National Parks Conservation Ass'n v. U.S. Environmental Protection Agency
788 F.3d 1134
| 9th Cir. | 2015Background
- EPA issued a Federal Implementation Plan (FIP) for Montana after the State declined to submit a SIP; the FIP set BART-based emission limits at Colstrip Units 1–2 (NOx and SO2) and limits for the Corette station.
- PPL Montana (operator/part-owner of Colstrip and Corette) and conservation groups (NPCA et al.) both challenged the FIP: PPL argued the controls/limits were too stringent; NPCA argued they were too weak.
- Key contested controls: for Colstrip Units 1–2, EPA required SOFA + SNCR for NOx (rejecting SCR) and a fourth scrubber for SO2; no new technology was required at Units 3–4 (post-1977) for BART purposes.
- Petitioners raised procedural and substantive objections: EPA’s cost-effectiveness analysis, baseline emissions choices, CALPUFF visibility-model results and margins of error, and internal consistency between Colstrip and Corette analyses.
- The Ninth Circuit reviewed EPA’s explanations under the arbitrary-and-capricious standard and remanded parts of the Rule for failure to provide reasoned explanations for certain BART and emissions-limit decisions.
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| 1) NOx BART at Colstrip Units 1–2 (SOFA+SNCR vs. SCR) | PPL: EPA failed to explain why SOFA+SNCR is required and SCR rejected; cost-effectiveness unexplained. | EPA: discretionary BART choice; costs and benefits disclosed; no bright-line metric required. | Court: Vacated — EPA’s decision arbitrary and capricious for failing to explain how it determined cost-effectiveness among options. |
| 2) SO2 BART at Colstrip Units 1–2 (fourth scrubber) | NPCA: EPA should have required more advanced replacement scrubbers; PPL: EPA underestimated cost and failed to justify cost-effectiveness. | EPA: incremental scrubber justified; replacement not required because existing removal >50%. | Court: Vacated — EPA did not adequately explain why incremental scrubber cost is reasonable given the small visibility gain. |
| 3) Consistency between Colstrip and Corette analyses | PPL: EPA’s differing results (finding controls cost-justified at Colstrip but not at Corette at similar $/ton rates) are inconsistent. | EPA: considered site-specific factors; concluded controls at Corette were not cost-justified. | Court: Vacated in part — EPA’s terse explanation for Corette was inconsistent with Colstrip analysis and unexplained. |
| 4) Use of CALPUFF and model predictability for small visibility changes | PPL: CALPUFF’s margin of error makes predicted improvements (e.g., ~0.085 deciviews) unreliable; EPA must show improvements are "reasonably anticipated." | EPA: CALPUFF is the approved model; regulations allow addressing sub‑perceptible impacts; model uncertainty acknowledged and 98th-percentile method reduces error. | Court: Vacated/remanded as to CALPUFF response — EPA failed to meaningfully address comment that predicted incremental improvements were within model error. |
Key Cases Cited
- Marsh v. Oregon Natural Res. Council, 490 U.S. 360 (deference to agency technical expertise)
- Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29 (agency must provide reasoned explanation; arbitrary-and-capricious standard)
- Am. Corn Growers Ass’n v. EPA, 291 F.3d 1 (vacatur where controls yield no appreciable haze benefit)
- Columbia Falls Aluminum Co. v. EPA, 139 F.3d 914 (agency must respond to modeling challenges)
- Chem. Mfrs. Ass’n v. EPA, 28 F.3d 1259 (agency need not rejustify approved model for each application, but must respond to specific model criticisms)
- Sierra Club v. EPA, 719 F.2d 436 (internal inconsistency in agency analysis is arbitrary)