Fifth Circuit Upholds SILs
On April 29, 2024, in Sierra Club v. LDEQ, --- F.4th ---, the Fifth Circuit Court of Appeals upheld the Louisiana Department of Environmental Quality's (LDEQ's) use of EPA’s “significant impact levels” (SILs) to both define when cumulative modeling is required and to determine whether a source “contributes” to a violation against a Sierra Club argument that Clean Air Act (CAA) section 165 requires “any” source contribution arguably above a national ambient air quality standard (NAAQS) to require further controls to prevent such impact. The Fifth Circuit also found that LDEQ’s permitting approach did not violate “best available control technology” (BACT) requirements or Louisiana’s “public trust” doctrine.
ANALYSIS
Commonwealth LNG sought a PSD permit and Title V permit from LDEQ for a proposed LNG export facility on the Calcasieu Ship Channel. It submitted an application and modeling showing that the emissions from the facility were below the SILs for all but four pollutants, that cumulative modeling for three of the four pollutants were below the NAAQS and that for the fourth, while the cumulative modeling for the 1-hour NO2 NAAQS was higher than the NAAQS level, the proposed facility’s contribution was below the SIL. Commonwealth proposed to control NOx at a level of 2.5 ppm. Sierra Club commented, challenging the use of SILs as violating the NAAQS, the use of AP-42 to set emission factors, and contending that BACT should be set at 2.0 ppm and not 2.5 ppb. Sierra Club also argued that granting the permit would violate the Louisiana public trust doctrine. LDEQ disagreed, responding to each of Sierra Club’s contentions, and issued the PSD and Title V permits. Sierra Club appealed to the Fifth Circuit.
The initial issue confronting the Fifth Circuit was whether it had jurisdiction. LDEQ argued that appeals of the PSD and Title V permits were state law matters. The Fifth Circuit disagreed, holding that because the state provisions were included in the state implementation plan (SIP) that they were “federal,” that the Natural Gas Act placed challenges to permits under “federal” law in the circuit court of appeals, and therefore the challenges to the air permits could be heard in federal court. Accordingly, it held it had jurisdiction over Sierra Club’s appeal.
The Fifth Circuit next took up Sierra Club’s contention that LDEQ could not use SILs to avoid conducting a full modeling review because the CAA and EPA’s guidance do not allow for de minimis exceptions to an applicant’s responsibility to show that they will not cause an exceedance of the NAAQS, citing Bluewater Network v. EPA, 370 F.3d 1 (D.C. Cir. 2004). LDEQ and Commonwealth countered that the Act does not specify details about the modeling and that EPA’s modeling guidelines should be respected. Ultimately, the Fifth Circuit held Sierra Club’s argument was not well founded, noting that the D.C. Circuit had found that “contribute” is “ambiguous” and subject to regulatory interpretation and that EPA’s interpretation explains that the SILs analysis shows that the NAAQS are not being exceeded. Based on this rationale, it held that LDEQ has not abused its discretion in using the SILs to streamline the scope of modeling review.
The next issue was whether LDEQ could use the SIL as a basis to show that the Commonwealth facility was not “contributing” to an exceedance. Sierra Club argued that there can be no exceedance. LDEQ’s response to Sierra Club’s comments on this issue had been to note that a “modeled” exceedance does not meant that there will be an actual exceedance with an explanation of why this was the case due to the assumptions used in modeling. LDEQ also compared the modeled exceedance against the SIL and found that the amount of exceedance was much less than the SIL, suggesting an insignificant risk of exceedance. The Fifth Circuit similarly found that this approach was not an abuse of discretion.
Sierra Club next challenged the use of AP-42 factors, arguing that they are not the “best” source of data, citing a 2020 EPA Enforcement Alert, and that if AP-42 factors were to be used, the “maximum” and not “average” value should be used. On the first argument, Commonwealth and LDEQ defended the use of AP-42 data on the basis that the facility was not yet built and it is therefore not possible to have site-specific values that were “better.” LDEQ then explained that is is “unlikely” that every piece of equipment would be emitting at the highest rate at the same time and this supports the use of the average factor as reasonable. The Fifth Circuit found that the use of AP-42 in this situation was appropriate.
Turning to BACT, Sierra Club had argued that a selective catalytic reduction system at 2.0 ppm should be used. LDEQ had explained its belief that the SCR systems would not achieve these levels because the application is different than the combined cycle electric generating turbines. The Fifth Circuit upheld LDEQ’s “reasoned discussion” on this point. On the oxidation catalyst system, Sierra Club had argued for a catalytic oxidation system but LDEQ found that system was not appropriate because they required the use of a scrubber and generated additional waste. The Fifth Circuit held that this was also a “reasoned” decision and refused to disturb it.
The last issue was Sierra Club’s challenge under the Louisiana public trust doctrine. LDEQ argued that this was not a “federal” issue but the Fifth Circuit held that it was incorporated into the SIP by the adopted statutes. It then reviewed the public trust doctrine claims, but found that LDEQ had adequately addressed the public trust through its consideration of the “IT questions” and showing that the burdens had been minimized to the extent appropriate. It rejected arguments that the public trust necessarily requires “more” than the CAA, at least in this situation.
Based on this analysis the Fifth Circuit upheld the LDEQ permits.
COMMENTARY
The Fifth Circuit’s Sierra Club v. LDEQ case is important because it addresses and rebuts the effort by Sierra Club and similar groups to eliminate any de minimis consideration under the CAA permitting programs. Instead of accepting the argument that “any” exceedance of the NAAQS calculated in any fashion is necessarily a violation of the CAA, the court accepted LDEQ's and EPA’s evaluation that impacts below the SILs are unlikely to be an “actual” increase that would “contribute” to a violation of the NAAQS. It is important to stress, however, that the SILs are not a panacea – they must be used with understanding and an explanation of the context and why that context show that an exceedance of the NAAQS will not occur.
The case is also useful for blunting the increasing number of challenges to AP-42.
The only unusual and potentially troubling aspect is the Fifth Circuit’s apparently growing acceptance of review of state agency decisions directly, without requiring challengers to first go through the state system. This will be an interesting development to watch and one that has the potential to undermine state program autonomy given differences in federal versus state court review.