distribution and uncertainties.
Text and structure: The phrase "ample margin of safety" is inherently indeterminate and does not fix a numeric risk threshold or require absolute zero risk. The statutory focus on public health supports EPA's health-first approach while allowing room to address uncertainties and feasibility in determining the margin of safety. Chevron deference: Because §112 does not dictate a precise risk level or methodology, EPA's two-step framework reasonably fills the statutory gap. The court rejected NRDC's view that §112 mandates the lowest technologically achievable (or zero-risk) standard, and it likewise rejected industry's contention that the statute invites open-ended cost-benefit analysis that could subordinate health protection. Risk methodology: The court approved EPA's use of MIR to anchor the "acceptable risk" inquiry, recognizing MIR's role in protecting the most exposed individual while acknowledging that other considerations—such as the number of people exposed, background exposures, and cumulative risks—are relevant when determining the margin of safety. The court emphasized that EPA must make clear how it balances these considerations consistent with the statute's health-protective purpose. Feasibility and costs: The court held that EPA may consider technological and economic feasibility when providing an "ample margin of safety," but not at the initial step of identifying what risk levels are acceptable from a health perspective. The statute does not authorize pure cost-benefit tradeoffs that would compromise health protection, but it does allow feasibility to inform how far beyond the "acceptable" level the margin of safety should extend. De minimis exemptions and record support: While agencies may craft narrow de minimis exemptions (consistent with the court's de minimis jurisprudence) where regulation would yield trivial benefits, EPA must justify such exemptions with evidence and reasoned explanation. The court found parts of the benzene rule insufficiently explained—e.g., certain thresholds, subcategory lines, or exemptions lacked adequate risk quantification or failed to account for uncertainties or population risk—and therefore remanded those elements for further consideration. In sum, the court sustained EPA's overarching interpretation but required the agency to shore up its reasoning and record for specific regulatory choices.
This decision is a touchstone for environmental and administrative law. It crystallizes the now-classic two-step "acceptable risk/ample margin of safety" framework under §112 and delineates the permissible role of feasibility and limited cost considerations in a health-based statute. It also reinforces Chevron deference while demanding rigorous, record-based reasoning—illustrating how courts both defer to agency policy choices within statutory bounds and police the quality of agency justification. Even after Congress overhauled §112 in the 1990 Clean Air Act Amendments (adopting technology-based MACT standards followed by a residual-risk review), this case continues to guide EPA's residual risk determinations under §112(f) and remains widely cited for risk assessment, de minimis authority, and the articulation of a margin-of-safety standard.
The U.S. Court of Appeals for the D.C. Circuit decided the case en banc in 1987. The full citation is 824 F.2d 1146 (D.C. Cir. 1987) (en banc).
It is a two-step concept. First, EPA identifies an "acceptable risk" level based primarily on health science (e.g., the MIR and related risk metrics). Second, EPA adds an "ample margin of safety" by tightening controls further to account for scientific uncertainties, the distribution of risk and numbers of people exposed, and technological and economic feasibility—without engaging in pure cost-benefit balancing that could undercut health protection.
No. The court rejected a rigid zero-risk mandate and did not impose a one-in-a-million (10^-6) bright line. It accepted EPA's use of an MIR on the order of 1 in 10,000 (10^-4) as potentially within the zone of "acceptable risk," provided EPA justifies its choice with record support and then adds an ample margin of safety. The precise risk level is not fixed by the court; it must be reasoned and evidence-based.
Costs and feasibility may inform the second, margin-of-safety step—how far beyond the acceptable risk EPA should go—but they cannot drive the initial determination of what risks are acceptable from a health standpoint. The statute's focus is health protection, not open-ended cost-benefit analysis, though feasibility can shape the extent of the protective margin.
The court recognized that EPA may adopt narrow de minimis exemptions where regulating would yield only trivial or no health benefits, but only if those exemptions are consistent with §112's protective purpose and are supported by a reasoned explanation and record evidence. In NRDC v. EPA, the court remanded certain exemptions and categorical lines because EPA had not adequately justified them.
In 1990, Congress shifted §112 to a technology-based MACT regime followed by a residual-risk review. NRDC v. EPA's two-step framework remains relevant to that residual-risk phase (§112(f)), which explicitly references protecting public health with an ample margin of safety. The decision continues to guide how EPA evaluates risk, uncertainty, and feasibility when determining whether additional controls are needed after MACT.
NRDC v. EPA established the bedrock two-step framework for implementing §112's mandate to protect public health with an ample margin of safety. By endorsing an "acceptable risk" determination followed by a further protective buffer that may consider feasibility and uncertainties, the court reconciled a health-first statute with practical implementation and scientific indeterminacy.
Equally important, the decision exemplifies modern administrative law: it applies Chevron deference to an ambiguous statutory phrase while insisting on reasoned, evidence-supported explanations for each regulatory choice. For law students, it offers a clear map of how courts evaluate agency risk regulation, accommodate scientific uncertainty, and cabin the use of cost considerations under health-protective mandates.
Need to cite this case?
Generate a perfectly formatted Bluebook citation in seconds.
Use our Bluebook Citation Generator →