## Facts

- Several oil companies operated offshore drilling platforms in Cook Inlet, Alaska, and related onshore separation facilities handling produced water from the offshore operations.
- The operations generated deck drainage (mixed liquids including oil and grease), produced water (co-produced with oil and gas), and sanitary wastes from crews on the platforms.
- The companies sought Clean Water Act § 402 NPDES permits that would allow offshore discharges of these wastes and limited discharges from the onshore facilities.
- EPA issued NPDES permits containing numeric effluent limitations (including oil and grease concentration limits stated as averages and maximums), derived from performance data and statistical methods intended to reflect “best practicable control technology currently available” (BPT).
- The permits allowed only limited discharges and provided narrow tolerance for treatment-system upsets and malfunctions.
- EPA held permit hearings that included adjudicatory features, but the presiding officer considered some material outside the hearing record and the proceedings did not follow APA formal adjudication requirements.
- The companies filed consolidated petitions for review in the Ninth Circuit challenging both the effluent limitations and EPA’s procedures.

## Issues

1. Whether EPA’s permit effluent limitations exceeded the Clean Water Act’s BPT requirement under §§ 301(b)(1) and 304(b)(1).
2. Whether EPA’s statistical methods and data assumptions used to derive daily and monthly numeric limits were inadequately supported or unreasonable.
3. Whether the permits unlawfully failed to account for inevitable operational upsets and malfunctions, making compliance impracticable.
4. Whether the Clean Water Act’s “opportunity for public hearing” requirement compelled APA formal adjudication procedures for NPDES permits, and whether consideration of extra-record material invalidated the permits.
5. Whether EPA’s permitting procedures violated due process.

## Decision

- The court upheld EPA’s general authority to set technology-based numeric effluent limits reflecting BPT and to use statistical techniques to convert performance data into enforceable limits.
- The court rejected the claim that NPDES permit issuance necessarily required APA formal adjudication procedures; the Clean Water Act required an opportunity for public hearing, not full APA trial-type procedures.
- The court found no due process violation based on the hearing structure, notice, and opportunity to present and challenge evidence.
- The court remanded to EPA for the limited purpose of reconsidering and/or better supporting specific aspects of the statistical analysis and the treatment of operational variability, including upsets and malfunctions.
- A dissent would have required stricter procedural protections and was more critical of the evidentiary and analytical foundation for the limitations.

## Legal Principles

- BPT is a technology-based standard that may be set by reference to performance of well-designed and well-operated control systems; it is not confined to a particular permittee’s current practices.
- EPA may employ reasonable statistical methods to translate treatment performance data into daily maxima and monthly averages, but the agency must explain and support key assumptions sufficiently to allow meaningful judicial review.
- Effluent limits tied to BPT must reflect practical operational conditions; agency reasoning should address predictable variability and the role of upsets and malfunctions in compliance and enforcement.
- The Clean Water Act’s requirement of an “opportunity for public hearing” for NPDES permits does not, by itself, trigger APA formal adjudication procedures under 5 U.S.C. §§ 554, 556, and 557.
- Combining investigative and decision-making roles in permit issuance does not, without more, establish unconstitutional bias; due process focuses on fair notice and a meaningful chance to be heard.

## Conclusion

The Ninth Circuit largely sustained EPA’s approach to Clean Water Act NPDES permitting and technology-based effluent limits, including the use of statistical tools, but ordered a limited remand for EPA to revisit and better justify particular analytical choices and the handling of operational upsets to ensure the limits were practicably achievable under BPT.


Marathon Oil Co. v. Environmental Prot. Agency, 564 F.2d 1253 (9th Cir. 1977)