AERMOD Tiering Approach Case Study for 1-Hour NO2

Environmental solutions delivered uncommonly well
AERMOD Tiering Approach Case Study
for 1-Hour NO2
Paper No 481
Prepared By:
Anthony J Schroeder, CCM – Managing Consultant
TRINITY CONSULTANTS
201 N. Illinois Street
16th Floor, South Tower
Indianapolis, IN 46204
(317) 451-8100
trinityconsultants.com
June 19, 2012

2
ABSTRACT
This study reviews 1-hour NO2 concentrations predicted by AERMOD for a hypothetical
source at four locations throughout the United States with hourly varying background
ozone concentrations. The sensitivity of the model-predicted concentrations to the tier
used (i.e., Tier 1 versus Tier 3) is presented based on distance from the source and
evaluated on a 1-hour NO2 NAAQS design concentration basis. All scenarios use hourly
varying ambient ozone concentrations corresponding to the same time period as the
meteorological data input to AERMOD. Also presented are results showing the sensitivity
of AERMOD outputs to those inputs needed for the Tier 3 approach (e.g., in stack ratio of
NO2/NOX, ambient equilibrium ratio, and ambient ozone concentrations), to the relative
magnitude of the emission rate modeled, and to the relative ambient ozone concentration.
INTRODUCTION
AERMOD is the U.S. EPA’s recommended model for evaluating near-field impacts, defined
as occurring within 50 km of the sources, caused by pollutant emission sources. This
recommendation has been reiterated by U.S. EPA in two recent guidance documents issued
with respect to the recently promulgated 1-hour NO2 NAAQS.1,2 In its March 1, 2011
clarification memo, EPA has also reiterated that a three tiered approach to predictions of 1-
hour average NO2 concentrations using AERMOD is appropriate for use in regulatory
applications. Many AERMOD users have traditionally used only Tier 1, where the entire NO
component of emitted NOX is assumed to immediately react to form NO2, when completing
regulatory assessments to demonstrate compliance with the annual NO2 NAAQS. The most
complex formulations associated with the Tier 3 approach have not been as widely used,
however. That has changed with the promulgation of the stringent 1-hour NO2 NAAQS, for
which the more complex Tier 3 formulations in AERMOD are often necessary to
demonstrate compliance through dispersion modeling.
The three tiered approach to NO2 modeling with AERMOD consists of three different
methods to estimate the quantity of NO in the NOX emitted from sources in the model that
is converted to NO2 at a particular receptor. NO2 is the pollutant actually assessed against
the 1-hour NAAQS so used of methods to estimate the NO to NO2 conversion that are not
overly conservative is critical. Each method is less conservative (i.e., generally yields lower
design concentrations) than the prior method, with Tier 1 generally yielding the highest
modeled concentrations and Tier 3 the lowest. In Tier 1, the assumption is made that all
NOX emitted from a source is fully converted to NO2 instantaneously. In the Tier 2
methodology, also called the Ambient Ratio Method (ARM), the assumption is made that
80% of the NOX emitted from a source is converted to NO2 instantaneously upon entering
the atmosphere when applied for 1-hour averaging periods. This ambient ratio is based on
two studies cited by EPA in its March 1, 2011 clarification memo. Tier 3 actually consists of
two methods that are similar in concept: the Ozone Limited Method (OLM) and the Plume
Volume Molar Ratio Method (PVMRM).3,4,5 In both methods the chemical mechanism of
ozone titration, in which NO interacts with ambient ozone to form NO2 and oxygen, is taken
into consideration. Algorithms implementing both Tier 3 methods are part of the standard

3
AERMOD code, although neither method is considered a default regulatory option
according to the Guideline of Air Quality Models.6
Sensitivity and bias testing of PVMRM and OLM was conducted in 2004 and 2005 on behalf
of the Alaska Department of Environmental Conservation.7,8 The sensitivity studies
focused on three locations in Alaska. The bias studies considered additional locations (e.g.,
Hawaii, New Mexico) and were primarily intended to test whether the PVMRM and OLM
algorithms within AERMOD yielded unbiased results compared with field data. Additional
limited bias testing and review of the PVMRM and OLM algorithms in AERMOD was also
conducted more recently in 2011.2,9 Those studies were also focused primarily on
determining if the predictions were unbiased or to determine if there were any issues with
the implementation of PVMRM or OLM in AERMOD.
The purpose of this paper is to conduct additional sensitivity testing of the PVMRM and
OLM algorithms in AERMOD using meteorological and ozone data from locations in
different climatic and urban/rural locations throughout the United States. EPA has
suggested that in the absence of source specific data an in-stack NO2/NOX ratio of 0.5 be
used, which is considerably higher than the default ratios (generally 0.1) used in many of
the test cases cited above. Therefore, the sensitivity of PVMRM and OLM results to this
input is evaluated. Additionally, previous sensitivity studies have shown that the NO2/NOX
ratio predicted using Tier 3 methodologies can be affected by the relative magnitude of NOX
emissions input to the model; thus the sensitivity of model predicted NO2/NOX ratios to the
emission rate magnitude is also considered for each location and in-stack ratio scenario
presented.
METHODOLOGY
To explore differences in modeled NO2/NOX ratios for varying PVMRM and OLM inputs,
meteorological and ozone data were gathered for four locations in the United States:
• Urban location in Illinois (IL).
• Rural location in eastern New Mexico/western Texas (NM).
• Urban location in Utah (UT).
• Rural location in Wisconsin (WI).
The following sections provide a description of the inputs used in the case studies
presented herein.
Meteorological and Ozone Data
Five years of concurrent hourly meteorological and ozone data were gathered for each of
the locations except for the NM location, for which only one year of data were available,
and the UT location, for which only four years of data were available. The peak hourly
ozone concentrations observed at each location over the period used in this study as well
as the period average concentrations are shown in Table 1. Additionally, the monthly
distribution of peak and average ozone concentrations are shown in Figures 1 through 4.

4
The highest peak ozone concentrations and highest month-to-month variability in
concentrations occur in the urban locations (IL and UT). The lowest peak concentrations
and lowest month-to-month variability occur at the NM location, which is located in a rural
setting. Average ozone concentrations are more consistent between the four locations with
the lowest overall average occurring in the urban IL location.
Table 1. Peak and Average Ozone Concentrations (ppb)
Ozone Concentration
Location Peak Average
IL 114.0 20.3
NM 59.0 28.7
UT 127.0 34.5
WI 87.0 31.4
Figure 1. Monthly Peak Ozone Concentrations at IL Location.
Figure 2. Monthly Peak Ozone Concentrations at NM Location.
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
1 2 3 4 5 6 7 8 9 10 11 12
Ozone(ppb)
Month
Average
Peak
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
1 2 3 4 5 6 7 8 9 10 11 12
Ozone(ppb)
Month
Average
Peak

5
Figure 3. Monthly Peak Ozone Concentrations at UT Location.
Figure 4. Monthly Peak Ozone Concentrations at WI Location.
Emission Source
A hypothetical emission source was defined for use in all AERMOD analyses. This source,
defined as a point source, was assigned the following release parameters:
• Stack height: 10 m
• Exhaust temperature: 366.5 K
• Exit velocity: 6 m/s
• Stack diameter: 1 m
An emission rate of 1 g/s, which is equivalent to approximately 7.9 lb/hr or 34.8 tpy, was
used for the emission source in a low emission rate set of model scenarios and an emission
rate of 100 g/s, which is approximately 794 lb/hr or 3,476 tpy, was used in a high emission
rate set of model scenarios.
Receptor Grid
A radial receptor grid, centered on the hypothetical emission source, was used in this
analysis. The radial grid allowed all receptors to be located at a fixed interval from the
emission source so that the ratio of NO2/NOX concentrations with distance could be plotted.
Receptors were placed at 100 meter spacing in each of 36 radial directions out to a distance
of 10 kilometers. So that the results shown in this study could focus primarily on
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
1 2 3 4 5 6 7 8 9 10 11 12
Ozone(ppb)
Month
Average
Peak
0.0
20.0
40.0
60.0
80.0
100.0
120.0
140.0
1 2 3 4 5 6 7 8 9 10 11 12
Ozone(ppb)
Month
Average
Peak

6
differences in the NO2 to NOX conversion methodology used and not possible influences of
terrain, flat terrain was assumed in all model analyses.
Tiering Options
For each of the four locations and both emission rate scenarios considered in this study,
several sub-scenarios are considered. First an AERMOD run was completed using Tier 1,
where full conversion of NOX to NO2 is considered. Then the Tier 1 results were multiplied
by 0.8 to implement the Tier 2 methodology or ARM. Next, a series of analyses were
completed with different input options using the PVMRM and OLM Tier 3 methods. The
“default” cases used an in-stack ratio of 0.5 and ambient equilibrium ratio of 0.9, consistent
with EPA’s March 1, 2011 guidance memo (Option A). Then the in-stack ratio was
decreased to 0.1, consistent with many prior sensitivity studies, with an ambient
equilibrium ratio of 0.9 (Option B). Finally, the ambient equilibrium ratio was reduced to
0.8 to match the Tier 2 ambient ratio in two additional scenarios, one with an in-stack ratio
of 0.1 (Option C) and one with an in-stack ratio of 0.5 (Option D).
To investigate the sensitivity of the ratio of NO2 to NOX concentrations over distance to the
relative magnitude of the ambient ozone concentrations, a final set of experimental model
runs were completed. These runs were all completed using the IL meteorological dataset
with three different levels of constant ambient ozone concentrations. The assumed
ambient ozone concentrations ranged from 75 ppb (the “high” scenario), to 40 ppb (the
“medium” scenario), to 10 ppb (the “low” scenario). All of the Tier 1, Tier 2, and Tier 3
scenarios described above were run for each of the three constant ozone background
concentrations.
RESULTS
In the following sections, results from dispersion modeling analyses using the different tier
options available in accordance with EPA guidance are compared. The results of the case
studies involving the four locations chosen for evaluation in this paper are presented in the
first section. In the second section, the results of an additional case study using the IL
location and constant ambient ozone concentrations are presented. Discussion regarding
the observed trends is also provided in each section.
Location Based Case Studies
The 1-hour NO2 design concentrations (i.e., the five year average of the 98th percentile of
the annual distribution of daily maximum concentrations) from each location and emission
rate are shown in Figures 5 through 8.

7
Figure 5. IL Design Concentrations for 1 g/s Scenario (left) and 100 g/s Scenario
(right).
Figure 6. NM Design Concentrations for 1 g/s Scenario (left) and 100 g/s Scenario
(right).
Figure 7. UT Design Concentrations for 1 g/s Scenario (left) and 100 g/s Scenario
(right).
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
1-hourNO2Concentration(µg/m3)
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0
0.0
20.0
40.0
60.0
80.0
100.0
120.0
0.0
2000.0
4000.0
6000.0
8000.0
10000.0
12000.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0
8000.0
9000.0

8
Figure 8. WI Design Concentrations for 1 g/s Scenario (left) and 100 g/s Scenario
(right).
The following observations can be made from the data presented in Figures 5 through 8.
• In general, use of the Tier 1 methodology results in the highest design
concentrations for all scenarios, with the Tier 2 methodology resulting in the next
highest design concentrations, and Tier 3 methodologies resulting in the lowest
design concentrations. The exception to this rule is that the use of OLM with an in-
stack NO2/NOX ratio of 0.5 (Options A and D) and lower emission rate results in
higher design concentrations than the use of Tier 2 for all locations. Additionally,
the use of PVMRM with the default options (in-stack ratio of 0.5 and ambient
equilibrium ratio of 0.9) and lower emission rate results in higher design
concentrations than the Tier 2 methodology.
• The magnitude of the AERMOD-predicted design concentration is similar for
Options A and D and Options B and C for particular Tier 3 methodology. Options A
and D both use an in-stack NO2/NOX ratio of 0.5 and Options B and C both use an in-
stack NO2/NOX ratio of 0.1 with the ambient equilibrium ratio equal to 0.9 for
Options A and B and 0.8 for Options C and D. Therefore, the magnitude of the
AERMOD-predicted design concentration is more sensitive to variations in in-stack
NO2/NOX ratio than the ambient equilibrium ratio.
• The exception to the previous observation occurs in the UT location lower emission
rate scenario where use of a lower in-stack ratio appears to have little impact on
design concentrations, particularly for the OLM scenarios. The ambient ozone
monitor used in the UT case study only operated during the summer ozone season
for the first two years of the four years of available data. In those first two years,
constant ozone monthly values ranging from 42 to 57 ppb are used for winter
monthly resulting in a month average ozone concentration that are considerably
higher in the substituted periods than for subsequent years when hourly data are
available. For example, the monthly average March ozone concentration is 57 ppb
in 2004 but only 30.8 ppb in 2005 and 30.6 in 2006. Overly conservative estimates
of ambient ozone concentrations during data void periods can lead to high estimates
of NOX to NO2 conversion ratios using Tier 3 methods in AERMOD.
• The use of Tier 3 methodologies results in much lower concentrations compared
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
90.0
100.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0
8000.0
9000.0
10000.0

9
with Tier 1 and Tier 2 in the higher (100 g/s) emission rate cases than in the lower
(1 g/s) emission rate cases as seen when the bar charts on the right are compared
with those on the left.
• Differences in design concentrations between Tiers 1, 2, and 3 do not appear to be
affected by whether the case is in a rural or urban location in the case studies
included here.
• In the higher emission rate scenarios, OLM and PVMRM result in similar, although
not identical, design concentrations as seen in the bar charts on the right.
The relationship of the NO2/NOX ratio with distance is shown for each of the four locations
and emission rate scenarios in Figures 9 through 12. The values shown on the vertical axis
in these figures are the maximum ratio of 1-hour NO2 design concentration (Tier 3) to 1-
hour NOX design concentration (Tier 1) for receptors located a particular distance from the
emission source. These figures do not show Options C and D, which use an ambient
equilibrium ratio of 0.8, because the variation in NO2/NOX ratios with distance is generally
similar to that shown for Options A and B. The primary difference between use of the
ambient equilibrium ratios of 0.8 and 0.9 is that for the 1 g/s emission rate scenarios, the
ratios converge to either 0.8 or 0.9 at locations distant from the source. Of note in this
series of figures is the following:
• For the higher emission rate scenarios, the NO2/NOX ratio equals the in-stack ratio
at the receptors nearest the source, as may be expected. However, for the lower
emission rate scenarios, the observed ratios are never as low as the in-stack ratios,
indicating that the NOX in the plume is assumed to have already reacted ambient
ozone to a decree that the NO2/NOX ratio has increased from the in-stack ratio by
the time the plume reaches the nearest receptors to the source.
• For the higher emission rate scenarios, OLM and PVMRM predict the same ratio of
NO2/NOX concentrations at receptors near the source. At receptors more distant
from the source, OLM predicts lower ratios than PVMRM.
• For the lower emission rate scenarios, the NO2/NOX ratios predicted by PVMRM are
lower than OLM for all source-receptor distances for a particular in-stack ratio. At
receptors more distant from the source, the NO2/NOX ratio converges to 1.0 for OLM
and 0.9 for PVMRM when an ambient equilibrium ratio of 0.9 is used.
• For the lower emission rate scenarios OLM generally predicts NO2/NOX ratios at or
near 1.0 for all receptor-source distances. This trend highlights one difference in
the OLM and PVMRM formulations used in AERMOD. In OLM the concentration of
NOX at a receptor location is compared with the ambient ozone concentration. If the
ozone concentration exceeds the NOX concentration, all of the NOX is assumed to
convert to NO2. In PVMRM, the number of moles of NOX present in the entire plume
is considered when the determination is made concerning the availability of
ambient ozone for conversion to NO2, not simply the concentration at the receptor
location. For scenarios with high ambient ozone or low NOX emission rates, this can

10
result in considerable differences between OLM and PVMRM predicted NO2/NOX
ratios.
Figure 9. IL NO2/NOX Ratios with Distance for Design Concentration: 1 g/s Scenario
(left) and 100 g/s Scenario (right).
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB

11
Figure 10. NM NO2/NOX Ratios with Distance for Design Concentration: 1 g/s
Scenario (left) and 100 g/s Scenario (right).
Figure 11. UT NO2/NOX Ratios with Distance for Design Concentration: 1 g/s Scenario
Figure 12. WI NO2/NOX Ratios with Distance for Design Concentration: 1 g/s Scenario
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB

12
Constant Ambient Ozone Case Study
The 1-hour NO2 design concentrations (i.e., the five year average of the 98th percentile of
the annual distribution of daily maximum concentrations) for each constant ambient ozone
concentration scenario and emission rate are shown in Figures 13 through 15.
Figure 13. 1-Hour Design Concentrations for High Ozone 1 g/s Scenario (left) and
High Ozone 100 g/s Scenario (right).
Figure 14. 1-Hour Design Concentrations for Medium Ozone 1 g/s Scenario (left) and
Medium Ozone 100 g/s Scenario (right).
Figure 15. 1-Hour Design Concentrations for Low Ozone 1 g/s Scenario (left) and
Low Ozone 100 g/s Scenario (right).
The variations in design concentrations by tier and option shown for the higher emission
rate scenarios are similar to those seen in the location case studies presented in the
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
0.0
1000.0
2000.0
3000.0
4000.0
5000.0
6000.0
7000.0

13
previous section. The variations in design concentrations for the low ambient ozone, lower
emission rate scenario are also similar to the location case studies with higher emission
rates. The lower emission rate scenarios with high and medium ambient ozone
concentrations resemble the distribution of design concentrations seen in the lower
emission rate UT location scenario where little impact on design concentrations was seen
using Tier 3 methods versus Tier 1 and results were not impacted by reducing the in-stack
NO2/NOX ratio input to AERMOD. This constant ambient ozone concentration case
highlights the fact that the largest differences in maximum modeled NO2 design
concentrations between Tier 1 and Tier 3 methodologies occur for high NOX emission rates,
low in-stack NO2/NOX ratios, and low ambient ozone concentrations.
The relationship of the NO2/NOX ratio with distance is shown for the three different
constant ambient ozone scenarios in Figures 16 through 18. These plots show that the
NO2/NOX ratio nearest the emission source is more sensitive to the ambient ozone
concentration for the lower emission rate scenarios than the higher emission rate
scenarios. At higher emission rates there is not enough ozone available for the emitted NOX
to be converted to NO2 prior to the plume reaching the nearest receptor to the source even
at the higher end of typically observed ambient ozone concentrations. Thus, regardless of
ambient ozone concentration, the NO2/NOX ratio is at or near the in-stack ratio for both the
OLM and PVMRM scenarios within approximately 1,000 meters of the source. As was seen
with the four location case studies, the use of PVMRM results in lower predicted NO2/NOX
ratios than OLM at all distances in the lower emission rate scenarios and within
approximately 1,000 meters of the source in the higher emission rate scenarios.
Figure 16. NO2/NOX Ratios with Distance for Design Concentration: High Ozone 1 g/s
Scenario (left) and High Ozone 100 g/s Scenario (right).
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB

14
Figure 17. NO2/NOX Ratios with Distance for Design Concentration: Medium Ozone 1
g/s Scenario (left) and Medium Ozone 100 g/s Scenario (right).
Figure 18. NO2/NOX Ratios with Distance for Design Concentration: Low Ozone 1 g/s
Scenario (left) and Low Ozone 100 g/s Scenario (right).
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
100 1000 10000
NO2/NOXRatio
Distance (m)
OLMA
OLMB
PVMRMA
PVMRMB

15
CONCLUSIONS
The case studies presented here highlight important sensitivities to be considered when
implementing the three-tiered NOX to NO2 conversion methodology to predict 1-hour NO2
concentrations using AERMOD. As would be expected, the Tier 1 assumption of full
conversion will always result in the highest predictions of NO2 concentrations. However,
there are some situations when use of the Tier 2 methodology will result in lower
predictions of NO2 concentrations than Tier 3 methodologies, particularly when using the
EPA-recommended default assumption of a 0.5 in-stack NO2/NOX ratio. It is also important
to note that the relative emission rate of the sources modeled can affect whether or not the
use of Tier 3 methodologies will result in considerably lower AERMOD-predicted NO2
concentrations compared with the full conversion assumption. Additionally, for a
particular ambient ozone concentration, the Tier 3 methodologies will yield a greater
difference in NO2 and NOX concentrations for higher emission rate sources than for lower
emission rate sources.
The case studies presented here also provide some cautions for AERMOD-users when
implementing Tier 3 methodologies for 1-hour NO2 modeling. Careful consideration should
be made to the in-stack NO2/NOX ratio used for all sources in an AERMOD analysis.
Maximum AERMOD predicted concentrations for a source frequently occur at receptors
located nearest the source and use of a lower, frequently more realistic, in-stack ratio than
the default of 0.5 can help to reduce model-predicted concentrations at locations of
maximum concentrations. AERMOD users should also be cautious when using ozone data
sets that do not contain complete hourly data. In many situations, nearby monitor data
may only be available for the summer ozone season. In those cases, some sort of
substitution must be completed for data void periods often with conservatively high
estimates of ozone concentrations made. Use of a constant ozone value for missing periods,
as was the case in the UT case study, can unnecessarily mute the NO2 reduction capabilities
of Tier 3 methods particularly for scenarios with lower emission rate sources.

16
REFERENCES
1. Applicability of Appendix W Modeling Guidance for the 1-hour NO2 National
Ambient Air Quality Standard, Memorandum from Tyler Fox of EPA to Regional Air
Division Directors, June 28, 2010.
2. Additional Clarification Regarding Application of Appendix W Modeling Guidance
for the 1-hour NO2 National Ambient Air Quality Standard, Memorandum from Tyler
Fox of EPA to Regional Air Division Directors, March 1, 2011.
3. Cole, H. and J. Summerhays, 1979, “A review of techniques available for estimating
short-term NO2 concentrations,” J. Air & Waste Manage. Assoc., 29, 812-817.
4. Hanrahan, P.L. 1999a, “The plume volume molar ratio method for determining
NO2/NOX ratios in modeling. Part I: Methodology,” J. Air & Waste Manage. Assoc., 49,
1324-1331.
5. Hanrahan, P.L. 1999b, “The plume volume molar ratio method for determining
NO2/NOX ratios in modeling. Part II: Evaluation Studies,” J. Air & Waste Manage.
Assoc., 49, 1332-1338.
6. U.S. EPA, Office of Air Quality Planning and Standards, Federal Register Vol. 70 / No.
216, pp. 68218-68261, 40 CFR 51, Appendix W, Revision to Guideline on Air Quality
Models, November 9, 2005.
7. MACTEC Federal Programs, Inc., Sensitivity Analysis of PVMRM and OLM in
AERMOD, Research Triangle Park, North Carolina, 2004.
8. MACTEC Federal Programs, Inc., Evaluation of Bias in AERMOD-PVMRM, Research
Triangle Park, North Carolina, 2005.
9. Epsilon Associates Inc., Technical Review of the Ozone Limiting Method (OLM) and
Plume Volume Molar Ratio Method (PVMRM) Codes in the ISC3 and AERMOD
Models, Prepared for the American Petroleum Institute, August 2011.
KEYWORDS
AERMOD, NAAQS, NO2

AERMOD Tiering Approach Case Study for 1-Hour NO2

More Related Content

What's hot (20)

Viewers also liked (15)

Similar to AERMOD Tiering Approach Case Study for 1-Hour NO2 (20)

More from BREEZE Software (20)

Recently uploaded (20)

AERMOD Tiering Approach Case Study for 1-Hour NO2