Entrepreneur: Start & Grow Your Business

Comparison of the Industrial Source Complex and AERMOD dispersion models: case study for human health risk assessment.

by Silverman, Keith C.^Tell, Joan G.^Sargent, Edward V.^Qiu, Zeyuan

Journal of the Air & Waste Management Association • Dec, 2007 • TECHNICAL PAPER

ABSTRACT

Air quality models are typically used to predict the fate and transport of air emissions from industrial sources to comply with federal and state regulatory requirements and environmental standards, as well as to determine pollution control requirements. For many years, the U.S. Environmental Protection Agency (EPA) widely used the Industrial Source Complex (ISC) model because of its broad applicability to multiple source types. Recently, EPA adopted a new rule that replaces ISC with AERMOD, a state-of-the-practice air dispersion model, in many air quality impact assessments. This study compared the two models as well as their enhanced versions that incorporate the Plume Rise Model Enhancements (PRIME) algorithm. PRIME takes into account the effects of building downwash on plume dispersion. The comparison used actual point, area, and volume sources located on two separate facilities in conjunction with site-specific terrain and meteorological data. The modeled maximum total period average ground-level air concentrations were used to calculate potential health effects for human receptors. The results show that the switch from ISC to AERMOD and the incorporation of the PRIME algorithm tend to generate lower concentration estimates at the point of maximum ground-level concentration. However, the magnitude of difference varies from insignificant to significant depending on the types of the sources and the site-specific conditions. The differences in human health effects, predicted using results from the two models, mirror the concentrations predicted by the models.

INTRODUCTION

Air dispersion models are designed to predict the fate and transport of emissions of pollutants into the atmosphere. Pollutants once emitted will mix with the ambient air, where physical processes, such as turbulence and chemical reactions, cause the primary pollutants to disperse and their concentration to decrease. In some cases, chemical reactions may cause the primary pollutants to produce secondary pollutants such as ozone. Air dispersion models predict the ambient air concentrations of a compound at specific spatial locations (called receptors) using mathematical equations that represent the numerous and complex meteorological processes responsible for dispersion. Inputs to an air dispersion model typically include meteorological data, source emission data in the form of a mass emission rate, dimensions of nearby structures, and local terrain information. The U.S. Environmental Protection Agency (EPA) and state environmental regulatory agencies have used air dispersion models to implement many regulatory programs.

Generally, EPA regulatory air dispersion modeling is conducted in accordance with the procedures outlined in 40 CFR 51 Appendix W Guideline on Air Quality Models. On November 9, 2005, EPA issued the final rule to replace the widely used Industrial Source Complex (ISC) air dispersion model with a new state-of-practice air dispersion model AERMOD in many air quality impact assessments. In accordance with EPA (2005), AERMOD is fully promulgated as a replacement to ISC. (1) The potential impact of these changes is of interest to regulatory agencies and regulated industries. Air dispersion modeling is used to predict the fate and transport of air emissions from industrial sources to comply with regulatory requirements, environmental and health standards, and facility design criteria. Modeling air concentrations at receptors in a community is a crucial first step in assessing any potential risk to human health or the environment. For example, in assessing human health effects due to inhalation of air toxics, the health outcome is directly related to the air concentration of the chemical predicted by the air dispersion model.

Because different air dispersion models are likely to produce different results under various conditions, it would be interesting to evaluate the nature and magnitude of these differences and their implications on the human health risk assessment of air toxics. Several studies have compared ISC and AERMOD and their PRIME versions. (2-10) The PRIME versions of the models include the Plume Rise Model Enhancements (PRIME) algorithm developed to correct some shortcomings discovered in the building downwash algorithm. Perry et al. (11) compared several existing air dispersion models in terms of modeled and observed concentration distributions and concluded that with few exceptions the performance of AERMOD is superior to that of the other applied models.

This study compared the EPA preferred models (AERMOD and AERMOD-PRIME) to two widely used alternative models (ISC and ISC-PRIME). Moreover, this study assessed the impact of the model changes on the calculation of both carcinogenic and noncarcinogenic inhalation risk to human health. The risk impacts are important because the model results directly influence decisions under the current risk-based air toxics programs of the 1990 Clean Air Act Amendments (CAA). For example, the Maximum Achievable Control Technology (MACT) program was designed to significantly reduce emissions from major sources through pollution-control technologies. Once the control technologies have been implemented, the CAA requires that risk assessments be performed to evaluate any residual human health risk. The results of these risk assessments will determine if major sources will need to implement further controls to reduce pollution, which are usually very expensive to implement. (12) Therefore, changes to the regulatory accepted air dispersion model could have important economic consequences to regulated industries.

In this application, the air dispersion models were tested using a point source and two nonpoint sources (an area and a volume source) that are located on two actual industrial sites. The point source represents a typical stack from a pollution control device. The volume source represents the fugitive emissions associated with a typical process building. The area source represents the fugitive emissions associated with large storage tanks. The modeled maximum ground-level air concentrations were used to evaluate the human health risk from exposure to air toxics using the exposure factors, toxicity factors, and risk equations typically used for the calculation of residual risk.

METHODS

Model Descriptions

The ISC model is a Gaussian dispersion model that assumes any release from a source disperses in a steady-state manner from the time of release until the time it reaches a receptor. Gaussian dispersion models assume that a normal distribution can characterize the horizontal and vertical spread of a plume. (13) On-site structures can affect wind flow and contribute to building downwash, which can have important ramifications in air quality modeling. The building downwash algorithms in ISC are designed to evaluate the extent of building downwash. These algorithms require additional input and therefore, the EPA Building Profile Input Program (BPIP) is run for all point sources (stacks) to generate necessary inputs required for execution of ISC. BPIP determines whether a stack is potentially subject to wake effects due to the surrounding structures and this information is supplied as an input to ISC. (15) In addition, ISC requires input data on source characteristics, receptor location, meteorological parameters, and topography. AERMOD incorporates the same down-wash algorithms as ISC but contains advanced algorithms for dispersion, plume rise, buoyancy, and the handling of complex terrain. AERMOD, like ISC, is a steady-state model and is most useful for analyzing short-range pollutant transport within 20 km of the source. (16) The main justification for replacing ISC with AERMOD was that AERMOD incorporates many of the scientific advances made in the 1970s and 1980s in understanding turbulence and dispersion in the planetary boundary layer (PBL). The PBL is the lowest portion of the atmosphere (1-2 km deep) where pollutants are emitted, transported, mixed, and dispersed. (17) The AERMOD meteorological preprocessor makes use of the surface characteristics of the land surrounding the site along with the hourly surface meteorological data to produce more realistic estimates of parameters that affect dispersion, such as albedo, bowen ratio, and surface roughness. (18)

The PRIME algorithm was developed to correct some shortcomings discovered in the building downwash algorithm used in the ISC model. (19) Using ISC with the PRIME algorithm (ISC-PRIME) should result in more realistic predictions of building downwash effects. The PRIME model algorithm was also added to AERMOD (AERMOD-PRIME). The PRIME models are better at handling the turbulent wake and reduced plume rise caused by the descending flow seen on the leeward side of the building. (19-21)

Air Dispersion Modeling Inputs

The two industrial facilities modeled are both manufacturing facilities located in the eastern United States. The terrain within 3 km of Site 1 is relatively flat. The terrain within 3 km of Site 2 is relatively flat to the west but the terrain to the east is variable and hilly with increasing elevations as distance from the site increases. Site 1 is surrounded by forest to the west and by a combination of grassland and deciduous trees to the east. Site 2 is surrounded by grassland to the west and dense forest with increasing elevation to the east. Both sites are considered rural.

Terrain elevation data representative for each site were obtained from various U.S. Geological Survey (USGS) data sources. Because 7.5-min terrain data at a scale of 1:24,000 were not readily available for both sites, 1[degrees] USGS digital elevation models (DEMs) at a scale of 1:250,000 were used instead. To determine surface roughness, a circle with a radius of 3 km was drawn around the center of each site and the circular area was divided into 12 sectors of 30[degrees] each, starting with sector 1, which was centered on 0[degrees](i.e., due north). The land use within each sector was classified as either water, urban, deciduous forest, coniferous forest, or grassland, consistent with EPA guidance. (18) The areal extent of each land use classification, in square meters and as a percentage of the total sector area, was determined using the Geographical Information System (GIS) software package, ArcView (ESRI).

In practice, emissions emanating from stacks, vent boxes, and tank vents are modeled as point sources. Fugitive emissions from tank farms are modeled as area sources in which the emission rate is divided by the source area to obtain an area-weighted emission rate. Fugitive emissions from process pads and buildings are modeled as volume sources and assigned dimensions on the basis of the building size in accordance with EPA guidance. (13) Following this practice, a representative stack, process building, and tank farm were chosen on each site for this study. Under the MACT rules, it is common practice for high production volume facilities to vent many of their process and fugitive emissions to a common pollution control device such as a scrubber or a thermal oxidizer unit, thus minimizing the number of emission points to be modeled. Table 1 summarizes the source parameters used in the modeling analyses.

The structures near the stacks are typical of structures found at industrial facilities (e.g., pipe racks, sheds, process pads, and process buildings) and are generally less than 6 m in height. The flow of air past structures can result in wakes and cavities forming on the downwind side of the structure, which contributes aerodynamic downwash. (13) As a rule of thumb, if the height of the stack is greater than 2.5 times the height of the nearby structures, the effects of aerodynamic downwash will be avoided. (14) In this study, the heights of both stacks were greater than 2.5 times the height of the nearby structures.

The commercial software packages, BREEZE ISC GIS Pro and BREEZE AERMOD GIS Pro (Version 4.0., Trinity Consultants Inc., 2002) were used for all modeling runs. A steady-state, unit emission rate of 1 g/sec x [m.sup.2] was used for all point and volume sources in all modeling runs. A steady-state, unit emission rate of 1 g/sec x [m.sup.2] was used for all area sources. Emissions were assumed to occur 8760 hr/yr with no downtime. Use of the unit emission rate allows the air modeling output (the ambient air concentration) to be expressed on a unit emission rate basis (i.e., [micro]g/[m.sup.3] per g/sec). The unit emission rate is not chemical specific and its use precludes having to run the model for each individual chemical emitted. To calculate the ambient air concentration of a particular chemical (in [micro]g/[m.sup.3]), the air modeling output (in [micro]g/[m.sup.3] per g/sec) is simply multiplied by the chemical emission rate (in g/sec).

A receptor grid for off-site receptors was set up using a Cartesian grid with a 100-m grid spacing out to a distance of 3 km from the approximate center of each site. Ground level was chosen as the height of all receptors. A fenceline was drawn around each site and the on-site receptors removed from the analysis. The distance from the stack to the fenceline was 150 m for Site 1 and 60 m for Site 2. The models were run in concentration mode for all sources using the 1-hr and total period averaging options, rural dispersion coefficients (ISC only), and the regulatory default options.

For Site 1, all runs were made with a meteorological dataset that contained 4 consecutive years of data and had been approved by the state agency for use in air dispersion modeling at this site. The ISC dataset originally contained a consecutive 5 yr of meteorological data. However, when the dataset was being compiled for AERMOD, from the same stations and for the same 5 yr, we were not able to locate upper air data for the first year. Therefore, for consistency, the ISC and AERMOD data were processed from the same stations for the same 4 yr. The predominant wind direction for Site 2 was from the southwest.

For Site 2, all runs were made with a meteorological dataset that contained 3 consecutive years of meteorological data and had been approved by the state agency for air modeling at this site. The meteorological data for ISC and AERMOD were processed from hourly surface and 12-hr upper air observations recorded at the same National Weather Service stations and for the same 3 yr. The predominant wind direction (i.e., direction from which the wind is blowing) for Site 1 was from the northwest.

For ISC modeling, the meteorological preprocessor, PCRAMMET, was used to create the ISC ready files. The AERMOD files were created using the AERMET preprocessor. In both models, the appropriate land use data were entered directly into the preprocessors. The appropriate land use data were determined from an assessment of the land usage in a 3-km radius around each site. In the case of AERMOD, seasonal variation and land use data were used in the preprocessor to yield different values of albedo, bowen ratio, and surface roughness for the four standard seasons.

All modeling output was collected in plot files that contained geographical coordinates (i.e., 'X' and 'Y' coordinates) for each receptor as well as the modeled ground-level air concentration for the appropriate averaging period. The modeled air concentrations were expressed as [micro]g/[m.sup.3]. However, because the modeled air concentrations were based on a unit emission rate of 1 g/sec they were expressed as [micro]g/[m.sup.3] per g/sec. The modeled air concentrations were multiplied by the source-specific emission rate to generate the predicted air concentrations. The modeled 1-hr and total period average air concentrations were imported into ArcGIS for data analysis and interpretation to assess the impact at the maximally exposed individual (MEI) as well as the spatial distribution of air concentrations and resultant human health risk.

Human Health Risk Assessment

The risk assessment evaluated the potential harm to the modeled receptors due to inhalation of the modeled maximum total period average (i.e., the exposure concentration). Risk assessors refer to the potential harm from exposure to carcinogens as risk and the potential harm from exposure to noncarcinogens as hazard. For noncancer effects, the exposure concentrations are compared with toxicity reference concentrations (RfCs). RfCs are an estimate (with uncertainty spanning perhaps an order of magnitude) of a continuous inhalation exposure to a chemical that is likely to be without an appreciable risk of deleterious effects to the human population (including sensitive subgroups) during a lifetime. For inhalation exposures, noncancer hazards are estimated by dividing the modeled exposure concentration (EC) by the RfC to yield a hazard quotient (HQ) for an individual chemical. (22) The HQ is calculated using eq 1.

HQ = EC/RfC (1)

A HQ of 1 or less for the inhalation pathway indicates that exposure to that chemical is not likely to result in any adverse health effects.

For carcinogenic effects, the lifetime incremental cancer risk (LICR) evaluates the degree to which a receptor may have an increased likelihood of developing cancer over a lifetime due to a lifetime of exposure to a chemical. (22) For carcinogenic effects, the exposure concentrations are compared with the inhalation unit risk (IUR) for a chemical. The LICR is calculated using eq 2.

LICR = EC x IUR (2)

For the great majority of chemicals, the LICR provides an upper-bound prediction of the risk of contracting cancer over a lifetime as a result of a lifetime of exposure (via inhalation) to the chemical at the modeled exposure concentration. LICRs are expressed as a unitless probability and are represented in scientific notation as a negative exponent of 10. For example, the probability of developing cancer of one chance in 10,000 is written as 1 x [10.sup.-4]. In reality, the actual risk may be lower than the predicted risk. (22) EPA cites an acceptable risk range of 1 x [10.sup.-4] to 1 x [10.sup.-6] for potential cancer risk. (23) Table 2 lists the RfCs and IUR values used in this case study. It should be noted that the above equations produce a quite simplistic and conservative estimate of hazard or risk. In reality, a distribution of hazard or risk would more accurately reflect the natural variability observed in humans.

RESULTS

Modeling Results

The study involved four separate model runs to predict the total period average and the 1-hr average air concentrations at all receptors. Because only point sources are affected by building downwash, the comparisons between the standard models and their enhanced versions incorporating the PRIME algorithm were only evaluated for point sources. Table 3 presents the maximum total period average and the maximum 1-hr average air concentrations predicted by the various models and scenarios described above.

AERMOD tends to predict lower maximum air concentrations than ISC for point sources. As presented in the first four rows in Table 3, except in the case of the maximum total period average concentration for Site 1 (in which AERMOD predicts a slightly higher concentration), AERMOD predicts much lower air concentrations than ISC during both averaging periods. The maximum 1-hr average concentration predicted by ISC is more than eight times higher than by AERMOD for Site 2.

Incorporation of the PRIME algorithm tends to decrease the predicted maximum average air concentrations. Comparing the second four rows of Table 3 to the first four rows, the enhanced models, with the PRIME algorithm, predict lower maximum average air concentrations than their standard models in six out of eight comparisons made in this case study. The two exceptions are the maximum total period average concentrations for Site 1. In those two cases, the enhanced models, with the PRIME algorithm, predict higher concentrations, but the differences between the standard and enhanced models are relatively small. Like the standard models, ISC-PRIME predicted higher maximum air concentrations than AERMOD-PRIME. In addition, the differences in the predicted maximum air concentrations between ISC and AERMOD, with and without the PRIME algorithm, are greater for Site 2 where the terrain is more complex than for Site 1.

For the area sources on the two sites, ISC predicted higher maximum air concentrations for Site 1 whereas AERMOD predicted higher maximum air concentrations for Site 2 for both averaging periods. The differences in model performance could be due to the terrain differences between the two sites and/or the enhanced treatment of plume dispersion and growth in AERMOD.

For the volume source on Site 1, AERMOD predicted the higher maximum total period average concentration whereas ISC predicted the higher maximum 1-hr average concentration. However, the magnitude of the observed differences is very small. For the volume source on Site 2, ISC predicted higher maximum air concentration levels than AERMOD for both averaging periods. In the case of the 1-hr averaging period, the maximum air concentration level predicted by ISC is five times higher than by AERMOD.

Overall, the results in Table 3 suggest that the magnitude of the variability between the models is greater for Site 2 than Site 1. We therefore decided to further compare the predicted air concentration levels for Site 2 using GIS mapping. For the point source located on Site 2, ISC-PRIME predicts higher total period air concentrations than AERMOD-PRIME in the receptors closer to the site as shown in Figure 1. As the receptor distance from the site increases, the predicted air concentrations and the shapes of the predicted concentration isopleths from the models become more similar. The same trend is observed for the 1-hr averaging period. The predicted maximum air concentrations, for both averaging periods, are higher for ISC-PRIME. For the 1-hr averaging period, ISC-PRIME predicted higher concentrations close to the source. As the distance from the source increases, the modeled concentrations for ISC-PRIME shift and become lower than those predicted by AERMOD-PRIME. For the total averaging period, the top 10% of the modeled air concentrations from ISC-PRIME are greater than those predicted by AERMOD-PRIME.

[FIGURE 1 OMITTED]

For the area source and volume source on Site 2, ISC and AERMOD predict similar concentrations for the total averaging period for receptors in close proximity to the site. For the 1-hr averaging period for the area source, ISC and AERMOD predict similar concentrations in proximity to the site, but ISC predicts higher air concentrations than AERMOD for the receptors that are away from the site. For the volume source, ISC consistently predicts higher concentrations for the 1-hr averaging period at all receptors.

Human Health Risk Assessment

The HQ and LICR were calculated at the receptor with the maximum total period average air concentration. Table 4 presents the HQs and LICRs calculated using the air concentrations predicted by the different air dispersion models for the point, area, and volume sources on the two sites. All calculated HQs and LICRs were below accepted thresholds of concern at the emission levels modeled. As stated previously, EPA considers a HQ less than 1 to be safe and cites an acceptable range of 1 x [10.sup.-4] to 1 x [10.sup.-6] for potential cancer risk, with 1 x [10.sup.-6] being considered the de minimus value.

There is a linear relationship between the predicted maximum average concentration and the HQ and LICR as indicated by eqs 1 and 2. Therefore, the predicted HQ and LICR values calculated using the results from the different air dispersion models simply mirror the results of the total period maximum average concentrations estimated by the models. For point sources, AERMOD predicts slightly higher air concentrations than ISC for Site 1 and lower air concentrations than ISC for Site 2. For Site 1, this results in a negligible difference in the calculated HI and LICR using both models. For Site 2, the calculated HI and LICR values using the air concentrations predicted by AERMOD are approximately one-third less than the values calculated using the air concentrations predicted by ISC. When the PRIME algorithm is considered, AERMOD-PRIME generates lower air concentration values (and subsequently lower HI and LICR values) than ISC-PRIME for both sites. For Site 1, this results in a negligible difference in the calculated HI and LICR using both models. For Site 2, the calculated HI and LICR values using the air concentrations predicted by AERMOD-PRIME are approximately one-third less than the values calculated using the air concentrations predicted by ISC-PRIME. For area sources, AERMOD generates lower air concentration values for Site 1 and higher air concentration values for Site 2 when compared with ISC. In the case of volume sources, the results are opposite: ISC generates slightly lower air concentration values for Site 1 and higher air concentration values for Site 2 when compared with AERMOD.

DISCUSSION

ISC and AERMOD generate different results because they embed different algorithms for dealing with plume dispersion, plume rise, and underlying surface conditions at the receptors. AERMOD takes into account wind and temperature changes above the stack top in stable meteorological conditions (i.e., little turbulence due to convection or buoyancy) and convective updrafts and downdrafts in unstable meteorological conditions (i.e., increased convective turbulence). However, ISC does not account for convective turbulence. Downdrafts can potentially bring pollutants down to the surface early on and with minimal dilution, thereby creating higher ground-level concentrations closer to the source. Updrafts can carry pollutants further downwind and in different directions. In unstable atmospheres, convective mixing causes an elevated plume to descend over distance. (15,21,24)

AERMOD also handles plume dispersion and plume growth rates differently than ISC. As a plume moves downwind from the release point, it grows in both the vertical and horizontal directions. ISC uses Gaussian models to calculate atmospheric dispersion in both the horizontal and vertical directions. However, AERMOD uses Gaussian models in both the horizontal and vertical directions only under stable conditions. Under unstable conditions, AERMOD uses a Gaussian model in the horizontal direction and a non-Gaussian probability density function in the vertical direction to account for the effects of vertical variations in wind and turbulence on air dispersion. AERMOD's treatment of the vertical air dispersion during unstable conditions is a more accurate portrayal of actual air movement. When the atmosphere is unstable, a surface release encounters turbulence at the ground and is rapidly diluted so that the maximum ground-level concentrations occur close to the source. In stable atmospheres, convective turbulence is minimal and plume dispersion is mainly effected by wind speed. Wind speed changes with height, with lower wind speeds occurring closer to the ground level. In regards to the plume growth rates, ISC uses either rural or urban plume dispersion curves that are a function of distance and one of six possible discrete stability classes. AERMOD uses profiles of vertical and horizontal turbulence that can be either measured or calculated from the meteorological dataset. The vertical profiles vary with height and use continuous growth functions rather than discrete stability classes. Use of turbulence-based plume growth with height gives AERMOD a substantial advancement over ISC. The greatest enhancement would most likely be seen during stable conditions when plume dispersion is minimal because of low turbulence. (15,21,24)

In the meteorological dataset used for Site 2, approximately 75% of the hours were classified as stable or neutral and 25% of the hours were classified as unstable. In this case study, ISC predicted higher maximum ground-level concentrations close to the sources for both the 1-hr and total period averaging time frames. AERMOD predicted lower maximum ground-level concentrations than ISC possibly because of its ability to better handle the stable periods. Because a majority of the meteorological hours were stable, AERMOD should allow for increased dispersion, which would result in lower maximum ground-level concentrations. The enhanced handling of ground-level releases in AERMOD may help explain why AERMOD predicts higher maximum ground-level concentrations for releases from the area sources.

The underlying surface conditions at the receptors were handled differently by the two models. ISC does not consider the underlying surface conditions at the receptors. However, AERMOD uses data on the underlying surface conditions at the receptors as well as data on seasonal variation in its meteorological preprocessor to calculate the values of albedo, bowen ratio, and surface roughness for all four seasons. AERMOD's consideration of the surface characteristics at the receptor are potentially important because the turbulence caused by surface friction will certainly affect the modeling results, especially in windy conditions. Wind speed is retarded because of the frictional effects caused by surface roughness. The more obstacles in the path of the wind, the greater the turbulence and frictional effects; therefore, the wind could blow much stronger several meters above a plowed field than it would above an urban area. (24) In the case of Site 2, the surface roughness in the forest surrounding the site was greater than the surface roughness in the grassland. The predicted maximum ground-level concentrations occurred in a sector predominated by forest. In this case, the increased surface roughness leads to increased turbulence, faster dispersion of the plume, and therefore, the maximum air concentration occurs closer to the emission source. In this case study, ISC, using rural dispersion coefficients, predicted higher air concentrations. The surface frictional effects may explain why AERMOD predicted higher 1-hr average concentrations for the receptors farther away from the sources.

One fundamental enhancement in air dispersion modeling was the addition of the PRIME algorithm for calculating building downwash effects into both ISC and AERMOD models. Buildings in the path of a plume create downwash effects, which essentially bring the contaminants to ground level more quickly than in a wide-open area. As discussed above, the modeling results presented in Table 3 tend to confirm that theoretical expectation that the PRIME versions of the air dispersion models predict lower air concentrations than their standard models. Another observation from Table 3 is that the variations in the predicted air concentrations between AERMOD and AERMOD-PRIME tend to be smaller than between ISC and ISC-PRIME especially for the 1-hr average air concentrations. Such an observation is reasonable because AERMOD embeds the more updated and realistic air dispersion processes that lead to lower ground-level air concentrations as discussed above.

CONCLUSIONS

Numerous federal and state environmental regulations in the United States incorporate risk assessment into the policy-making process. In most cases, these decisions are made based on a quantitative human health risk assessment in which the estimated risks are compared with a set of acceptable criteria. A quantitative risk assessment dealing with industrial emissions to the atmosphere requires adequate site-specific data and realistic physical air dispersion modeling. An air dispersion model that tends to overpredict or underpredict air concentrations of pollutants can directly affect the outcome of a risk assessment, which can have a profound impact on decision-making. Because AERMOD has replaced ISC, it would be necessary to compare AERMOD and ISC to evaluate the potential impacts of the replacement on human health risk assessments.

This study evaluates the performances of AERMOD, ISC, and their PRIME versions for point, area, and volume sources on two realistic sites. In general, the switch from ISC to AERMOD and the incorporation of the PRIME algorithms tends to generate lower air concentration estimates for point sources at the point of maximum air concentration. However, the magnitude of difference varies from insignificant to significant depending on the types of the sources and the site-specific conditions. The spatial distribution of the predicted maximum air concentration for the point source shows ISC predicted higher air concentrations nearer the site than AERMOD for both averaging periods. As the distance from the site increased, the predicted air concentrations and the shapes of the concentration isopleths become similar. Therefore, the impact of the proposed change on resultant risk to human health when considering chronic exposure can be significant due to the decreased air concentration maximums predicted by AERMOD. The biggest differences between the models appear to occur for the 1-hr averaging period. This could have implications for acute risk assessments and odor evaluations where decisions are often made based on the predicted maximum 1-hr average air concentrations.

In all modeled scenarios, the predicted points of the maximum average air concentrations were separated by no more than 300 ft. In reality, the point of maximum exposure may not have an actual receptor present and it potentially can be located in the middle of a road or in a body of water. It is therefore essential to assess the spatial distribution of air concentrations of pollutants when determining the magnitude and extent of the health impact of air emissions from industrial sources. The spatial distribution will allow for a more accurate assessment of impact or risk to human health and the environment. GIS packages are especially suited to handling the large concentration datasets generated by air dispersion models. The GIS allows for viewing and interpretation of these large spatial and temporal datasets and is a valuable tool for assessing risk. Spatial analysis of risk will allow industry and policy makers to assess site-specific exposures, evaluate the extent of risk to exposed populations, and determine the potential risk to environmental systems such as specific water bodies. Spatial issues may therefore become a key component of environmental and health decision-making in the future.

Because there is a relationship between the modeled air concentration, the HI, and the LICR, the predicted values using different air dispersion models will vary. In this case study, AERMOD and AERMOD-PRIME tended to estimate lower air concentrations. In one case, the LICR based on the air concentration predicted by AERMOD is about one-third of the LICR based on the air concentration predicted by ISC. In this study, all the calculated HIs and LICRs are below accepted thresholds of concern. This case study was performed to gain some practical experience with the models. However, it only examined a limited number of sources on a limited number of sites. Future studies should address sites with multiple sources and sites located in areas of extremely complex terrain.

REFERENCES

1. 40 CFR Part 51: Revision to the Guideline on Air Quality Models: Adoption of a Preferred General Purpose (Flat and Complex Terrain) Dispersion Model and Other Revisions. Fed. Regist. 2005, 70(216), 68218-68261.

2. Hall, D.J.; Spanton, A.M.; Bennett, M.; Dunkerley, F.; Griffiths, R.F.; Fisher, B.E.A.; Timmis, R.J. Evaluation of New Generation Atmospheric Dispersion Models; Int. J. Environ. Pollut. 2002, 18, 22-32.

3. Hanna, S.R.; Egan, B.A.; Purdum, J.; Wagler, J. Evaluation of the ADMS, AERMOD, and ISC3 Dispersion Models with the OPTEX, Duke Forest, Kincaid, Indianapolis and Lovett Field Datasets; Int. J. Environ. Pollut. 2001, 16, 301-314.

4. Irwin, J.S.; Carruthers, D.; Stocker, J.; Paumier, J. Application of ASTM D6589 to Evaluate Dispersion Model Performance; Int. J. Environ. Pollut. 2003, 20, 4-10.

5. Radonjic, Z.; Garisto, N. Dispersion Modeling, Comparison to Available Data and Model Inter-Comparison at Pickering Nuclear Generating Station (PNGS) Using ISCST3, ISC-Prime, AERMOD, and CALPUFF; In Proceedings of A & WMA's Air Quality Modeling: New Methods for a New Reality; A & WMA: Pittsburgh, PA, 2004.

6. Ramakrishnan, D.; Zwicke, G.; Wall, D.; Remsberg, A.M, Jr. A Performance Comparison of AERMOD vs. Current Guideline Models in a Real World Scenario; In Proceedings of A & WMA's Guideline on Air Quality Models: The Path Forward; A & WMA: Pittsburgh, PA, 2003; Paper 03-A-27.

7. Sax, T.; Isakov, V. A Case Study for Assessing Uncertainty in Local-Scale Regulatory Air Quality Modeling Applications; Atmos. Environ. 2003, 37, 3481-3489.

8. Schulze, R.H. Procedures Used by the United States Environmental Protection Agency (U.S. EPA) to Develop and Adopt Newer Dispersion Models; Int. J. Environ. Pollut. 2001, 16, 483-494.

9. Schulze, R.H.; Dai, W.; Otto, C.M. Managing Air Quality during Regulatory Changes; Int. J. Environ. Pollut. 2003, 20, 121-127.

10. Tarde, J.A.; Westbrook, J.A. Air Quality Modeling in a Highly Industrialized Valley Regime: a Comparison of AERMOD-PRIME to ISCST-PRIME and ISCST3 Results for [PM.sub.10] Emissions. In Air & Waste Management Association's Guideline on Air Quality Models: the Path Forward; A & WMA: Pittsburgh, PA, 2003.

11. Perry, S.G.; Cimorelli, A.J.; Paine, R.J.; Brode, R.W.; Weil, J.C.; Venkatram, A.; Wilson, R.B.; Lee, R.F.; Peters, W.D. AERMOD. A Dispersion Model for Industrial Source Applications; Part II: Model Performance Against 17 Field Study Databases; J. Appl. Meteorol. 2005, 44, 694-708.

12. Brownell, F.W. Clean Air Act. In Environmental Law Handbook, 16th ed.; Sullivan, T., Ed.; Government Institutes: Rockville, MD, 2001, Chapter 5, pp 218-222.

13. User's Guide for the Industrial Source Complex (ISC3) Dispersion Models Volume I: User Instructions; EPA-454/B-95-003a; U.S. Environmental Protection Agency. Office of Air Quality Planning and Standards: Research Triangle Park, NC, 1995.

14. Boubel, R.W.; Fox, D.L.; Turner, D.B.; Stern, A.C. Fundamentals of Air Pollution, 3rd ed.; Academic: San Diego, CA, 1994.

15. Users Guide to the Building Profile Input Program (BPIP); EPA-454/R-93-038; U.S. Environmental Protection Agency; Office of Air Quality Planning and Standards: Research Triangle Park, NC, 1995.

16. Compendium of Reports from the Peer Review Process for AERMOD; U.S. Environmental Protection Agency; Office of Air Quality Planning and Standards: Research Triangle Park, NC, 2002; available at http://www.epa.gov/scram001/7thconf/aermod/dockrpt.pdf (accessed 2006).

17. Sorbjan, Z. Air Pollution Meteorology. In Air Quality Modeling--Theories, Methodologies, Computational Techniques, and Available Databases and Software. Vol. 1: Fundamentals; Zannetti, P., Ed.; EnviroComp Institute and A & WMA: Pittsburgh, PA, 2003; Chapter 4, pp 37-100.

18. Revised Draft User's Guide for the AERMOD Meteorological Preprocessor (AERMET); U.S. Environmental Protection Agency; Office of Air Quality Planning and Standards: Research Triangle Park, NC, 1998; available at http://www.epa.gov/scram001/7thconf/aermod/aermetug.pdf (accessed 2006).

19. Paine, R.J.; Lew, F. Consequence Analysis for ISC-PRIME, Prepared for the Electric Power Research Institute (EPRI), Palo Alto, CA; EPRI Publications No. TR-2460026, National Technical Information Service (NTIS) Publications No. PB98-156516; U.S. Department of Commerce, NTIS: Springfield, VA, 1997.

20. Schulman, L.L.; Strimaitis, D.G.; Scire, J.S. Addendum to ISC3 User's Guide. The Plume Rise and Building Downwash Model, 1997; available at http://www.epa.gov/scram001/7thconf/iscprime/useguide.pdf (accessed 2006).

21. Trinity Consultants. BREEZE ISC and AERMOD Users Guide, Version 3.5. Trinity Consultants: Dallas, TX, 2001.

22. Air Toxics Risk Assessment Reference Library, Vol. 1: Technical Resource Manual; EPA-453-K-04-001A. U.S. Environmental Protection Agency; Office of Air Quality Planning and Standards: Research Triangle Park, NC, 2004.

23. National Research Council. Science and Judgement in Risk Assessment; National Academies: Washington, DC, 1994.

24. Beychok, M.R. Fundamentals of Stack Gas Dispersion, 3rd ed. Newport Beach, CA, 1994.

About the Authors

Dr. Keith Silverman is a director and Dr. Joan Tell is a senior project engineer in the Global Safety and the Environment Department at Merck & Co., Inc. based in Whitehouse Station, NJ. Dr. Edward Sargent is the managing director of EV Sargent LLC and Dr. Zeyuan Qiu is an assistant professor in the Department of Chemistry and Environmental Sciences at New Jersey Institute of Technology, Newark, NJ. Please address correspondence to: Keith Silverman, Merck & Co., Inc., 2 Merck Drive, Whitehouse Station, NJ 08873; phone: +1-908-423-4102; fax: +1-908-735-1496; e-mail: keith_silverman@merck.com.

Keith C. Silverman and Joan G. Tell

Global Safety and the Environment, Merck & Co., Inc., Whitehouse Station, NJ

Edward V. Sargent

EV Sargent LLC., Watchung, NJ

Zeyuan Qiu

Department of Chemistry and Environmental Sciences, New Jersey Institute of Technology, Newark, NJ