Differences Between the 1999 and 1996 NJDEP and USEPA Risk Analyses
USEPA’s 1999 NATA includes emissions and ambient concentrations for 177 air toxics plus diesel particulate matter (PM). It also includes an exposure and risk assessment (cancer risk and noncancer hazard) for 133 of these air toxics, plus diesel PM, based on toxicity data for chronic exposure.
For 1996, EPA focused on exposure concentrations and risk for 33 pollutants.
About 80 chemical-specific tables with county-level ambient concentration data are available at the USEPA 1999 NATA web site. NJDEP carried out its own risk assessment on those chemicals, comparing chemical-specific cancer or noncancer health benchmarks (described at LINK) to the modeled ambient concentration averages for each county and the entire state. 22 pollutants were of concern in New Jersey, that is, the NATA modeled concentration exceeded the health benchmark, either at the county or state level.
Six pollutants had countywide modeled concentrations equal to zero, meaning that they were not emitted in New Jersey or in nearby areas of bordering states.
Hazardous Air Pollutants with Zero Modeled Concentrations in New Jersey
Pollutant |
CAS No.* |
Benzidine |
92875 |
Coke Oven Emissions |
|
Ethyl carbamate |
51796 |
4-Dimethyl aminoazobenzene |
60117 |
4,4-Methylenebis(2-chloroaniline) |
101144 |
Toluene-2,4-diamine |
95807 |
*CAS No. – Chemical
Abstracts Service (CAS) Registry Number - a unique
number assigned by the CAS to each distinct chemical
substance recorded in the CAS registry.
The following 52 compounds had risk ratios that were equal to or less than one, based on New Jersey state- or countywide average modeled concentrations. Therefore, individually, they do not pose a risk in any counties in New Jersey.
Hazardous Air Pollutants with State- or Countywide Risk Ratios Equal to or Less Than One
|
Pollutant |
CAS No. |
1 |
Acetonitrile |
75058 |
2 |
Acrylamide |
79061 |
3 |
Acrylic acid |
79107 |
4 |
Acrylonitrile |
107131 |
5 |
Allyl chloride |
107051 |
6 |
Aniline |
62533 |
7 |
Antimony Compounds |
|
8 |
Benzotrichloride |
98077 |
9 |
Benzyl chloride |
100447 |
10 |
Beryllium Compounds |
|
11 |
Bis(chloromethyl) ether |
542881 |
12 |
Chlorine |
7782505 |
13 |
Chloroprene |
126998 |
14 |
Cobalt Compounds |
|
15 |
Cyanides |
|
16 |
1,2-Dibromo-3-chloropropane |
96128 |
17 |
Dichloroethyl ether |
111444 |
18 |
Diethanolamine |
111422 |
19 |
Dimethyl formamide |
68122 |
20 |
2,4-Dinitrotoluene |
121142 |
21 |
p-Dioxane |
123911 |
22 |
Epichlorohydrin |
106898 |
23 |
Ethyl acrylate |
140885 |
24 |
Glycol Ethers |
|
25 |
Hexachlorobenzene |
118741 |
26 |
Hexachloroethane |
67721 |
27 |
Hexamethylene Diisocyanate |
822060 |
28 |
Hexane |
110543 |
29 |
Hydrazine |
302012 |
30 |
Hydrochloric acid |
7647010 |
31 |
Hydrofluoric acid |
7664393 |
32 |
Lead Compounds |
|
33 |
Maleic anhydride |
108316 |
34 |
Manganese Compounds |
|
35 |
Mercury Compounds |
|
36 |
Methyl bromide |
74839 |
37 |
Methylene Chloride |
75092 |
38 |
4,4,-Methylenedianiline |
101779 |
39 |
4,4-Methylenediphenyl diisocyanate |
101688 |
40 |
Phosgene |
75445 |
41 |
Polychlorinated Biphenyls |
|
42 |
Polycyclic Organic Matter |
|
43 |
Propylene Dichloride |
78875 |
44 |
Propylene oxide |
75569 |
45 |
Quinoline |
91225 |
46 |
Titanium tetrachloride |
7550450 |
47 |
Toluene |
108883 |
48 |
2,4-Toluene diisocyanate |
584849 |
49 |
o-Toluidine |
95534 |
50 |
1,1,2-Trichloroethane |
79005 |
51 |
Vinyl Chloride |
75014 |
52 |
Xylenes |
106423 |
1999 Chemicals of Concern in New Jersey
The following chemicals that USEPA included in the 1999 NATA modeling are of concern in New Jersey, because their modeled average ambient concentrations are above their health benchmarks in a county or across the entire state.
Pollutant |
Number/Name
of Counties
Above
Health Benchmarks* |
Primary Emissions Source |
Acetaldehyde |
21 |
Onroad & background |
Acrolein |
21 |
Mobile, area |
Arsenic Compounds |
4 (Atlantic, Camden, Cape May, Gloucester) |
Major & area |
Benzene |
21 |
Mobile |
Bis(2-ethylhexyl)phthalate |
21 |
Background |
1,3-Butadiene |
21 |
Onroad & background |
Cadmium Compounds |
1 (Warren) |
Area |
Carbon Tetrachloride |
21 |
Background |
Chloroform |
20 |
Area, background |
Chromium VI |
14 |
Area |
1,4-Dichlorobenzene |
4 (Atlantic, Essex, Hudson, Passaic) |
Area |
1,3-Dichloropropene |
1 (Hudson) |
Area |
Diesel PM |
21 |
Mobile |
Ethylene Dibromide |
21 |
Background |
Ethylene Dichloride |
11 |
Background |
Ethylene Oxide |
2 (Hudson, Warren) |
Area |
Formaldehyde |
21 |
Mobile, background |
Methyl chloride |
21 |
Background |
Naphthalene |
14 |
Area |
Nickel Compounds |
1 (Camden) |
Area, major |
Perchloroethylene |
10 |
Area, background |
1,1,2,2-Tetrachloroethane |
21 |
Background |
*For more information on which areas
are impacted by the chemicals of concern, see the chemical-specific
maps click here.
Beryllium, hydrazine, and polycyclic organic matter (POM) were chemicals of concern in New Jersey in the 1996 NATA, but for 1999, their modeled air concentrations have been reduced to where they no longer show a high risk. Since USEPA constantly makes changes to its emission inventory methods, with the goal of improving them, it is difficult to pinpoint exactly why the concentrations have decreased. USEPA states that “…it is not meaningful to compare the two national-scale assessments. This is because any change in emissions, ambient concentrations, or risks may be due to either improvements in methodology or to real changes.” (See 1999 National-Scale Air Toxics Assessment – Frequent Questions, number 10.
Chromium. For 1999, USEPA grouped emissions of the toxic hexavalent form of chromium separately from other chromium compounds. For 1996, emissions of hexavalent chromium were derived from total chromium emissions, assuming that 34% of all chromium compound emissions were hexavalent. This led to a gross overestimation, particularly for facilities that did not emit any hexavalent chromium. For example, in USEPA’s 1996 NATA risk evaluation, a census tract in Camden County was identified as having the highest cancer risk in the country, based on the assumption that 34% of a specific facility’s chromium emissions were hexavalent. In reality, the facility did not emit any hexavalent chromium at all.
The chemicals listed below were added to the list of New Jersey chemicals of concern, either because USEPA added them to the NATA analysis for 1999, or because their emissions increased:
- Bis(2-ethylhexyl)phthalate
- Ethylene oxide
- Methyl chloride
- Naphthalene
- 1,4-Dichlorobenzene
- 1,3-Dichloropropene
- 1,1,2,2-Tetrachloroethane
Trichloroethylene, one of the 33 pollutants in 1996 NATA, does not currently have county summaries available on the USEPA web site, so it could not be included in the NJDEP risk assessment.
Differences Between the NJDEP and USEPA Risk Analysis
The most dramatic differences between NJDEP’s and USEPA’s 1999 NATA risk analysis are the results for formaldehyde, chloroform, methyl chloride, and diesel PM. This comes from using different toxicity information. NJDEP used the same toxicity information that is used in the NJDEP Air Quality Permitting Program, and can be found here. USEPA describes its toxicity information in Health Effects Information Used in Cancer and Noncancer Risk Characterization for the 1999 National-Scale Assessment. Although most of the toxicity information used by NJDEP is the same as USEPA’s, there are some critical differences, described in more detail below.
Differences in Toxicity Information
Formaldehyde. NJDEP’s cancer risk estimates for formaldehyde are much higher than USEPA’s because of the different toxicity factors used. The number used by USEPA is about 2400 times smaller. NJDEP used a cancer unit risk factor (URF) from USEPA’s own Integrated Risk Information System (IRIS) database. However, for its 1999 NATA analysis, USEPA chose to use a URF developed by the CIIT Centers for Health Research (formerly the Chemical Industry Institute of Technology). USEPA says that this toxicity value will be put through the usual IRIS review process, but the process takes some time. However, the newest studies being done seem to show that formaldehyde is even more toxic than previously thought. In light of this information, NJDEP decided that using the more stringent value from IRIS was most appropriate. USEPA describes its approach at its 1999 NATA web site.
The health benchmark for formaldehyde used by NJDEP is 0.077 ug/m3, based on a URF of 1.3 x 10-5/ug/m3. USEPA used a URF of 5.5 x 10-9/ug/m3, which converts to a health benchmark of 182 ug/m3. Not surpisingly, USEPA did not find formaldehyde to contribute significantly to the overall cancer risk estimated in NATA. Using NJDEP’s approach, the contribution from formaldehyde to overall cancer risk in New Jersey is quite significant, second only to diesel PM.
Chloroform. Again, NJDEP used a toxicity value from USEPA’s IRIS database that USEPA chose not to use. USEPA did not evaluate cancer risk from chloroform at all, stating that they did not use the IRIS number because they are “currently developing a new dose-response assessment for this chemical.”
Methyl Chloride. USEPA did not evaluate carcinogenic risk. NJDEP did, using a URF from a 1997 USEPA document.
Diesel PM. NJDEP’s estimate of risk from diesel PM is discussed in detail at DEP’s diesel air toxics page.
All of NJDEP’s toxicity values and their sources can be found in the documents “Unit Risk Factors for Inhalation,” and “Reference Concentrations for Inhalation”.
Differences in Exposure Assessment
Another major difference in the risk
analysis approaches taken by USEPA and NJDEP is that
USEPA carried out an exposure assessment using a model
called HAPEM, the Hazardous Air Pollutant Exposure Model. This
is a screening-level human exposure model designed to
estimate inhalation exposures of population subgroups
to hazardous
air pollutants (HAPs). This model uses a series
of assumptions about where different groups of people
spend different parts of their day (“microenvironments”). The
net effect of using HAPEM is that it decreases the ambient
air concentrations calculated for NATA using the ASPEN
dispersion model. This results in a decrease in
the concentrations that NATA estimates people are exposed
to. There is some debate about whether USEPA is
justified in going through the step of applying HAPEM
to the modeled ambient concentrations. The assumptions
about exposures in different micorenvironments may not
be that accurate, and may be more variable than accounted
for in the model. There are particular concerns
that HAPEM does not adequately represent indoor exposures. Many
studies comparing indoor air to outside (ambient air)
levels of pollutants have shown that indoor exposures
are usually higher. The end result of USEPA using
HAPEM for the NATA risk analysis is that the risks are
lower than those estimated by NJDEP, because we used
the original ASPEN ambient concentrations for our evaluation.
back
to top |