In a letter to Mr. Bart Cassidy dated June 14, 2018, EPA’s Office of Air Quality Planning and Standards (OAQPS) rejected an argument in the context of plantwide applicability limits (PALs) that a unit should be considered “new” for purposes of determining baseline actual emissions (BAE) if the BAE period fell within 2 years of initial startup of the unit. Instead, OAQPS states that a unit is “new” only if, at the time of the permit application, it is within 2 years of unit startup.
Mr. Cassidy, on behalf of an unidentified client, wrote to EPA requesting clarification that “an emissions unit that had commenced operation within two years of the selected baseline period constitutes a ‘new source’ under the NSR regulation.” Letter, at 1 (quoting clarification request).
OAQPS began its response by quoting the PSD PAL regulations, which state that the PAL level “shall be established as the sum of the baseline actual emissions (as defined in paragraph 40 CFR 52.21(b)(48)) of the PAL pollutant for each emission unit at the source; plus an amount equal to the applicable significant level for the PAL pollutant.” 40 CFR 52.21(aa)(6)(i). The regulations then reference the definition at 52.21(b)(48)(iii), which defines a “new emissions unit” as “any emissions unit that is (or will be) newly constructed and that has existed for less than 2 years from the date such emissions unit first operated.” OAQPS argued that this definition of “new emissions unit” contains no reference to the baseline period nor does it provide that a “new” emissions unit is one that began operating within two years of that period. EPA further argued that the present tense (“that has existed”) used in the regulation supported its position, whereas if the interpretation urged was true the regulation should read in the past tense (“that had existed”), and further that if the interpretation urged was true the regulation should refer to the baseline period to frame the point of reference. Because neither element was present, OAQPS argued that the requestor’s interpretation was implausible and inconsistent with the regulation itself.
OAQPS then cited the preamble, which states:
You must first identify all your existing emissions units (greater than 2 years of operating history) and new emissions units (less than 2 years of operating history since construction).
67 Fed. Reg. 80186, 80208 (Dec. 31, 2002) (emphasis added in EPA letter). OAQPS read the “first” as requiring the identification of unit as “existing” or “new” prior to the identification of the baseline period, which has not yet been identified at this point in the process. OAQPS also noted that the preamble states that any units, whether “currently classified” as new or existing, that were added after the PAL period would have their PTE used. Id. OAQPS argued that "currently classified" in this context clearly refers to the present time, not the past.
Next, OAQPS quoted the Technical Support Document (TSD), where a question about how the BAE is calculated was asked. In that response, EPA had stated “We agree with the commenter in the case of a new emissions unit (unit that not does not yet have a 2-year operating history) that the baseline emissions rate should be the unit’s potential to emit, since a unit with less than 2 years of normal operation at the time of a physical or operational change does not have sufficient operating history to determine its actual emissions.” TSD, at I-2-13 (Nov. 2002). Finally, OAQPS noted that the requestor’s interpretation:
would allow an emission unit with more than a decade of actual operational data to continue to use its potential to emit as its baseline actual emissions rate. This is certainly not what we intended in the definition of baseline actual emissions adopted in the 2002 NSR final rule.
Letter at 3. OAQPS thus concluded: “all emissions units that have existed and operated for 2 or more years from the time of the PAL permit application are "existing emissions units" that contribute to the PAL level at an amount equal to average rate, in tons per year, at which the unit actually emitted the pollutant during the selected 24-month period.”
This is an interesting letter. For those concerned that EPA is giving up the NSR program to industry interests, its insistence that actual emissions be used rather than PTE is an encouraging sign of rigor. Nevertheless, the question, based solely on the regulatory text, is likely a closer one than OAQPS would like to recognize, but if consideration is given to the wider context of the NSR program and its supporting materials, then OAQPS is pretty clearly correct in its interpretation. One factor OAQPS did not consider is the concern of the New York court that any “actual” increase in emissions be considered in determining whether a modification occurs, and this factor would also support OAQPS’s conclusion.