Jump to main content.


EPA's Observations Regarding Results from the 1996 ASPEN Dispersion Modeling

Information provided for informational purposes onlyNote: EPA no longer updates this information, but it may be useful as a reference or resource.
Feb 26, 2000

Synopsis of Observations

  • Ambient levels of 18 HAPs are due to emissions from major and area and other sources only; 15 also come from mobile sources. Diesel PM originates only from mobile sources.
  • 6 HAPs (carbon tetrachloride, chloroform, ethylene dibromide, ethylene dichloride, hexachlorobenzene, and PCBs) have ambient levels which are dominated on average by a background which is due to historic emissions (the background contributions for these compounds are generally due to the chemically persistent nature of these compounds combined with their ability to be re-emitted into the atmosphere after depositing on land or water).
  • On the average nationwide, ambient levels for 6 of the HAPs (acetaldehyde, acrolein, benzene, 1,3-butadiene, formaldehyde, and lead compounds) are dominated by contributions from mobile source sectors.
  • When summarized at the State level, very few areas of the country show substantially different levels for individual pollutants than any other area. A few States tend to show higher levels for several pollutants, and these tend to be the States which have higher populations and a higher percentage of urban areas than rural areas. This is intuitively consistent with the notion that many of these pollutants are either associated with mobile sources or areawide sources which tend to locate near areas of higher population.
  • A small percentage (much less than 1%) of county-level maximum predicted concentrations for several pollutants are extremely high, based on knowledge from current measured ambient levels. These unreasonably high estimates are likely due to artifacts in the formulation of the modeling system, and should not be interpreted literally. EPA will continue to investigate such anomalies to ensure that they are indeed artifacts and to correct them for future assessments. In the meantime, analysts should focus their interpretations on less extreme statistics (e.g., 5th to 95th percentile values).
  • Several States show higher 95th percentile levels for a few pollutants, relative to the other States, than one would intuitively expect. While the actual 95th percentile values may be accurate, it may not necessarily indicate that these States are significantly higher than the rest of the States. Rather, it may indicate that the emission inventories for certain pollutants are comprehensive in a few States and not so in the others. This may provide us with an indication of the pollutants and the States where additional emission inventory work needs to be performed for future assessments, and thus merits further investigation.
  • Details on these and other observations are provided below.
Detailed observations from ASPEN output charts

I. Source sector percent contribution charts

     

  1. 18 of the HAPs originate from major and area and other sources only (methylene chloride, trichloroethylene, carbon tetrachloride, chloroform, ethylene dibromide, ethylene dichloride, hexachlorobenzene, PCBs, coke oven emissions, hydrazine, propylene dichloride, acrylonitrile, 1,3-dichloropropene, ethylene oxide, perchloroethylene, 1,1,2,2-tetrachloroethane, quinoline, and vinyl chloride). The remaining 15 (acetaldehyde, acrolein, benzene, 1,3-butadiene, formaldehyde, dioxins/furans, POM/7-PAH and lead, arsenic, beryllium, cadmium, chromium, manganese, mercury, and nickel compounds) originate from major, area, and mobile source sectors. Diesel PM originates solely from mobile sources.

     

  2. Of the 18 HAPs which are not emitted from mobile sources, 8 have significant background contributions; 2 of these (methylene chloride and trichloroethylene) result from a mix of stationary sources and background, and 6 of these (carbon tetrachloride, chloroform, ethylene dibromide, ethylene dichloride, hexachlorobenzene, and PCBs) have ambient levels which are dominated on average by background. The background contributions for these compounds are generally due to the chemically persistent nature of these compounds combined with their ability to be re-emitted into the atmosphere after depositing on land or water. In this case, the dominance of the background contribution indicates that current emissions (from all sources sectors) are very low on a national basis.

     

  3. Ambient levels of 4 HAPs (coke oven emissions, hydrazine, propylene dichloride, and acrylonitrile) are dominated by emissions from major sources.

     

  4. Ambient levels of 4 HAPs (1,3-dichloropropene, ethylene oxide, perchloroethylene, and 1,1,2,2-tetrachloroethane) are dominated by emissions from area and other sources.

     

  5. Ambient levels of 2 HAPs (quinoline and vinyl chloride) result on average from a mix of emissions from both major, area and other sources.

     

  6. As a result of missing source contributions from the onroad mobile source sector, there appears to be missing emissions information for a few pollutants (notably benzene and 1,3-butadiene) associated with mobile sources in the States of Vermont and New Hampshire. This problem is being pursued with those States and will be corrected for the final assessment.

     

  7. On the average nationwide, ambient levels for 6 of the HAPs (acetaldehyde, acrolein, benzene, 1,3-butadiene, formaldehyde, and lead compounds) are dominated by contributions from mobile source sectors. Of these, 3 HAPs (acetaldehyde, benzene, and 1,3-butadiene) have contributions from onroad mobile sources which exceed those from nonroad. Two HAPs (formaldehyde and acrolein) have nonroad contributions which exceed those from onroad mobile sources. Only one HAP (lead compounds) has mobile source contributions which come almost exclusively from nonroad sources.

    [NOTE: Just because ambient levels are dominated on average by emission contributions from one source sector does not mean that emissions from another source sector can never dominate in any particular local area.]

     

  8. Ambient levels of 7 metal HAPs (arsenic, beryllium, cadmium, chromium, manganese, mercury, and nickel compounds) and 2 semi-volatile HAPs (dioxins/furans and POM/7-PAH) have emission contributions from mobile source sectors which are small. Ambient levels of all of the metal HAPs have significant contributions from both major, area and other sources.

     

  9. Of the 6 HAPs whose ambient levels are dominated by mobile source contributions, only 2 (benzene and formaldehyde) have significant contributions from "background". This means that these 2 compounds either have significant contributions from natural sources (not generally included in the emission inventory) or that they can transport long distances (beyond 50 km from their source) in the atmosphere, or both. The other 4 HAPs have no significant background contribution, meaning that they have no significant natural sources and that they do not tend to transport beyond 50 km from their original source (generally because they chemically degrade in or fall out of the atmosphere before they can transport that far.)

     

  10. Of the metals and semi-volatiles emitted by mobile sources, only mercury and dioxins/furans have significant background contributions. In general, these background contributions are due to the chemically persistent nature of these 2 compounds combined with their ability to be re-emitted into the atmosphere after depositing on land or water (the remaining metals and semi-volatiles do not re-emit into the atmosphere as easily after deposition).

     

  11. Ambient levels of diesel PM are dominated by nonroad mobile source contributions (75% versus 25% contributed by the onroad mobile sources).
II. Statewide concentration distributions (boxplots)

     

  1. General observations
    1. When summarized at the State level, very few areas of the country show substantially different levels for individual pollutants than any other area. A few States tend to show higher levels for several pollutants, and these tend to be the States which have higher populations and a higher percentage of urban areas than rural areas. This is intuitively consistent with the notion that many of these pollutants are either associated with mobile sources or areawide sources which tend to locate near areas of higher population.
    2. Several States show higher 95th percentile levels for a few pollutants, relative to the other States, than one would intuitively expect. While the actual 95th percentile values may be accurate, it may not necessarily indicate that these States are significantly higher than the rest of the States. Rather, it may indicate that the emission inventories for certain pollutants are comprehensive in a few States and not so in the others. This may provide us with an indication of the pollutants and the States where additional emission inventory work needs to be performed for future assessments, and thus merits further investigation.
    3. Washington, DC has higher 5th percentile levels than most every State for most every pollutant. This is to be expected when comparing an exclusively urban area to States, which are a mixture of rural and urban areas.

     

  2. Major, area and other source dominated VOCs
    1. . Ambient levels of ethylene oxide vary moderately from State to State, with Minnesota and Wisconsin appearing to have significantly higher levels than the rest of the nation.
    2. Ambient levels of acrylonitrile show wide State-to-State variability with consistently low median values. This wide variability is expected for a pollutant whose emissions are dominated by major source contributions.
    3. Ambient levels of hydrazine vary only moderately across the nation, with a few areas of high concentration showing up in Arkansas, New Jersey, New York, and North Carolina.
    4. Ambient levels of quinoline show wide State-to-State variability, with a few areas of higher concentrations in Ohio, West Virginia, Indiana, Kentucky, and Michigan.

     

  3. Pollutants dominated by background
    1. Ambient levels due to current emissions of carbon tetrachloride, PCBs, and ethylene dibromide appear to be dwarfed by the background due to the environmental persistence of historic emissions.
    2. In general, ambient levels due to current emissions of hexachlorobenzene appear to be dwarfed by the background due to the environmental persistence of historic emissions. A few areas of high concentration show up in New York, Oklahoma, and Virginia.
    3. In general, ambient levels due to current emissions of ethylene dichloride appear to be dwarfed by the background due to the environmental persistence of historic emissions. A few areas of high concentration show up in Illinois, Indiana, New Jersey, and Pennsylvania.

     

  4. Chlorinated hydrocarbons from major, area and other sources
    1. Ambient patterns for chloroform indicate that background generally dominates as noted above. Statewide averages for New Hampshire and Washington exceed their respective 95th percentiles; this indicates the presence of a few very high ambient predictions which are skewing the average value. Since the contributing sources from both of these States are primarily major, this indicates the potential for high ambient levels near individual or groups of major sources.
    2. Ambient levels of perchloroethylene (or tetrachloroethylene) show moderate variability across the nation, with a few areas of high concentration appearing in Colorado, Illinois, California, New York, and Massachusetts.
    3. Wisconsin appears to have a few areas with very high concentrations of propylene dichloride due to emissions from the major source sector. All other areas of the country show very low ambient levels. ???
    4. Ambient levels of trichloroethylene vary moderately from State to State, with a few areas of high concentrations appearing in Wisconsin, Washington, Massachusetts, and Georgia.
    5. Ambient levels of vinyl chloride vary significantly from State to State, but Delaware appears to have an unusually high number of areas with concentrations greater than the rest of the nation. ???
    6. Ambient levels of 1,1,2,2-tetrachloroethane vary from State to State, with a few areas of high concentration showing up in California and Colorado.
    7. Ambient levels of methylene chloride vary only moderately from State to State, with a few areas of high concentration showing up in California, Massachusetts, New York, North Carolina, and Tennessee.
    8. Ambient levels of 1,3-dichloropropene show moderate variability from State to State, with a few areas of high concentrations appearing in New York, California, New Jersey, Maryland, and Massachusetts.

     

  5. Metals
    1. Ambient levels of arsenic compounds are rather consistent from State to State, with a few areas of higher levels in Michigan, Nebraska, North Carolina, South Carolina, and West Virginia.
    2. Ambient levels of beryllium compounds tend to be higher in California and North Carolina than the rest of the nation.
    3. Ambient levels of cadmium compounds tend to be higher in Wisconsin than the rest of the nation.
    4. Ambient levels of chromium compounds show high variability from State to State.
    5. Ambient levels of manganese compounds are generally consistent, with a few areas of higher concentration showing up in Indiana, North Carolina, and West Virginia.
    6. Ambient levels of mercury tend to be dominated by background in most States, with a few areas of higher concentration showing up in Maine and Wisconsin.
    7. Ambient levels of lead compounds vary nationwide, with areas of higher concentration appearing in Nebraska and New York.
    8. Ambient levels of nickel compounds vary from State to State, with a few areas of higher concentration showing up in New York, Oklahoma, Utah, and West Virginia.

     

  6. Particulate air toxics
    1. Ambient levels of POM and 7-PAH (a subset of POM) vary from State to State, with generally higher levels showing up in Illinois, Michigan, and Minnesota. The 7-PAH subset alone shows some generally higher levels in Oregon and Washington.
    2. Ambient levels of dioxins/furans are bracketed by upper and lower bound estimates, driven by differing assumptions about grouping the pollutant emission inventory estimates. Several States show wide variation between the bounds: California, Illinois, Michigan, Minnesota, and Missouri. This gives rise to a wide variation in the national pattern.
    3. Ambient impacts of coke oven emissions vary widely as might be expected for an emissions inventory dominated by point sources. Areas of higher concentration appear in Illinois and Indiana.

     

  7. Mobile source dominated VOCs
    1. Minnesota appears to have significantly higher 1,3-butadiene levels when compared to the national distribution (the median is more than 3 times higher than that for the nation). Otherwise, 1,3-butadiene levels vary only moderately across the entire nation, with slightly higher concentrations appearing in more densely populated areas.
    2. Formaldehyde levels in a few areas of Louisiana appear to be quite high, but the median for the State is quite comparable to the rest of the nation, which shows moderate variation from State to State.
    3. Ambient levels of benzene, acetaldehyde, and acrolein vary moderately across the entire nation, with slightly higher concentrations appearing in more densely populated areas.
  8. Diesel PM
    1. Ambient levels of diesel PM vary moderately from State to State, with a few areas of higher concentrations in California, Colorado, Illinois, Massachusetts, New Jersey, and New York.v

Local Navigation


Jump to main content.