Over the last three years, the U.S. Environmental Protection Agency (EPA) has moved quickly on several fronts to set a new course on Clean Air Act (CAA) regulation. From promulgation of the first ever greenhouse gas emission regulations to new ambient air quality standards to unprecedented regulation of fossil-fuel fired utilities and industrial boilers, the Obama administration has undertaken a complete makeover of the federal clean air program. EPA’s recently issued Cross-State Air Pollution Rule, addressing interstate transport of certain air emissions in the eastern United States, is one example of the push to expedite this new air quality vision. It also illustrates EPA’s willingness to test the limits of its authority under the CAA. While the new rule applies exclusively to fossil-fuel power plants, according to EPA, the rule will establish the methodology for future interstate transport rules, which will likely apply to industrial facilities as well. The new rule is the subject of over fifty petitions for administrative reconsideration, forty-five petitions for judicial review, and eighteen motions asking the court to stay the rule pending review. The challenges call into question both the process and substance of the final rule.
Cross-State Rule basics
The Cross-State Rule replaces the Bush administration’s Clean Air Interstate Rule (CAIR). CAIR was vacated on numerous grounds by the D.C. Circuit, but on reconsideration in December 2008, the court left CAIR in place pending EPA’s remand rulemaking. North Carolina v. EPA, 531 F.3d 896 (D.C. Cir. 2008), reconsidered at, 550 F.3d 1176 (2008). The remand rule, proposed August 2, 2010, was originally known as the Clean Air Transport Rule (75 Fed. Reg. 45,210 (Aug. 2, 2010)), but was renamed the “Cross-State Air Pollution Rule” or CSAPR in the final rule. 76 Fed. Reg. 48,208 (Aug. 8, 2011). While CAIR and CSAPR share the same basic framework, as the name change suggests, CSAPR is very different from its predecessor.
CSAPR is designed to implement the “Good Neighbor” provision of the CAA. The CAA is based on a framework of cooperative federalism in which EPA develops national ambient air quality standards and states implement those standards through State Implementation Plans, or SIPs. The Good Neighbor provision requires SIPs to include rules sufficient to prevent air emissions within that state from significantly contributing to air quality problems in other states. 42 U.S.C. § 7410(a)(2)(D)(i)(I). CSAPR specifically addresses interstate air quality impacts under the 1997 ground-level ozone and annual fine particulate matter (PM2.5) air quality standards. CSAPR also addresses interstate impacts under the 2006 daily PM2.5 standard. Central to both rules are three separate cap and trade allowance programs that target sulfur dioxide (SO2) and nitrogen oxides (NOx) emissions. Both of these pollutants are precursors to the formation of PM2.5 year-round. NOx emissions are also a precursor to the formation of ground-level ozone, but only during the warmer months. Thus, the rule creates an annual trading programs for SO2 and NOx to reduce PM2.5 and a separate ozone seasonal NOx trading program (effective May to September) to reduce ozone.
One of the “fundamental flaws” the D.C. Circuit identified for the prior rule, CAIR, was that it allowed unlimited interstate trading of emissions allowances without respect to interstate impacts. The court found that unlimited trading violated the Good Neighbor provision by allowing states to essentially buy the right to contribute significantly to another state’s air quality problems. In response, CSAPR places significant limits on interstate allowance trading. Specifically, CSAPR establishes geographic limits by splitting the annual SO2 program into two distinct trading areas, and it also sets allowance volume limits for each state under all of the programs, effectively capping emissions for each affected state. These limits make CSAPR much more stringent than CAIR.
While the overall geographic reach of the two rules is similar, CSAPR includes fewer New England states and expands the programs west. Massachusetts, Connecticut, and Delaware are out, but Nebraska, Oklahoma, and Kansas are in. CSAPR maintains CAIR’s two-phased emissions reduction approach, with the first reductions required in 2012 and additional reductions required in many states by 2014.
The methodology that EPA says will be its template for future interstate transport rules is very complex. CSAPR establishes a multi-step process that relies heavily on modeling to identify and quantify (i) interstate air quality impacts, (ii) a state’s “significant contribution,” and (iii) emissions caps. In this process, EPA first identifies areas projected to have ambient air quality concerns, either nonattainment or maintenance problems, under the relevant standards. EPA then identifies each state that has a measurable impact on the identified areas of concern in other states. A state is “linked” to air quality concerns downwind if its contribution to ambient levels of ozone and/or PM2.5 is greater than one percent of the applicable ambient standard. Once a state has been linked, the next step is to quantify the state’s “significant contribution,” that is, the emissions that must be reduced to address the downwind problems.
To quantify significant contribution, EPA models the targeted emissions—in this case, SO2 and NOx from power plants—and identifies emissions reductions that are both cost effective and resolve all or most of the downwind problems. This quantification analysis is regional, not state-specific. EPA selects a regional investment level for emissions reductions (e.g., $500 per ton of SO2 removal) and then assesses the modeled impacts on each area if all states linked to that area make reductions achievable at that level of investment. The cost-effective reductions that achieve the desired result are defined as the state’s “significant contribution” to be eliminated under the Good Neighbor provision. Then, EPA establishes state-by-state emissions budgets or caps for each pollutant based on emissions remaining after eliminating that “significant contribution.”
Rush to regulate
Although the D.C. Circuit declined to set a deadline for EPA’s remand rule, EPA has moved very quickly to finalize the new rule—some say too quickly. EPA published the proposed rule in August 2010 and finalized it just one year later. The proposal, detailing EPA’s new methodology, exceeded 1,300 pages before it was published in the Federal Register and was accompanied by numerous technical support documents detailing the modeling that is the basis for the rule. Despite the volume and complexity of the rule, EPA denied requests for extension of the sixty-day comment period, including requests from numerous states.
During and following the public comment period on the proposal, EPA issued three successive Notices of Data Availability, each announcing proposed revisions to the models and assumptions underlying the rule. For example, the agency announced new versions of the models used to project future emissions and emissions impacts, new emissions inventories, and new fuel price assumptions. EPA also announced alternative methods for allocating emissions allowances to individual units under the program.
Notwithstanding the new data releases, the agency did not issue a supplemental proposed rule using the new models and data, which would have shown the impacts of the many proposed revisions on the state emissions budgets and allowance allocations. As a result, the final rule caught many states and affected sources by surprise. Based on the updated models and assumptions, many states and sources received substantially reduced emissions budgets and allowance allocations as compared to the proposal. Some were cut by almost half. In addition, based on the new modeling, some states were included in programs they were not subject to in the proposal. For example, the final rule included Texas in all three regulatory programs whereas it was only included in the seasonal NOx program in the proposal. Several states and industry petitioners argue that EPA’s rulemaking approach in this case violates basic public notice and comment requirements of the CAA and the Administrative Procedure Act.
The same day EPA finalized the rule, it began revising it. With the final rule, EPA simultaneously issued a proposal to expand the number of states subject to the seasonal NOx program. In October 2011, EPA proposed additional revisions to the final rule, including changes to the emissions budgets for several states to address errors in the underlying data and assumptions.
Despite significant changes from the proposal and the fact that the final rule is still in flux, compliance is required in short order. The annual SO2 and NOx reduction requirements take effect January 1, 2012—less than five months after publication of the final rule. The seasonal NOx reduction requirements begin May 1, 2012.
To achieve this unusually short compliance deadline, EPA has issued CSAPR as a Federal Implementation Plan (FIP), rather than allowing states to develop SIPs. That is, EPA not only determines the amount of emissions that must be reduced in each state to comply with the Good Neighbor provision, but it also determines how those reductions will be achieved in each state by deciding which sources must reduce and setting the allowance allocations for those sources, including set asides for new sources on a state-by-state basis. Numerous comments on the proposal took issue with this FIP-first approach as contrary to the CAA’s basic principle of cooperative federalism that calls for EPA to set ambient standards but states to determine implementation of those standards. In the final rule, the agency defended the FIP-first approach by taking the position that the CAA required the states to submit SIPs designed to eliminate significant contribution within three years of EPA’s promulgation of the ambient standards. States, on the other hand, argue that they have EPA-approved rules in place that implement CAIR, and following the remand, it was EPA’s responsibility to provide states with a replacement rule redefining their “significant contributions” (i.e., the new amounts states are obligated to reduce via their SIPs).
There are numerous other challenges to the rule as well. For example, several states and utilities argue that CSAPR unlawfully forces them to shoulder a portion of other states’ emissions reduction obligations under the Good Neighbor provision.
CSAPR illustrates this EPA’s unusual sense of urgency, even at the expense of procedural obligations under the CAA and the Administrative Procedures Act, and its willingness to test the limits of its authority under the CAA. The D.C. Circuit left no question that CAIR requires a makeover—the question is whether this version will fare any better upon review.
On December 30, 2011, the U.S. Court of Appeals for the District of Columbia Circuit issued an order staying CSAPR and reinstating CAIR pending judicial review.