DC Circuit Upholds NSR “Reasonable Possibility” Test
In a 2-1 decision in New Jersey v. EPA, released on March 5, 2021, the District of Columbia Circuit Court of Appeals (DC Circuit) held that EPA did not act “arbitrarily and capriciously” in setting the “reasonable possibility” standard that triggers additional NSR recordkeeping. In reaching this conclusion, the DC Circuit rejected challenges to New Jersey’s standing to bring the lawsuit and then, in turn rejected New Jersey’s claims that EPA had failed to adequately consider enforcement issues arising from the alleged vagueness of the reasonable possibility standard and possible underestimation of emissions. Accordingly, the DC Circuit upheld the current regulation.
This case arose out of EPA’s respond to a remand in New York v. EPA, 413 F.3d 3 (D.C. Cir. 2005) (New York II), where the DC Circuit held that “EPA acted arbitrarily and capriciously in determining that sources making changes need not keep records of their emissions if they see no reasonable possibility that these changes constitute modifications for NSR purposes.” Id. At 11. EPA remanded so that EPA could either “provide an acceptable explanation” or “devise an appropriately supported alternative” for when records were or were not required.
In 2007, EPA proposed and finalized its response, which determined that records should be kept if the amount of emissions increase was projected to exceed 50% of the relevant significant emission rate for any pollutant. EPA also provided that if a source used the “demand growth exclusion” in 40 C.F.R. § 52.21(a)(1)(xxviii)(B)(3), then any amount excluded pursuant to that provision should be added back to determine whether recordkeeping was required. New Jersey filed a petition for reconsideration with EPA and challenged the rule in the DC Circuit, which was held in abeyance while EPA reconsidered. EPA initially accepted the petition for reconsideration, but in 2019 informed New Jersey that it would not change the rule. New Jersey then resumed the DC Circuit lawsuit.
In court, New Jersey asserted that EPA’s rule was arbitrary and capricious because it (1) inadequately accounted for enforcement difficulties from the “projected actual emissions” methodology; (2) adopted an arbitrary 50% trigger, and (3) failed to explain how authorities could enforce the NSR program without records. EPA argued that it had fully considered these issues in crafting the final reasonable possibility rule and its recordkeeping and reporting requirements. Industry intervened to support EPA and also urged that New Jersey lacked standing to bring its challenge.
The DC Circuit began with the challenge to New Jersey’s standing. Article III standing requires “a concrete and particularized injury that is imminent and not conjectural, that was caused by the challenged action, and that is likely to be redressed by a favorable judicial decision.” Slip op. at 8, citing Lujan v. Defenders of Wildlife, 504 U.S. 555, 560-61 (1992). Because this blog is focused on NSR and not standing, suffice it to say that the majority determined that New Jersey’s allegation that some sources might not report their modifications properly could result in both an increased burden on New Jersey in enforcing NSR in its own state and increased pollutants that its citizens might breathe. The dissent characterized this analysis as “purely speculative” and would have denied standing.
After addressing standing, the DC Circuit turned to the merits. New Jersey argued that “inclusion of demand growth and startup, shutdown, and malfunction elements as well as netting analyses make predicting emissions under the actual-to-projected-actual methodology susceptible to manipulation” and that sources have compelling incentives to apply “unsubstantiated or overly optimistic pre-construction analyses” to evade NSR requirements. New Jersey’s ire was particularly directed to excluding analyses that fell under the 50% test using “demand growth” from mandatory reporting.
The DC Circuit found, however, that EPA had considered and addressed these issues by triggering recordkeeping provisions whenever a projected emissions increase added to excluded demand growth emissions exceeded the 50% trigger, which created a record for enforcement review. The court found no contrary evidence that EPA’s resolution did not address the enforcement concern about demand growth.
On the netting issue, the court noted that netting is not included and therefore dismissed this issue, citing 40 C.F.R. § 51.165(a)(6)(vi)(A). On the startup, shutdown, and malfunction issues, the court found that EPA’s explanation that these emissions were included on both the baseline and projected side of the ledger meant that they usually offset and that there is “no reason to expect greater amounts of these types of emissions in the post-project projections.” Once again, the court found nothing in the record challenging EPA’s resolution.
Moving to the 50% trigger, New Jersey challenged the value as “completely subjective” and “vague.” The DC Circuit rejected the vagueness challenge, noting that the rule sets forth how the emissions are to be calculated and the numeric values to which they are compared. As to the choice of 50% as the threshold, the court held that EPA “is not required to identify the optimal threshold with pinpoint precision” but most only “identify the standard and explain its relationship to the underlying regulatory concerns,” which the court held that EPA had done. The court noted that EPA had stated it was balancing burdens on both sources and states, had sought comment on the threshold, and that half of the commenters had supported the 50% (or higher) trigger. The court also noted that EPA adjusted the rule from the proposal to provide more recordkeeping for cases where demand growth was invoked and that might cause the significant threshold to be exceeded. The court held that EPA had reasonably determined that its trigger would capture projects that had a higher risk of noncompliance, which was the gravamen of the court’s remand.
Finally, the court rejected New Jersey’s challenge that EPA’s reliance on non-NSR records was impermissible, holding that EPA was not required to achieve “perfect NSR compliance” and that its obligation was “to analyze the trade-off between compliance improvement and the burdens” record imposed and to “articulate a reasoned judgment as to why any proposed additional burden would not be justifiable in terms of the likely enhancement of compliance.” The DC Circuit held that, on the overall record, EPA had considered the concerns raised by New Jersey and other commenters and articulated a reasonable justification for its position. The DC Circuit thus affirmed the “reasonable possibility” rule and dismissed New Jersey’s petition for review.
The decision in New Jersey v. EPA is important because it resolves the last major remaining challenge to the 2002 NSR Reform Rule, which replaced the prior “actual to potential” test with the new “actual to projected actual” test, introduced plantwide applicability limits, and made other improvements to the NSR programs. The decision is also welcome because it upholds a reasonable accommodation between the competing workability and enforceability elements of the NSR program. In this case, EPA considered the concerns raised by New Jersey and explained how it sought to address them and the court found that explanation adequate.
The most disappointing aspects of this case are (1) New Jersey’s efforts to undo a rule after 12 years, without apparent thought to the consequences on sources complying with the extant rule during that period; and (2) the majority’s blithe acceptance of New Jersey’s theory that sources will engage in criminal NSR evasion unless the regulators review every action that a source takes. Judicial proliferation of the unwarranted assumption that the regulated community at large is criminal undermines the ability of the regulated community, regulators, and larger community to engage with each other to solve common problems and is deeply regrettable.