EPA Issues NSR Reform Package

David Friedland and Laura McAfee
Beveridge & Diamond, PC
Washington, DC
DFriedland@bdlaw.com; LMcAfee@bdlaw.com

After years of debate, EPA has decided to proceed with long-anticipated reforms to the Clean Air Act's New Source Review (NSR) program. The NSR program requires companies to obtain preconstruction permits and install stringent control technology for major new sources and modifications to those sources that exceed certain thresholds.

In a package released June 13, EPA concluded that for power plants and refineries, "the NSR program has impeded or resulted in the cancellation of projects that would maintain or improve reliability, efficiency, or safety." Letter from C. Whitman, Administrator, EPA, to the President (June 13, 2002) at 1 (on file with EPA). The impacts extend to many other industries as well - "at existing industrial facilities outside the power plants sector, the NSR program does discourage projects that improve capacity or efficiency but do not result in increases in actual air emissions." Id. The NSR Reform package will be implemented over the coming months and years as a final rule on some issues (those that were proposed by EPA in 1996 under the Clinton administration, including plantwide applicability limits (PALs), exemptions for "clean units" and pollution control projects, and the "actual to future actual" methodology) and a proposed rule for notice and comment on other issues (including clarification of the "routine repair" exemption and of EPA's debottlenecking and aggregation policies). The rules will be very controversial, and will likely trigger a protracted legal and political battle. A group of 44 senators has written to the President, urging him not to go forward with the reforms. Letter from John Edwards et al., Senators, to C. Whitman, Administrator, EPA (Aug. 1, 2002).

At press time, only summaries of what will be included in the rules are available, the most detailed of which are EPA's "New Source Review: Report to the President" (June 2002) (Report), and "New Source Review: Recommendations" (June 2002) (Recommendations). The rules themselves may differ from the summary documents. In this article, we discuss four key reforms, as described in the Report and the Recommendations: 1) the routine repair exemption; 2) the actual to future actual test; 3) debottlenecking; and 4) aggregation of emissions.

1. Routine Maintenance, Repair and Replacement

Activities that constitute routine maintenance, repair and replacement (routine repair) are excluded from NSR permitting. 40 CFR § 52.21(b)(2)(iii)(a). Over the years, there has been a significant dispute between industry and EPA over what constitutes routine repair. EPA applies a four part test to determine whether a particular capital project is routine, evaluating the purpose, nature and extent, frequency, and cost of the project. This test was upheld by the Seventh Circuit in 1990. See WEPCO v. Reilly, 893 F.2d 901, 910-12 (7th Cir. 1990) (WEPCO).

In the reform package EPA will propose three broad reforms to the routine repair exclusion: (1) new language authorizing replacements with functionally equivalent components; (2) clearer guidelines on the types of projects that qualify for the exclusion; and (3) new, cost-based "safe harbors." We discuss each of the reforms below.

Functional equivalence. Over the past few years, industry has argued that EPA follows an unnecessarily strict interpretation of "routine repair," which limits the exemption to projects using precisely the same component parts. For example, when Detroit Edison sought to replace turbine blades with a more energy-efficient design, EPA rejected the company's claim that the project was "routine repair," because the company was not using identical turbine blades. Letter from Francis X. Lyons, Regional Administrator, USEPA, to Henry Nickel, Counsel for Detroit Edison Co. (May 23, 2000). EPA justified its finding by explaining that, because the project would cost substantially more than that "typically associated with routine blade maintenance activity," would further the purpose of significantly enhancing efficiency, and would result in efficiency great than that which would have been achieved by simply replacing deteriorated blades with the same type of blades, the project could not be deemed a "routine repair." Id.

In the reform package, EPA agreed that its narrow interpretation "may discourage some energy efficiency improvements." Report at 16. To resolve this concern, EPA proposes to issue new regulations clarifying that replacement components need not be precisely identical. Rather, as long as the replacement equipment "serves the same function and ... does not alter the basic design parameters of the unit [,]" the project would be exempt from NSR review. Recommendations at 5.

Types of projects. Under EPA's four-part test discussed above, the purpose of a capital project is a key factor in determining whether it qualifies for the exemption. However, existing EPA guidance does little to identify what types of "purposes" would support applicability of the exemption. Moreover, the Agency routinely took the position that efficiency-related projects were not entitled to the exemption.

Under NSR Reform, EPA proposes to adopt new regulations providing "clear guidelines" for the types of projects that may typically be considered routine repair. Specifically, projects undertaken to "facilitate, restore, or improve efficiency, reliability, availability, or safety within normal facility operations" may qualify for the exemption. Recommendations at 5. Energy efficiency projects are specifically singled out as routine: "EPA will affirm that existing NSR rules are not intended to discourage activities that increase efficiency. The Agency will propose that energy efficiency improvements undertaken through routine maintenance, replacement, and repair activities will be considered to be RMR&R." Id. at 6. EPA may further publish a list of particular projects that will automatically be considered "routine" for particular types of sources; however, the absence of a particular project from such a list would not automatically disqualify it from the exemption, but would rather simply require a case-by-case evaluation. Id.
at 5.

"Safe harbors." Finally, EPA proposes to adopt "safe harbors", i.e., automatic exemptions for projects that fall within certain cost thresholds. EPA proposes to bring two NSPS provisions into the NSR program: 1) the 50 percent "reconstruction" threshold; and 2) the "capital expenditure" concept, which exempt projects if their cost falls below the annual asset guideline repair allowance, as set forth by the IRS for that particular industry category Recommendations at 4. Projects that exceed these cost thresholds would not automatically trigger NSR, but would instead be reviewed on a case-by-case basis. Id.

EPA may not, however, exempt all low-cost projects from NSR under this provision. The reform package mentions EPA's concern that the safe harbors "might allow a facility to undertake relatively low-cost projects (such as installation of new burners or painting equipment) that can increase emissions significantly and should not automatically be excluded from NSR." Recommendations at 5. Thus, EPA may identify specific types of projects that are not eligible for exemption.

2. Actual to Future Actual Test

Under the current regulations, NSR applies only to "major modifications," defined as a change that results in a "significant net emissions increase" - that is, where the "increase in actual emissions" from the change exceeds certain thresholds. 40 C.F.R. § 52.21(b)(2)(i), (b)(3)(i), (b)(23). Many in industry read the phrase "actual emissions" to mean what it says, i.e., actual emissions to the atmosphere. The definition of "actual emissions" supports this interpretation, referring to "the actual rate of emissions of a pollutant from an emissions unit[.]" Id.
§ 52.21(b)(21)(i).

However, the regulations also state that when a unit has not "begun normal operations," EPA may use the facility's post-change "potential to emit" - that is, the maximum capacity of the unit to emit under its physical and operational design. § 52.21(b)(21)(iv), (b)(4). EPA has taken the position that every modified unit by definition has not "begun normal operations," and so post-change emissions must be based on the facility's potential to emit, i.e., presuming that the facility will operate at maximum capacity 24 hours a day, 365 days a year. See Fact Sheet appended to Report, at 2.

Industry has long argued that EPA's use of potential to emit instead of actual emissions causes "paper" emissions increases to trigger NSR, when in practice, a project would not increase emissions to the atmosphere, and may even decrease emissions. The courts have not clearly resolved this issue. In Puerto Rican Cement v. EPA, 889 F.2d 292 (1st Cir. 1989), the First Circuit upheld EPA's use of potential to emit as applied to the replacement of two old cement kilns with one new kiln of a different, improved design, reasoning that the more efficient design could induce the company to increase usage. On the other hand, in WEPCO, the Seventh Circuit rejected the use of potential to emit as applied to rebuilt utility boilers, finding "no support in the regulations for the EPA's decision wholly to disregard past operating conditions at the plant." 893 F.2d at 917. Notably, Puerto Rican Cement involved construction of a new unit, while WEPCO involved modifications to an existing unit, so the two cases are not clearly at odds with each other.

EPA now recognizes that comments received regarding the NSR program have portrayed "a credible case that some capacity or efficiency projects that do not increase actual emissions are not undertaken because they trigger NSR under the actual-to-potential test." Report at 21. Accordingly, as part of the NSR Reform package, EPA plans to finalize proposed regulations that require the "actual to future actual" test for a much larger universe of projects. For new units and complete unit replacements, EPA would continue to apply the actual to potential test. For all other modifications, however, EPA would apply the actual to future actual test. Recommendations at 3. The actual emissions baseline would be the highest consecutive 24-month period over the past ten years, as modified by the source's "current emissions factor" (i.e., taking into consideration any new emissions limitations or reductions that have occurred since that time). Id.

EPA will also confirm that there is a "causation" requirement in NSR permitting, that is, under the express language of the regulations, only emission increases "resulting from" a particular change may be considered in determining whether the change triggers NSR review. See 57 Fed. Reg. at 32326-28 (1992) (the so-called "WEPCO" rule for electric utilities). EPA plans to clarify that this requirement is not limited to the electric utility industry:

Consistent with pre-existing statutory and regulatory requirements, only emissions increases caused by a given change are considered in measuring the emissions increase associated with the change. In particular, as part of the actual to projected future actual methodology, EPA will continue to apply the causation test incorporated into the WEPCO rule. EPA will exclude from the emissions increase calculation that portion of the post-change emissions that both: (1) could have been accommodated before the change within the representative baseline period; and (2) is attributable to an increase in projected capacity utilization at the unit that is unrelated to the particular change.

Recommendations at 3.

3. Debottlenecking

The NSR regulations focus upon the specific unit that is modified. See, e.g., 40 C.F.R. § 52.21(j)(3) (requiring the installation of control technology on the unit that is modified and that experiences an emissions increase). Some projects, however, can impact emissions from upstream or downstream units, as well - for example, if a source removes a bottleneck in a production line by replacing a small unit with a larger one.

EPA has given conflicting guidance as to whether emissions increases from these upstream and downstream units must be considered in determining whether NSR is triggered. While most of the Agency's published guidance states that these emissions must be included; e.g., Letter from K. Henry, EPA Region 3, to J. Daniel, VA Dep't of Envtl. Quality (Nov. 23, 1998), other guidance has led companies to believe the opposite. See, e.g., Letter from J. Daniel, VA Dep't of Envtl. Quality to K. Henry, EPA Region 3 (Sept. 23, 1998)(stating that verbal and written guidance provided at an EPA-sponsored NSR workshop, as well as verbal guidance from named EPA employees, led him to believe that these emissions did not have to be included).

As part of NSR Reform, EPA will propose new regulations clarifying that these upstream and downstream emissions generally need not be evaluated. Rather, when calculating actual emissions associated with a physical change or change in the method of operation, "sources generally should look only at the unit undergoing the change." Recommendations at 6. "Upstream" or "downstream" units should be considered "only when the permitted emissions limit of the upstream or downstream unit would be exceeded or increased as a result of the change." Id.

4. Aggregation

EPA's aggregation policy has caused similar confusion. Generally, to determine whether a modification causes a "significant net emissions increase," a facility must compare net emissions increases from the proposed project against all other "contemporaneous" increases and decreases (all other increases and decreases over a five-year period prior to the project). 40 C.F.R. § 52.21(b)(3). However, EPA has long taken the position that if a project is not "significant" in and of itself, then other contemporaneous increases and decreases need not be considered. See "New Source Review Workshop Manual" (EPA Draft 1990) at A.36. In other words, if a particular project is limited to installing a unit which, by itself, does not have the potential to emit a "significant" amount of any pollutant, then that project is exempt from review regardless of any other contemporaneous emissions increases or decreases.

The one exception to this rule is where a facility deliberately splits a single, "significant" project into several smaller projects to avoid NSR review. Thus, for example, if a facility installs two units, each of which has the potential to emit 30 tpy of SO2, installing both units at the same time would trigger NSR, because the 60 tpy increase in SO2 emissions exceeds the 40 tpy "significance" level for SO2. If the facility instead attempts to avoid NSR by installing the units six months apart, the emissions from both units will be aggregated, the project will require permitting, and the facility could be subject to an enforcement action for attempting to circumvent NSR.

Over time, however, EPA has expanded the exception beyond the "sham" permitting situations the Agency originally targeted to require aggregation of small projects even in the absence of an intent to circumvent NSR. For example, in 1992, EPA overturned Minnesota's determination that various small projects at a 3M facility over an 18-month period were separate minor modifications that did not trigger NSR. See Memorandum from J. Rasnic, Dir. SSCD, EPA, to G. Czerniak, Chief, Air Enforcement Branch, EPA Region 5 (1992). 3M justified the project by explaining that each was "individually conceived and separately funded" by the company's various research divisions, an explanation that EPA appears to have accepted. Id. at 5. Nevertheless, EPA suggested that NSR review was required even in the absence of any intent to circumvent the law, because the projects were "consistent with the source's overall production goals or plans for a short planning period." Id. at 4.

As part of the NSR Reform package, EPA plans to clarify through a new rulemaking that projects generally are not aggregated together in determining whether NSR applies. Specifically, projects are treated as "separate and independent" unless "the project is dependent upon another project to be economically or technically viable[,]" or the two projects have been intentionally split to avoid NSR. Recommendations at 6.