January 01, 2013

The Cross-State Air Pollution Rule: Will EPA learn from experience?

If as Oscar Wilde said “experience is simply the name we give our mistakes,” then the U.S. Environmental Protection Agency (EPA) has had plenty of experience trying to regulate interstate air emissions. The Cross-State Air Pollution Rule (CSAPR) was simply EPA’s most recent attempt. It was short-lived. The D.C. Circuit vacated CSAPR in EME Homer City Generation, L.P. v. EPA, 696 F.3d 7 (D.C. Cir. 2012) before the rule could even take effect.

Brief history of interstate emissions regulation

Air emissions do not respect political boundaries. Emissions from one state may affect air quality—or National Ambient Air Quality Standards (NAAQS)—in other states. The Clean Air Act addresses the interstate transport of air emissions by requiring states to include provisions in their State Implementation Plans (SIPs) to regulate emissions “which will . . . contribute significantly to nonattainment in, or interfere with maintenance by, any other state with respect to any such [NAAQS].” 42 U.S.C. § 7410(a)(2)(D). This good neighbor provision, as it is called, is one of 20 requirements for SIPs in the Clean Air Act. It is at the heart of the dispute over EPA’s attempts to regulate interstate air emissions.

Congress added the good neighbor provision to the Clean Air Act in 1990. See Pub. L. No. 101-549, § 101(b), 104 Stat. 2399, 2404 (1990). After the 1990 amendments, EPA and several states formed the Ozone Transport Assessment Group (OTAG). See 63 Fed. Reg. 57,356 (Oct. 27, 1998). As its name implies, OTAG studied the transport of ozone and ozone precursors, and strategies to ameliorate the effects of such transport. OTAG, however, could not reach a consensus on its recommendations. In 1998, EPA nevertheless relied on OTAG’s work to issue a nitrogen oxides (NOx) SIP call, in which the agency formally informed 22 states and the District of Columbia that their SIPs failed to adequately address NOx emissions, precursors to ground-level ozone. Litigation followed.

The D.C. Circuit addressed challenges to the 1998 NOx SIP call in Michigan v. EPA, 213 F.3d 663 (D.C. Cir. 2000). The decision was a mixed bag. Notably, in resolving petitions filed by states, industry, and conservancy groups, the court held that EPA could consider cost in determining an upwind state’s emission reduction obligations under the good neighbor provision. Specifically, the court held that EPA could use cost to limit required reductions to “only a subset of each state’s contribution.” But the court also held that EPA erred by including Wisconsin, Missouri, and Georgia in the SIP call, and in defining Electric Generating Units.

In 2005, EPA promulgated a more comprehensive interstate transport rule, which it termed the Clean Air Interstate Rule or CAIR. 70 Fed. Reg. 25,162 (May 12, 2005). EPA’s CAIR proposal spanned 85 pages in the Federal Register and garnered significant public comment with EPA’s response to comments document totaling more than 1,000 pages. Through CAIR, EPA sought to tackle the interstate transport of precursors for both ozone and fine particulate matter, namely NOx and sulfur dioxide (SO2) emissions. To do this, EPA defined the good neighbor obligations of 28 states and the District of Columbia, allocated NOx and SO2 budgets for each, and created a cap-and-trade program for NOx and SO2 emissions. Again, litigation followed.

The D.C. Circuit initially vacated CAIR in North Carolina v. EPA, 531 F.3d 896 (D.C. Cir. 2008). There, the court explained that the good neighbor provision “gives EPA no authority to force an upwind state to share the burden of reducing other upwind states’ emissions.” Instead, “[e]ach state must eliminate its own significant contribution to downwind pollution” and EPA “may not require some states to exceed the mark.” So, while Michigan allowed EPA to use cost to lower an upwind state’s obligations, as interpreted by EME Homer, North Carolina held that EPA may not use cost to increase an upwind state’s obligations, which CAIR allowed, in part, with unlimited interstate trading of emission allowances (premised on regional cost considerations) without respect to interstate impacts. The D.C. Circuit thus vacated CAIR. But, upon petitions for rehearing the court later converted its mandate into a remand without vacatur, effectively leaving CAIR in place pending further EPA regulatory action. The court reasoned that leaving CAIR in place temporarily would “preserve the environmental values covered by CAIR.” North Carolina v. EPA, 550 F.3d 1176 (D.C. Cir. 2008) (per curiam).

Replacing a “fundamentally flawed” rule

The D.C. Circuit called CAIR “fundamentally flawed.” Yet this fundamentally flawed rule has now been in place since 2005 with emission reductions required beginning in 2009. As a result, many facilities have installed new controls at significant cost to curb NOx and SO2 emissions and have Title V operating permits that require the operation of these controls. In Florida, for example, the combination of additional controls and Title V requirements under CAIR has contributed to an approximately 64 percent reduction in annual NOx emissions between 2008 and 2011. See EPA Air Markets Program Data. Downwind air quality can only have improved as a result. So, it follows that EPA should have felt no rush to promulgate a replacement for CAIR. It did.

In August 2010, EPA proposed CSAPR. The proposal required 256 pages in the Federal Register and was accompanied by voluminous technical material. Three Notices of Data Availability followed with each proposing revisions to CSAPR’s underlying models and assumptions. Substantial public comments were submitted with EPA issuing multiple “response to comment” documents, including a 3,009 page “primary” response.

A short 12 months later, EPA issued a final rule that caught many by surprise. Among other things, emissions allocations—or budgets—for the states were significantly different than those in the proposed rule. In fact, the Office of Management and Budget observed that “the sheer magnitude of change to the budgets of all the states results in a significantly different rule than originally proposed.” Summary of Interagency Working Comments on Draft Language under EO 12866 Interagency Review, Document EPA-HQ-OAR-2009-0491-4133, at 11 (posted July 11, 2011). Simultaneously with finalization of CSAPR, EPA even issued a proposal expanding the rule’s scope to other states. Two proposals correcting errors in CSAPR soon followed.

EPA’s haste to implement CSAPR was also evident in the departure from its approach to implementing the 1998 NOx SIP call and CAIR. Unlike the earlier rules, which provided several years to comply, CSAPR imposed a compliance deadline of January 1, 2012—a mere five months from publication of the final rule in the Federal Register—and simultaneously imposed Federal Implementation Plans (FIPs). This precluded states from having any initial opportunity to develop SIPs. Naturally, litigation followed.

In EME Homer, the D.C. Circuit had before it 45 petitions for review and 18 separate motions asking the court to stay CSAPR pending review. The court stayed CSAPR on the eve of its January 1, 2012 effective date but expedited briefing on the merits of the case. On August 21, 2012, a divided court vacated CSAPR, while leaving CAIR in place, for two independent reasons.

First, the D.C. Circuit held that EPA exceeded the scope of the good neighbor provision by potentially requiring upwind states to reduce emissions in excess of their “significant contribution.” According to the court, EPA could use a numeric threshold to include a state in CSAPR—to determine that a state significantly contributed to downwind air quality problems; however, that threshold then established a floor below which EPA could not require states to reduce emissions. EPA ignored this floor by relying on a cost-based standard to determine a state’s emissions reduction target without regard for its contribution determined by the threshold. Thus, while EPA limited interstate trading under CSAPR in response to North Carolina, the court reiterated that “EPA may not use cost . . . to force an upwind State to ‘exceed the mark.’” The court also found that EPA erred by failing to account for the relative contribution of emissions from other upwind states to a downwind state and those of the affected downwind state itself.

Second, the D.C. Circuit held in EME Homer that EPA erred by simultaneously promulgating FIPs. The court explained that EPA must first quantify a state’s good neighbor obligations, if any, before it can require the state to submit a SIP. EPA cannot preemptively find a SIP deficient and then usurp the state’s prerogative to comply with the good neighbor obligations as it sees fit.

The proposed CSAPR regulation of Florida illustrates the rule’s flaws. EPA included Florida in CSAPR because of modeled links to two air quality monitors near Houston, Texas. Yet EPA’s own modeling projected that these monitors would have no air quality exceedences by 2014 without any reductions mandated by either CAIR or CSAPR. See Air Quality Modeling Final Rule Technical Support Document, Document EPA-HQ-OAR-2009-0491-4140, at Appendix B-30, B-31 (posted July 11, 2011). Moreover, EPA failed to consider the downwind state’s relative contribution to its own air quality problem. It failed to consider, for example, that mobile sources from the Houston-area emitted 153,556 tons of NOx in 2008, which dwarf EPA’s 2012 projections of 91,072 tons of NOx emissions from all Florida power plants absent CAIR or CSAPR reductions. Compare 2008 National Emissions Inventory Data, with Emissions Inventory Final Rule Technical Support Document, Document EPA-HQ-OAR-2009-0491 at 103 (posted July 12, 2011). Also, while EPA linked eight other states to these monitors, it required Florida to bear the lion’s share of emission reductions. 76 Fed. Reg. 48,208, 48,246, 48,250-251 (Aug. 8, 2011). Again, it did so without regard for Florida’s relative contribution.

Judge Rodgers filed a lengthy dissent in EME Homer. Relying on the dissent, on October 5, 2012, EPA and other intervenors filed petitions seeking rehearing en banc. That petition is still pending before the court as of this writing.

Replacing a “fundamentally flawed” rule . . . again

Should CSAPR’s vacatur withstand further scrutiny by the D.C. Circuit or even the U.S. Supreme Court, EPA must ensure that a replacement is consistent with Michigan, North Carolina, and now EME Homer. EPA could decide to a build a new rule from the ground-up or it may attempt to salvage portions of CSAPR. EPA’s recent actions suggest that it favors the latter approach.

In at least two actions since EME Homer, EPA has relied on CSAPR’s modeling (linking upwind to downwind states) to conclude that states have satisfied their good neighbor obligations. 77 Fed. Reg. 61,724 (Oct. 11, 2012); 77 Fed. Reg. 63,228 (Oct. 16, 2012). In fact, EPA specifically stated that nothing in the EME Homer opinion suggests that this aspect of CSAPR was flawed or invalid. So, according to EPA, while the D.C. Circuit may have invalidated CSAPR’s methodology for determining required emission reductions, the methodology linking upwind states to downwind states remains valid.

On remand, this initial rulemaking suggests EPA may seek to continue to rely on this aspect of CSAPR. Prognostications aside, the audience can only wait, watch, and wonder whether EPA has finally learned from experience.