A Review of 1-Hour NO2 NAAQS Air Quality Modeling

During the development of this 4TR article, proposed revisions to Appendix W to 40 CFR Part 51 “ Guideline on Air Quality Models (Guideline) were published in the Federal Register. As a result of revisions which propose to establish Tier 3 nitrogen dioxide (NO₂) atmospheric chemical reaction screening techniques as default methods, applicants will no longer need to gain U.S. EPA approval after the Guideline revisions are finalized. In the meantime, the guidance provided below related to the Tier 3, NO₂ atmospheric chemical reaction screening techniques must be provided for U.S. EPA regional approval for permits issued before the Guideline is finalized. The Guideline is anticipated to be finalized in the spring of 2016.

I’ve been involved in quite a few natural gas combined-cycle power plant projects and it seems like the critical issue is related to showing compliance with the 1-hour NO₂ National Ambient Air Quality Standard (NAAQS) [188 micrograms per cubic meter (μg/m³) or 100 parts per billion (ppb) based on the average (5 years or 1 year depending on whether you using NWS or onsite data) of the 98^th percentile of maximum daily 1-hour concentrations]. This is especially challenging when evaluating start-up related emissions. As natural gas combined-cycle power plants have become more efficient, it has led to the ability to start-up faster to quickly meet energy demand. Even though these units are more fuel efficient than ever, the fact that they can now be started up in less than one hour (usually daily, Monday through Friday) to meet peak energy demand has actually made complying with the 1-hour NO₂ NAAQS more difficult for two (2) reasons:

1. Since these more efficient units start-up in less than one hour, the start-up emissions can’t be averaged across multiple hours and
2. The total number of start-ups that occur in a year.

For these two (2) reasons, State agencies don’t feel that start-up NO₂ emissions meet the definition of an intermittent emissions unit as summarized in U.S. EPA’s March 1, 2011 Memorandum “Additional Clarification Regarding Application of Appendix W Modeling Guidance for the 1-hour NO₂ NAAQS and are, therefore, requiring start-up NO₂ emissions to be evaluated as continuously occurring (i.e., the most conservative approach). Because of interpretation, we have had to work through some less conservative options for showing compliance with the 1-hour NO₂ NAAQS for evaluating start-up NO₂ emissions and even for normal operation process sources.

First of all, gone are the days of using the conservative assumption of full conversion of nitric oxide (NO) to NO₂, which is referred to as the Tier I approach. In all of the 1-hour NO₂ modeling that we have completed, a less conservative Tier 2 or Tier 3 approach has been required to demonstrate compliance with the 1-Hour NO₂ NAAQS. The current status of these methods is summarized well in U.S. EPA’s September 30, 2014 Clarification of the Use of AERMOD Dispersion Modeling for Demonstration Compliance with the NO₂ NAAQS memorandum. However, I’ve summarized them below with some things to keep in mind if using either of these options.

• Tier 2 Method multiply Tier 1 results by empirically derived NO₂/NO_Xratios with the national default ratio of 0.8 for hourly NO₂.
• Tier 3 Method detailed screening methods may be used on a case-by-case basis. At this time, the Ozone Limiting Method (OLM) and the Plume Volume Molar Ratio Method (PVMRM) are considered to be appropriate screening techniques.

The Tier 2 ambient ratio method (ARM) is a simple approach that can be used that does not require additional approval beyond your State agency. In addition the current version of AERMOD has an option to automatically perform the ratio. Depending on the magnitude of your 1-hour NO_X emission rate and surrounding terrain, this approach is a viable option for demonstrating compliance with the 1-hour NO₂ NAAQS in certain scenarios. Another more recent Tier 2 method is the Ambient Ratio Method 2 (ARM2), which is based on an evaluation of the ratios of NO₂/NO_X from U.S. EPA’s Air Quality System (AQS) record of air quality data. ARM2 was developed by categorizing all the AQS data into bins of 10 ppb increments for NO_X values less than 200 ppb and into bins of 20 ppb for NO_X in the range of 200-600 ppb. The ARM2 equation is then used to compute NO₂/NO_X ratios based on the total NO_X levels. Since ARM2 is currently classified as a non-default beta option in AERMOD, it is currently treated like a Tier 3 option that requires U.S. EPA regional approval. However, in addition to Tier 3 revisions described above, the Appendix W revisions propose to replace ARM with ARM2 as the Tier 2 option. In any event the following four (4) criteria must be met to justify that ARM2 is appropriate for a particulate project:

1. The Tier 1, total conversion results should be around 150-200 ppb. Total conversion results closer to 200 ppb may be justified if the project site is located in an area of lower background ozone concentrations.
2. If Tier 1 total conversion results are greater than 200 ppb, ARM2 may be appropriate if the project site is located in an area with generally low background NO₂ levels (less than about 20-30 ppb).
3. If Tier 1 total conversion results in greater than 200 ppb, ARM2 may be appropriate if it has been demonstrated that the primary sources (i.e., greater than 95% of modeled NO_X emissions) have an NO₂/NO_X in-stack ratio (ISR) less than 0.2. U.S. EPA preference is for actual testing results to identify sources ISRs.
4. Background ozone concentrations should generally be below 80-90 ppb, with no more than seven (7) hourly concentrations in a year greater than 80-90 ppb. ARM2 is a less conservative option that requires a better understanding of actual ISRs.

The Tier 3 options (OLM and PVMRM) incorporate the atmospheric chemical reaction referred to as titration to convert NO to NO₂ in the presence of ozone. The Tier 3 options also currently require U.S. EPA regional approval (until Appendix W revisions are finalized) for use and source-specific ISR stack test data and representative hourly ozone data for the period corresponding to the meteorological dataset being used for the air quality modeling evaluation. Most ozone ambient monitoring programs only collect data during the ozone season (typically April 1^st through September 30^th). Therefore, in most cases, supplementing ozone data from more distant year-round ozone ambient monitoring stations is required. In cases where representative year-round ozone ambient monitoring stations aren’t available another approach is to extrapolate winter conditions from the shoulder seasons from ozone monitors only operated during the ozone season.

In most cases, Tier 3 represents the least conservative option for 1-hour NO₂ NAAQS air quality modeling. However, because each site has a unique set of specific inputs; it is always recommended to evaluate available methods to determine the most appropriate option for your particulate site. Another recommendation that was included in U.S. EPA’s September 30, 2014 memorandum when using a Tier 3 option is to use a default ISR of 0.2 for any local sources included in a NAAQS air quality modeling evaluation that is greater than four (4) kilometers from the applicant site. This updated default ISR for local sources has made the difference between showing compliance with the 1-hour NO₂ NAAQS and not. Much like ARM2, there is the possibility that the Tier 3 options will be upgraded to default status in the final Appendix W revisions.

In addition to less conservative options for accounting for NO to NO₂ conversion, another important component to a 1-hour NO₂ NAAQS air quality modeling evaluation is the incorporation of representative background NO₂ concentrations. There are currently two (2) options outlined in U.S. EPA guidance memoranda for incorporating 1-hour NO₂ background concentrations. The Tier 1 option is to add the monitoring design value concentration [three (3) year average of the 98^th percentile of daily maximum 1-hour concentrations] to the modeled design concentration. This approach is very conservative and in most cases the Tier 2 option for incorporating background concentrations is required. The Tier 2 approach allows for the incorporation of seasonal and diurnal fluctuations. Specifically, the 3rd highest monitored NO₂ concentration for each hour (1-24) from each day over one (1) season from the last three (3) available years is calculated and the appropriate value is added to the modeled concentration. The most recent version of AERMOD has the ability to incorporate this approach or alternative approaches for varying background monitored concentrations to be added to modeled concentrations.

The last thing to consider when evaluating the 1-hour NO₂ NAAQS is the operating schedule of the source being modeled. This is especially true when evaluating start-up conditions. This approach should be used with caution because there is always the possibility that whatever operating schedule that is incorporated into the model will end up as an operating limit in your permit. However, close attention should be paid to limits already accepted as part of a permit either through emission restrictions or operating hours. In most cases, an operating schedule can be developed for air quality modeling purposes that is more conservative than existing limits but less conservative than assuming 8,760 hours of operation. Since this approach falls into an area with little guidance from U.S. EPA, it is recommended that you work closely with your State agency before proceeding very far down this path.

It should be noted that less conservative approaches for incorporating background concentrations and operating schedules as outlined above can also be utilized for the other 1-hour NAAQS, sulfur dioxide (SO₂), as well as for the fine particulate (PM_2.5) 24-hour and annual NAAQS. However, PM_2.5 comes with a host of other air quality modeling issues that I will save for another 4 The Record.

Since the incorporation of more stringent 1-hour NAAQS (NO₂ and SO₂), it has become necessary to incorporate less conservative modeling options when addressing Prevention of Significant Deterioration (PSD) requirements. Since the most recent version of Appendix W predates the 1-hour NAAQS (Appendix W was last amended in November 2005), U.S. EPA has had to play catch up by releasing guidance memoranda outlining these approaches. Now that proposed revisions have been published with favorable updates to NO₂ modeling techniques, I strongly encourage the regulated community to take this opportunity to provide comments in support of the update of ARM2 as a Tier 2 option and Tier 3 options gaining regulatory default status.

For more information on 1-hour NO₂ NAAQS air quality modeling please Dan Dix at ddix@all4inc.com or 610.933.5246 x118.