EPA Proposes Sweeping NSR Changes

On February 22, 2024, EPA released a pre-publication draft of its proposed rule on “Prevention of Significant Deterioration (PSD) and Nonattainment New Source Review (NNSR): Regulations Related to Project Emissions Accounting,” expected for imminent release.  The proposed rule proposes sweeping changes to the definition of “project,” requests comment on whether to adopt additional controls to prevent double counting of emissions reductions; proposes to require that all Step 1 emissions reductions must be “enforceable,” and proposes numerous changes to the “reasonable possibility” recordkeeping requirements that would subject substantially all projects to reporting requirements while also increasing the required detail. 

ANALYSIS

Readers may recall that EPA issued the initial Project Emissions Accounting interpretation in a memorandum in 2018, followed by a final rule in 2020.  Petitions for reconsideration were filed.  Opponents charged that EPA’s proposal did not assure that projected emissions reductions would actually occur; that they were not enforceable; that they may not be “contemporaneous,” that sources would combine “unrelated” projects to avoid NSR, and that it would be impossible to enforce.  While EPA rejected those comments at the time, EPA stated that it would consider those petitions in a future rulemaking. This proposal represents EPA’s proposed action on the issues raised by commenters opposing the Project Emissions Accounting changes and the petitions for reconsideration. 

Definition of Project

EPA states that it agrees with the concerns expressed that sources would combine unrelated projects to avoid NSR.  It also stated that it believes that the Project Emissions Accounting memo and rule were based upon the 2018 reconsideration of “project aggregation” that adopted the “substantially related” test for determining when projects are and are not related.  With this understanding, EPA proposes to revise the definition of “project” for PSD and NNSR purposes as follows:

Project means a discrete physical change in, or change in the method of operation of, an existing major stationary source, or a discrete group of such changes (occurring contemporaneously at the same major stationary source) that are substantially related to each other. Such changes are substantially related if they are dependent on each other to be economically or technically viable. [Additional language for NNSR in 40 CFR 51.165:  In an extreme ozone nonattainment area, a “project” means each discrete operation, emissions unit, or other pollutant-emitting activity.]

EPA clarifies in the preamble that it is not changing the 3-year rebuttable presumption that “projects” separated by more than three years are generally not “substantially related,” although it caveats this statement by noting that phased projects likely are substantially related and suggests that projects on the same process line are more likely to be substantially related than projects on unrelated parts of the facility.  EPA also states that it has not proposed a meaning for “contemporaneously” used in this definition but requests comment on whether such a definition should be established.  For NNSR purposes, EPA clarifies the definition to reflect the statutory provision that “any change” at an existing major source in an extreme ozone nonattainment area must go through NNSR and that this obligation cannot be avoided by taking Step 1 reductions.

EPA states that it is proposing this definition to address concerns raised by commenters opposing the PEA because it would allow “unrelated” projects to be combined avoiding applicable NSR requirements by merging unrelated emissions reductions with a project that would otherwise trigger NSR.  EPA reiterates that it also remains concerned about disaggregation of projects to avoid NSR requirements and urges state and local permitting authorities to review all cases where applications for major and minor NSR permits or multiple minor NSR permits are submitted in a short time period.

The definition of “project” will also be relevant to EPA’s proposed changes to the Reasonable Possibility reporting, discussed below.

Double Counting

EPA requests comment on whether it should adopt rules that prevent “double counting” of emissions reductions, giving an example of a prior project at a source that is causing emissions to go down and the source including those emissions reductions in its forecast to demonstrate that the significant emission rate (SER) is not exceeded.  EPA postulates that this may be necessary because such emissions reductions are not “caused by” the project and therefore should not be considered in the calculation.  EPA expresses some concern about the complexities that could become involved.

Requiring Enforceable Step 1 Emissions Reductions

One of the principal concerns of opponents of the PEA rule is that the emissions reductions are not enforceable either initially or that they will be maintained over time.  EPA states that it agrees with this concern and proposes to amend the PSD and NNSR rules to require such enforceability:

The “sum of the difference” as used in paragraphs (c), (d) and (f) of this section shall include both increases and decreases in emissions calculated in accordance with those paragraphs. A decrease may only be accounted for in the significant emissions increase determination if it meets the requirements [for Step 2 enforceability].

Proposed § 51.166(a)(7)(iv)(g).  EPA states that this is necessary to address the requirement in CAA section 110 that “each implementation plan submitted by a State include enforceable emission limitations.”  EPA notes that “there may be reason to require that such decreases be enforceable,” citing the “predominant impact that one or more claimed decreases in emissions involved in a project could have on the determination of whether the project constitutes a major modification” and that this justifies “additional safeguard … to ensure that such decreases actually occur and that they are maintained.”  It states that the changes it is proposing to “reasonable possibility” reporting may be enough to show that decreases actually occur but not if the decreases don’t occur as projected.  EPA expresses concern that “source owners or operators may overestimate emissions decreases at Step 1 with no recourse provided actual emissions are not significant.”  EPA requests comment on this and whether it is appropriate to treat decreases and increases differently.  EPA also requests comment on whether Project Emissions Accounting should be revoked and only increases considered.

"Reasonable Possibility" Recordkeeping and Reporting

EPA proposes to reanalyze “the trade-off between compliance improvement and the burdens of data collection and reporting” in the proposal and decides that substantial additional recordkeeping and reporting is required.  Additional recordkeeping and reporting requirements that EPA identified as necessary include:

• A description of the project that includes the name of the project, the project’s intended objective(s), each physical change and/or change in the method of operation associated with the project objectives(s), and estimated timeline for the project, including an estimation of when the project would begin actual construction and begin regular operation;

• A description of the applicability test used to determine that the project is not a major modification, including the baseline actual emissions, the projected actual emissions, the amount of emissions excluded under demand growth and an explanation of why excluded, and the potential to emit, as applicable, and any netting calculations, if applicable.

• Before beginning actual construction, all sources that in Step 1 exceed 50% of the significant emission rate without considering any reductions or exclusions must submit the information listed above to the permitting authority, but is not required to receive any determination from the reviewing authority.

• If a source uses an emissions decrease of any kind to determine that the project is not a major modification, then reasonable possibility recordkeeping and reporting are required.

The EPA also clarifies existing policy that emissions increases from new, modified, and non-modified units “that could realize a change in utilization as a result of the project, including increases resulting from removal of a process bottleneck (what we often call ‘de-bottlenecking)” must be included.  EPA states its belief that these steps are necessary “to ensure that unenforceable emissions decreases … will … be qualitatively equivalent to the increases they purportedly offset.”  EPA also clarifies its belief that “differ” requires both reporting when emissions exceed both the significant emission rate and the projections and when the decreases that were proposed are not what was forecast and may exceed the baseline emission rate by a significant amount.

EPA also expresses concern that if a source comes in below the reasonable possibility threshold, “the reviewing authority may not be able to verify that activities were properly aggregated and that decreases accounted for in the NSR applicability process actually occur” and justifies expanding the prior notice to all sources and to any project involving any decrease to address this risk.  EPA states its belief there will be no increase in burden because units should be doing it anyway and submitting it as part of the minor NSR process.  EPA claims that it has received feedback from state permitting authorities and environmental groups that existing recordkeeping is insufficient, but cites only a single comment on the PEA rule.

Revisions to Netting in NNSR

EPA proposes to clarify that in Serious and Severe areas, all net emissions increases over five years must be added together and that Project Emissions Accounting is not permissible at all in Extreme areas, based on its reading of the Clean Air Act.

Implementation

EPA states that the proposed rule would take effect immediately in directly administered and delegated states.  EPA proposes to find that the proposed revisions would modify the “minimum program elements” and hence all states with SIP-authorized programs must submit revisions within three years.

COMMENTARY

From a source owner/operator or permitting authority’s perspective, there is little to like in the proposed rule.  Aspects of the proposal lend themselves to unending litigation and second guessing, which does little to improve permitting but increases the burdens on permitting authorities, reduces innovation as sources seek to minimize exposure to regulatory risk, and degrades public confidence in the NSR program.

Definition of Project

EPA is not well-advised to tackle the definition of “project.”  Projects depend upon a myriad of technical and economic details including current and forecast market conditions that may or may not occur and which can easily be upset by external forces.  EPA’s proposal also thrusts permitting authorities into the difficult position of trying to determine whether projects at a source are or are not “substantially related.”  EPA further amps up the difficulty by pronouncing that aggregating too much is a violation, aggregating too little is a violation, and submitting too many permits should be scrutinized as evidence of a possible violation.   

Determining whether projects are substantially related is challenging for counsel with years of experience in the program and industry and is an unreasonable burden to thrust upon permit writers, who often have limited experience with the industry or available time to fiddle with the question.  It is also beside the point:  the courts have repeatedly held that only a “significant emissions increase” that is actual (for modifications) counts.  If what is done would causally exceed the significant emissions rate, then NSR is required; if it doesn’t, the statute does not create a carte blanche for others to slice the project up and say “if the source had done it another way, NSR would have been triggered, therefore NSR is required.”  It is likely that Congress sought to regulate the projects being done, not hypothetical alternative projects.

In short:  the 2018 rule got it right.  This proposal gets it wrong.  Splitting up a project to avoid NSR needs to be prevented.  Adding projects together and keeping the total under the significant emission rates poses no more threat to the environment than a single “discrete” project below the significant emission rate and actually may be better because the combined projects are likely reducing pollution on a per unit basis.  EPA’s rule has no such benefit.

Double Counting

This is a solution in search of a problem.  If emissions from the facility are trending downward and that trend is expected to continue, that is a relevant fact in forecasting the projected future actual emissions.  It doesn’t matter whether it is “caused” by the project or not.  The “demand growth exclusion” expressly recognizes that non-project factors may impact projections.  EPA is correct not to propose any new language to address putative “double counting.”

Step 1 Enforceability

One can legitimately debate the merits of whether Step 1 emissions should or should not be enforceable as a practical matter.  On the environmental side, enforceability makes perfect sense: sources should “be compelled” to achieve their emissions reductions.  In reality, it is less clear.  The success of many projects is unclear and permitting often precedes formal commitment of capital expenditure to support the project.  A source owner/operator may well find itself in the position of having to choose between taking enforceable emissions restrictions on existing equipment when it is uncertain that the project will ever occur.  If the project fails, the source is saddled with those restrictions and no benefit.  This risk—heightened by EPA’s proposal--is a powerful deterrent to innovation or even replacement of obsolescent machinery, likely resulting in higher pollution per unit of production over time.

In at least one respect, EPA’s concern is misplaced.  Most “modifications” involve replacement of at least some equipment.  For replaced equipment, there is no question that the emissions reductions for that equipment are achieved because the equipment is gone.  If the equipment is gone, so are the emissions.

We also note that on March 1, 2024, the D.C. Circuit rejected EPA’s claim that Section 110 requires “all SIPs” to specify “emissions limitations,” finding that emissions limitations are only required when “necessary and appropriate.”  EPA relies upon the requirement that SIPS must include "emissions limitations” to support its contention that Step 1 requires “enforceable limitations” yet it is difficult to see why limits are “necessary and appropriate” for equipment replacements because the reduction will occur in any event.  If EPA persists with its proposal unchanged, it could be vulnerable to a summary reversal.

Reasonable Possibility Recordkeeping and Reporting

EPA agreed with opponents of Project Emissions Accounting that much more recordkeeping is required to make the program enforceable and ensure that reductions occur.  As outlined above, EPA adopted sweeping, burdensome changes and applied them to almost all projects.

First, EPA should drop the reasonable possibility reporting for replaced equipment.  Equipment that is gone has no “reasonable possibility” of emitting. 

Second, the blanket extension of reasonable possibility reporting to any project involving a decrease far exceeds what can be supported.  The Clean Air Act and its implementing regulations confers jurisdiction over emissions increases that EPA has defined as “significant.”  It is not clear that a trivial decrease included in a project meets the standard for regulation articulated in Alabama Power v. Costle or in any of the subsequent legislation.

Third, the proposed recordkeeping is significantly too expansive.  EPA requires identification of “each physical change and/or change in the method of operation associated with the project objective(s).”  This definition is not even limited to new or modified units but also includes affected units and, in fact, is not limited to “units” at all.  Further, EPA requires identification of “change in the method of operation” but this is a poorly defined concept.  As written, the proposal may well be void for vagueness.  It should be limited to emissions units.  Reporting timelines should be eliminated or clearly defined as non-enforceable because they provide little meaningful information as most projects deviate from timelines due to construction issues, local permit delays, contractor availability, and supply chain glitches. 

Nonattainment NSR

These changes are useful clarifications.

Summary

If EPA wanted to force more litigation on its state and local permitting partners, it could hardly have chosen a better tool.  Every aspect of the proposal—particularly the redefinition of project and the hopelessly overbroad notice requirements—creates additional litigation opportunities to challenge permitting authority action.  Even less desirably, the proposal forces permitting authorities to fight on ground which they cannot reasonably navigate because it requires them to be experts in the underlying industry's technology and economics.  In an era where Congress and the President are trying to move big projects forward to prepare for economic transformation and climate change, it seems like EPA is moving in the opposite direction.