Northwest International Air Quality Environmental Science and Technology Consortium

NW AIRQUEST

Criteria pollutant design values

Frequently Asked Questions



Q: How are background concentrations determined at areas without monitors?

A: By using monitoring data when available, along with archived air quality modeling data from the AIRPACT3 model. Models may not always correctly simulate air pollutant concentrations, but have good spatial coverage. The monitors are used to adjust the modeled value, and model data are used in a relative sense to account for the spatial variability of pollutant levels.


[top]


Q: Can these data be used for determining compliance with Federal standards?

A: No. Though monitoring and modeling data used in this application are processed according to rules governing the calculation of design values, these data may differ from the official numbers due to various reasons.


[top]


Q: Which regulatory agencies have signed off on this tool?

A: At present this is mostly a minor NSR permitting tool that Idaho, Oregon and Washington use. Always speak with your permitting authority before using these data in a permit application. There may be times when it is advisable to use a modified version of the data, use the O3 background for PVMRM, or the PM10 values with no extremes, etc.


[top]


Q: Are these background concentrations used for quick screening only?

A: These background concentrations could be used in support of most minor source permit applications in Washington, Oregon and Idaho, after speaking with your permitting authority. Modeling may be conducted with a screening model like AERSCREEN, or a refined model like AERMOD.


[top]


Q: AIRPACT already includes most existing sources near the new source I want to model, so I don't need to include their contributions in a model run (like AERMOD), right?

A: Though the AIRPACT3 model includes emissions from major point sources, it does not resolve the plumes from near source impacts. Point source emissions are significantly dispersed over the 144km2 grid cell in which the emissions originate. As a result, modeled concentrations from these point sources will have low concentration gradients within the area of interest. In addition monitoring data from source compliance monitors and other monitors located near large stationary sources (SO2 from Trail, BC, and Pocatello, ID), were omitted from the interpolation because of the potential significant contribution to ambient concentrations from these sources. As such the effect of double counting nearby sources will be small, and there is little justification assigning negative emission rates to such sources in AERMOD or AERSCREEN.


[top]


Q: Where can I see maps of design values of different pollutants?

A: Please click on the relevant hyperlink below:


[top]


Q: How much do these concentrations change from year to year?

A: While some sites do have design values than change substantially each year, most do not. As such it is thought that interpolated design values in unmonitored areas will not have large inter-annual changes. However if your work is critically dependent on the most recent design value, these 2009-2011 design values are probably not appropriate.


[top]


Q: Are exceptional events removed from the monitoring data?

A: No. All valid monitoring data, whether collected with a Federal Reference Method (FRM) sampler or not, are used. Note: Removing "flagged" high monitor values, even if not officially concurred by EPA, may be appropriate to obtain representative background values for a refined analysis. Refined analyses should be discussed with your state permit modeling coordinator. Data for 2009–2011 may be obtained using EPA's AirData website. The data may be downloaded either including or excluding data flagged by state agencies as "exceptional events" even if the flags are not concurred by EPA, however you should discuss specific flagged data with your appropriate state agency personnel.


[top]


Q: What is the "PM10- no extremes" all about?

A: A more reasonable estimate of the PM10 "design value" that is less dependent on a few extreme events.

The PM10 design value is determined by the 4th highest daily mean concentration, in 3 years. However not all PM10 samplers (particularly the FRMs) operate on a daily basis. According to the 1987 SIP guidance, the PM10 design value is dependent on the number of valid samples, as follows:

If the number of valid PM10 samples in 3 years is: Select:
< 347 1st high (i.e. maximum over 3 years)
348 - 695 2nd high
696 - 1042 3rd high
> 1043 4th high

Thus, if a PM10 sampler operating on a 1-in-3 schedule did not report a valid sample on an average of 7 days per year, its design value could be determined by the highest value it recorded over the 3 year period. This is very conservative compared to standards of other pollutants. For example, the PM2.5 24-hr standard and NO2 1-hour standard allow the highest 2% days in each of the 3 years to be excluded. The PM10 standard in contrast allows a maximum of 0.38% of the highest days to be excluded. Often these highs are driven by exceptional events, the demonstrations of which could still be pending.

Most high PM10 concentrations in these three states are caused by windblown dust or medium range transport events. These occur during periods of high winds. These winds also produce good dispersion of pollutants from local sources. As such it would be a gross over-estimate to pair a high PM10 concentration from a local source (likely to occur during the absence of high winds), with a high impact windblown dust event, for instance. The exception however is if the local source generates a lot of PM10 during high wind events.

To address this concern, the PM10 background design value calculation was repeated by omitting from the monitoring data, days that were not part of their lognormal distribution. Please see Figure 1.

Figure 1: Lognormal Quantile-Quantile plots of PM10 for identifying extremes. Numbers in each panel are the six highest days in three years. Green lines run through the 10th and 90th percentiles. IMPROVE site name abbreviations are shown in uppercase.


Please discuss with your permitting authority whether the use of this PM10 product is suitable for your application or not.


[top]


Q: What is "Ozone for PVMRM" all about?

A: It is the month with the highest 75th percentile of daily maximum ozone concentrations. In other words:

  1. Calculate the daily maximum of the running 8hr average ozone level (D8M)
  2. Find the monthly 75th percentile of these D8Ms.
  3. Pick the highest from (2) above.

The ozone design value is driven entirely by the high summer values and does not account for lower values during the rest of the year. However the Plume Volume Mixing Ratio Method (PVMRM) used in AERMOD to account for year-round conversion of NO → NO2 in the presence of ozone, needs to account for the lower ozone concentrations as well. This metric is an attempt to find a single, year-round value that balances both the high and low ozone periods. Several metrics were tested and the monthly maximum 75th percentile D8M was chosen because it provided some site-to-site variability, while remaining somewhat conservative.

Figure 2: 2009-2011 monthly ozone data distribution. This figure shows the relationship among the mean ozone level (red line), the monthly maximum 75th percentile D8Ms (green line), and the design value (blue line).

[top]


Q: What if I have more questions?

A: Please contact Ranil Dhammapala (360-407-6807, ranil.dhammapala@ecy.wa.gov) at the Washington State Department of Ecology for questions regarding this application. For questions regarding the use of background concentrations in preparing an air quality dispersion modeling project for New Source Review, please contact the appropriate state modeling contact person:


[top]

Civil & Environmental Engineering, PO Box 642910, Washington State University, Pullman WA 99164-2910, 509-335-2576