Screening Modeling

Screening modeling encompasses a number of conservative analytical modeling techniques for estimating extreme upper bound concentrations. These "worst-case" estimates are based on a standard set of meteorological conditions and simplified assumptions/representations of source-receptor geometry. The primary purpose of screening modeling is to quickly and easily eliminate sources, whose impacts are low enough that they will not pose a threat to ambient air quality standards or ambient increment standards, from further analysis. Another purpose of screening modeling is to identify the "worst-case" operating scenario or load case for a particular source(s) that require a more refined modeling analysis to demonstrate compliance with ambient air quality standards and ambient increment standards.

Screening modeling involves a basic look at the emissions from the proposed operating scenarios using appropriate models with standardized "worst-case" meteorological data. Screening modeling tends to be easy-to-run, quick and conservative (i.e., tends to overpredict air contaminant concentrations).

Screening modeling may be used by itself to demonstrate compliance with ambient air quality standards or to indicate the proposed operating scenario most likely to produce the maximum predicted air contaminant concentrations to be further analyzed via refined modeling.

Screening Models

The Guideline on Air Quality Models provides a complete list of available air quality screening models and modeling techniques. Some basic guidance on a few of the more commonly used models can be found later in this page, however, the "Guideline on Air Quality Models" should be consulted for further details on the models suggested in this guideline and for those modeling situations not specifically covered in this guidance.

SCREEN 3

SCREEN3 is the current regulatory screening model for air permitting applications. The SCREEN3 model is based on the same steady-state Gaussian plume algorithms as ISC3. SCREEN3 is applicable for estimating ambient impacts from point, area, and volume sources, in addition to flares, out to a distance of 50 kilometers and has the ability to handle building downwash and complex terrain.

AERSCREEN

AERSCREEN is a single-source screening version of AERMOD that will produce conservative impact estimates without the need for refined meteorological or detailed terrain data.

At this time, AERSCREEN is currently being tested by various entities and is not a EPA-approved guideline model, therefore, SCREEN3 can be used as the screening model for AERMOD. AERSCREEN guidance will be available here once the model is released and promulgated.

All EPA-approved screening models can be downloaded at the SCRAM website.

Screening Modeling Receptors

Receptors for screening modeling should be selected so as to provide detailed horizontal and vertical resolution of the terrain surrounding the source being modeled. For screening purposes, receptors typically are arrayed along a single axis or radial, and a wind direction selected so that the emissions from the source(s) will be directed towards the receptors. Each receptor is specified by a distance from the source and its elevation (as used in the current version of the SCREEN model with worst-case meteorological conditions). The following are two acceptable methods for selecting screening receptors:

Simple Terrain Receptors

Method 1

1. Select receptors, based on distance from the source, by plotting circles of radii equal to 100 meter intervals out to 2 kilometers from the source; 500 meter intervals for distances out to 4 kilometers; and 1 kilometer intervals out to 10 kilometers.

2. Assign a "worst-case" terrain height to each radius by identifying the highest elevation within the band formed by circles of radii midway between the 2 adjoining receptor circle radii.

Example: To find the correct elevation for the 100 meter receptor in FIGURE 1 below, draw circles of radii of 50 meters and 150 meters. Then look inside that ring and choose the highest elevation within that band and assign it to the 100m receptor, in this case, 480 feet AMSL. Continue on with this placement methodology for the receptor spacing outlined in step 1.

Figure 1

If screening modeling results show that concentrations are still increasing at the 10 kilometer receptor, then the grid must be extended beyond 10 kilometers. Stack top height terrain elevations may be used beyond ten 10 kilometers when running the current versions of the SCREEN dispersion model. Receptors in areas not meeting the definition of "ambient air" (e.g., receptors within the "production area" of a source, see Chapter 116 of the MEDEP-BAQ Regulations) may be eliminated.

Method 2

1. Select all receptors, on the basis of distance from the source, as described in method 1.

2. Select a second set of receptors on the basis of elevation. Find the closest terrain elevation 10 feet above stack (source) base elevation and tabulate a receptor at that distance and elevation. Continue to find the closest terrain elevation to the stack at successively higher 20-foot intervals above stack base and tabulate. Stop at 10 kilometers from the source.

3. Combine these 2 sets of receptor heights and distances to form a single set of screening modeling receptors.

Example: Suppose that the stack in the following example (FIGURE 2) is 60 feet tall and has a base elevation of 430 feet AMSL (stack top height of 490 feet). The first receptor placement will be at the closest occurrence of terrain 10 feet above the stack-base height (in this case, 440 feet). Additional receptors will be placed at 460, 480, 500, 520, 540, 560, 580, 600 feet and so on until you reach 10 kilometers from the source. Combine with the set of receptors derived using method 1.

Figure 2

If screening modeling results show that concentrations are still increasing at the 10 kilometer receptor, then the grid must be extended beyond 10 kilometers. Stack top height terrain elevations may be used beyond 10 kilometers. Receptors in areas not meeting the definition of "ambient air" (e.g., receptors within the "production area" of a source, see Chapter 116 of the MEDEP-BAQ Regulations) should be eliminated.

It is important to note that screening receptors are to be selected to minimally 10 kilometers. This geographical extent is important to demonstrate that the maximum impact area is included in the analysis and also to determine the significant impact area (which will be used to determine the geographical extent for a refined modeling analysis, if applicable). MEDEP-BAQ believes a complete screening receptor grid provides valuable information relative to the analysis.

Complex Terrain Receptors
The selection of complex terrain receptors for use in screening modeling is based on two key elevations: stack-top elevation and plume centerline elevation less ten (10) meters.

Method

1. Calculate the stack-top elevation above mean sea level (AMSL) for each stack and round to the nearest 10 feet.

2. Calculate the plume centerline (AMSL) less 10 meters using the current version of the SCREEN model and round to the nearest 10 feet.

3. Assign a receptor to the closest occurrence of terrain at the calculated stack-top elevation. Continue placing receptors at ten foot intervals until the calculated plume centerline (less ten meters) value is reached.

Complex terrain receptors are to be selected using the methodology described above for each stack being modeled.

Additional or alternative receptors may be required by MEDEP-BAQ depending on the complexity of the terrain and source relationships.

Additional Receptors

There are also certain cases where additional special purpose receptors may have to be selected for inclusion in the analysis. Some examples of special purpose receptors are:

• Fenceline receptors at the closest occurrence of "ambient air", as defined by Chapter 116;

• Receptors within the wake region of the controlling building (outside of the production area, if any);

• Receptors at state/international borders;

• Receptors in Class I areas; and

• Receptors at any designated sensitive locations (typically requested by MEDEP-BAQ).

Building-Related Considerations

For screening analyses where the stack height is less than its formula GEP height, the height and the maximum projected width of the overall controlling structure (as determined through the use of any BPIP program) is to be used. As stated previously, hand calculations will NOT be accepted for making this determination!

Cavity Analysis

The methodology for performing a cavity analysis depends on which model is used to demonstrate compliance:

• Models that uses either the Schulman/Scire or Huber/Snyder algorithm (SCREEN, etc.):

A cavity analysis is to be conducted for all stacks (whose cavity effects may possibly extend into ambient air as defined by Chapter 116) that are below 60% of full GEP height using any EPA approved model.

• Models that incorporate the PRIME downwash algorithm (AERMOD-PRIME, etc.)

A cavity analysis is to be conducted for all stacks (whose cavity effects may possibly extend into ambient air as defined by Chapter 116) that are below 100% of full GEP height using any EPA approved model.

Regardless of the model used, if the cavity area does not or is not likely to extend into ambient air (as defined by Chapter 116), the cavity analysis may be omitted.

All cavity methodologies and techniques are to be included in the protocol and (if applicable) discussed with MEDEP-BAQ prior to performing the analysis.

Meteorology

The set of 54 meteorological conditions given in the following table is required for general screening purposes when using the current version of SCREEN. The table encompasses a full range of atmospheric dispersion conditions which might be encountered. These meteorological conditions are built into the current version of the SCREEN model.

STANDARD SCREENING METEOROLOGICAL CONDITIONS

Wind Speed (m/s)

Stability Class

1.5

2.5

3.5

4.5

•

Additional Notes:

1. Ambient temperature of 293 Kelvin must be used.
2. Mixing heights are automatically calculated in the current version of the SCREEN model.

Since there are many screening models other than those discussed here (each with it's own unique set of meteorological data), it is important to consult with MEDEP-BAQ to help determine if you have the correct data necessary to run the proposed model!

General Screening Modeling Guidance

SCREEN

The preferred screening method for facilities with a single emission point is the current version of SCREEN, which assesses impacts in simple and complex terrain, and handles building downwash, cavity effects, inversion breakup fumigations and coastal fumigations. Guidance can be found in Screening Procedures for Estimating the Air Quality Impact of Stationary Sources and the current version of the SCREEN Users Guide.

COMPLEX I - VALLEY MODE (CI-VM)

For facilities with single or multiple emission points, the current version of the COMPLEX I model in VALLEY MODE [IOPT(27) = 1] or other model approved by EPA or MEDEP-BAQ may be used for complex terrain screening.

For CI-VM screening, the single assumed "worst-case" meteorological condition of Class F stability and wind speed of 2.5 m/s at stack top is used. These conditions are assumed to occur with the same wind direction for 6 hours. CI-VM will produce a 24-hour prediction. A one-hour prediction can be obtained by multiplying the 24-hour prediction by 4. The CI-VM screening 24-hour prediction should be used directly for that averaging period. Other averaging periods can then be derived by using the conversion factors in the following table:

Multiply 1-hour result by:	To get the:
0.9	3-hour concentration
0.7	8-hour concentration
0.08	Annual concentration

The conversion factors listed above were taken from Section 4.2 (page 4-15) of Screening Procedures for Estimating the Air Quality Impact of Stationary Sources (EPA 1992c). There exists a certain degree of flexibility to which the user may diverge from the conversion factors (see Section 4.2), however, for general use in screening modeling, the factors listed above are likely to be appropriate.

When using CI-VM, use the following options:

a. Gradual plume rise should be used to estimate concentrations at nearby elevated receptors if plume impaction is likely on any elevated terrain closer to the source than the distance from the source to the final plume rise, otherwise use final plume rise;

b. Stack tip downwash;

c. Set buoyancy induced dispersion [IOPT(4) = 1];

d. Ambient temperature of 293 Kelvin;

e. Mixing height of 9999 meters;

f. Wind profile exponents should be set to 0.0, respectively, for all 6 stability classes only when using the "CI-VM equivalent" option [IOPT(27) = 1] otherwise for all other regulatory uses of COMPLEX I [IOPT(25) = 1], set the wind profile exponents to 0.07, 0.07, 0.10, 0.15, 0.35 and 0.55, respectively for rural modeling;

g. Set ZMIN = 10;

h. Set terrain adjustment [IOPT(1) = 1]; and

i. Set the terrain adjustment values to 0.5, 0.5, 0.5, 0.5, 0.0, 0.0, respectively for the 6 stability classes.

For applications of CI-VM, 36 ten-degree wind sectors are traditionally used to capture the maximum predicted impacts. Wind sectors with smaller angles (less than 10 degrees) are not typically needed, however, the applicant may choose to refine the number of sectors should it become necessary.

Although not a requirement, a useful feature of the CI-VM model is to select the use of the "High 5 Table" (set IOPT(19) to "Include Average Concentration and High-Five"). This will be helpful to both the consultant, as well as MEDEP-BAQ, to quickly identify the maximum impact predicted by the model for ALL sources. The High-Five Table will NOT predict the maximum impact for any specific source(s) in a multi-source analysis.

Intermediate Terrain Screening

Current regulations requires that concentrations from both the simple terrain model and the complex terrain model be evaluated in intermediate terrain (the area between stack-top height and plume centerline height) and the higher of the two predicted concentrations used at a receptor point after comparison.

For screening modeling, a case-by-case analysis is required where judgments can be made whether the controlling concentrations are associated with the simple terrain or the complex terrain model estimates.

Averaging Periods

Ambient air quality standards, ambient increments, and significance levels cover a variety of time (or averaging) periods depending on the air pollutant in question.

The SCREEN simple terrain and cavity algorithms and ISC model produces 1-hour concentration values which will need to be converted to other averaging periods using the conversion factors listed in the table below:

Multiply 1-hour result by:

To get the:

0.9

3-hour concentration

0.7

8-hour concentration

0.4

24-hour concentration

0.08

Annual concentration

The CI-VM model produces 24-hour concentration values which will also need to be converted to other averaging periods. To do this, first convert the 24-hour prediction to a 1-hour value by multiplying by 4, then use the factors in the above table to derive 3-hour, 8-hour, and annual concentrations. The CI-VM screening 24-hour prediction should be used directly for that averaging period.

The conversion factors listed above were taken from Section 4.2 (page 4-15) of Screening Procedures for Estimating the Air Quality Impact of Stationary Sources (EPA 1992c). There exists a certain degree of flexibility to which the user may diverge from the conversion factors (see Section 4.2), however, for general use in screening modeling, the factors listed above are likely to be appropriate.

Significance Determination

If predicted concentrations from screening modeling are below their respective significant impact levels (as defined in MEDEP-BAQ Regulations, Chapters 100 and 110) for a given regulated pollutant, then no additional modeling is necessary for that pollutant.

Significant Impact Area

The area which is significantly impacted by the source, based on screening modeling, shall be defined as a circular area surrounding the source with radius equal to the maximum distance to which screening no longer predicts concentrations that exceed significance levels. This area, which is significantly impacted by the source, may be redefined using the results of sequential modeling for the source only (see Refined Modeling section for methodologies).

Interactive Sources

If screening modeling results indicate significance levels being exceeded by the source being modeled, then other nearby sources may need to be included in the analysis. If applicable, MEDEP-BAQ will provide the applicant with a list of any sources that may have to be included in the multi-source modeling analysis and the model input data for these sources. The list will contain all data required for model input including source location(s), emission rates, stack parameters, and model-ready BPIP input files or necessary building dimensions for the applicant to determine direction-specific building parameters.

Although the results tend to be very conservative, it is possible for a combined source analysis to demonstrate compliance using only screening modeling. If the total predicted maximum concentrations of all sources, modeled separately, when added to background does not exceed ambient air quality standards or if the total predicted maximum ambient increment concentration does not exceed ambient increment limits, then compliance with the ambient air quality standards and ambient increment standards shall have been demonstrated and no further modeling will be required. This approach applies to simple, complex and intermediate terrain modeling.

If single and/or combined source modeling results cannot demonstrate compliance with MAAQS or increment standards, refined modeling will be necessary.

Further guidance on modeling contributions from other nearby sources using a screening model can be found in Screening Procedures for Estimating the Air Quality Impact of Stationary Sources.

meteorology