HAZARDOUS AIR POLLUTANT HANDBOOK: Measurements, Properties, and Fate in Ambient Air - Part 4 pot

The results of that first stage survey have been reported.4,5Ambient concentrations for 70 of the 188 HAPs were compiled through 1987 in the National VOC Data Base1,2, which was updated

As a result, data with which to evaluate the potential public health risks from the 188 HAPs may not be readily available

This chapter summarizes the results of a survey of ambient air concentrations of the 188 HAPs.

To the extent possible, the definition of an ambient measurement used in this survey was that stated

in section 3.1, i.e., a measurement in the open atmosphere away from direct source impacts, and suitable for assessing the pollutant exposure of the general population Thus, the results of this survey should be useful in estimating public exposure to the HAPs Just as important, the survey has identified significant gaps in our knowledge of the ambient levels of several HAPs Filling these gaps should be given a high priority, so that the public health risks from these chemicals can

be evaluated with a satisfactory degree of certainty.

For the purposes of this survey, the 188 diverse chemicals designated as HAPs were organized according to the chemical classes and volatility classes identified in Chapter 2 This classification was useful because, as shown in Chapter 3, similar chemicals are frequently measured together, using similar measurement methods The survey was conducted in two stages In the first, infor- mation on ambient concentrations of the 188 HAPs was located through keyword searches of appropriate computerized databases, in review articles, reference books, proceedings of relevant air-quality conferences, and in unpublished datasets from urban air monitoring studies The results

of that first stage survey have been reported.4,5Ambient concentrations for 70 of the 188 HAPs were compiled through 1987 in the National VOC Data Base1,2, which was updated concurrently with the first stage of this survey.3 For this survey, the ambient data in the 1988 version of the national database1,2 were summarized and supplemented with ambient data from other measurement programs The search strategy for the

118 HAPs not included in the National VOC Data Base differed somewhat from the 70 included Those 118 chemicals were the subject of computerized and manual searches of the literature to locate ambient data For each chemical, a keyword search was conducted through the computerized databases of STN International (Columbus, OH) The databases searched included the Chemical Abstracts (CA) files from 1967 to 1993, Chemical Abstracts Previews (CAP) current files, and National Technical Information Service (NTIS) files from 1964 to 1992 To focus on data pertinent4

to toxics exposure of the U.S population, the search was restricted to English-language citations authored in the United States The strategy used both abstract and basic index searches to increase the likelihood of finding relevant citations.

Master sets of literature citations were set up in each of the STN files searched These master sets were then combined with the chemical names and CAS registry numbers of the compounds, to produce citation listings specific to each HAP If the initial reviews indicated information of value, the listed citations were then reviewed by title, abstract, and in their entirety For all of the 188 HAPs, data were obtained from published reviews, reference texts, and from proceedings of meetings such as the annual EPA/Air and Waste Management Asso- ciation (AWMA) annual symposium on toxic air pollutants By contacting the respective lead scientists, recent data were also obtained from unpublished field studies

The list of 188 HAPs includes some redundant entries in the form of chemical groups (e.g., xylenes, cresols) and their individual constituent isomers These chemicals may be used in industrial settings as the mixed isomers, but are generally measured in the atmosphere as individual isomers Searches were performed for both the individual and mixed isomers, but ambient data were found primarily for the individual isomers The HAP denoted as polycyclic organic matter (POM) com- prises numerous individual compounds, and the compounds measured are not always clearly defined

in reports of ambient measurements For consistency, and to emphasize potential health risks from POM, this survey focused on eight individual POM compounds identified as possible or probable human carcinogens.6,7 Those eight compounds are benzo[a]pyrene, benzo[a]anthracene, dibenzo[a,h]anthracene, chrysene, benzo[b]fluoranthene, benzo[k]fluoranthene, indeno[1,2,3,c- d]pyrene, and benzo[g,h,i]perylene Ambient data were compiled for the sum of these eight POM compounds.

The second stage of this survey of ambient HAP concentrations relied upon a recently assembled database of ambient monitoring data.8 Developed for the EPA’s Office of Air Quality Planning and Standards (OAQPS), that database encompasses a larger number of chemical species, studies, and measurements than does the National VOC Database.1–3 Although there is no federal mandate to

do so, numerous state and local agencies conduct sampling programs for toxic air pollutants, including the 188 HAPs To identify these sources of ambient data, EPA directed a search through

a number of different means that included professional organizations such as the State and Territorial Air Pollutant Program Administrators (STAPPA), the Association of Local Air Pollution Control Officers (ALAPCO), and the AWMA, as well as Internet information provided on state environ- mental agencies and other contact referrals Once these agencies were identified, cooperative sources contributed suitable ambient monitoring data to OAQPS A database of these ambient concentration measurements was then compiled.8

To produce the most complete and comprehensive archive of ambient measurements, the data obtained from state and local monitoring efforts were combined with similar data from the Aero- metric Information Retrieval System (AIRS) Administered by the OAQPS, AIRS is a computer- based repository of U.S air pollution data Data contributed to AIRS are largely the result of regulatory monitoring of criteria pollutants by state and local agencies However, depending on the attainment status of a region, some non-criteria pollutant monitoring is mandatory, as is submission

of the data to AIRS All collected data are merged into the archive on a regular basis; the latest update occurred in the fall of 2000

The ambient data archive8 was made available for this survey by staff of Battelle’s Statistics and Data Analysis Systems department, who are assembling a web-based, readily accessible version

of the database The database contains several sub-databases that link to one another by one or more common fields, eliminating redundant data and making the database smaller and easier to work with Those sub-databases include information on the sampling program, sampling site, measurement method, pollutant identification, ambient pollutant concentration, and detection limit

The intent of this ambient concentration survey was not to catalog every data point or sample Rather, the aim was to compile information on typical concentrations (i.e., mean and/or median), the range of concentrations observed, and the number, locations, and time periods of the measure- ments The purpose of this approach was to provide concentration data suitable for estimating population exposures to the 188 HAPs In general, the scarcer the ambient data for a given HAP, the greater the effort spent to find such data Additional information such as the detection limit of the measurements and the number of results below the detection limit was also recorded when available

In keeping with the aim of providing data for health risk assessment, the focus of this survey was

on ambient data in populated (urban to rural) areas of the U.S To that end, effort was made to exclude data from remote sites, and data indicating strong, direct, local source contributions In some cases, such exclusion was called for by clear identification of the origin of the samples However, in many cases, identification was ambiguous and, in the absence of clear information, elevated concentration results were generally retained in the dataset There was no attempt to exclude measurements that may have been subject to some impact of local urban sources, because those data properly represent the upper range of concentrations to which urban residents may be exposed.

Ambient air concentrations of hazardous air pollutants are compiled in Table 4.1 (see Appendix following Chapter 4), which lists all 188 HAPs in the same order as in Title III of the CAA, with alternate names as well, if they were stated in the HAPs list Table 4.1 gives the name and CAS number for each compound; the locations and years of measurements; the number of ambient measurements (N); the mean, range, and median (if available) of the measured data; the number

of the pertinent reference in the associated reference list that follows Table 4.1 ; and additional comments on the data, such as the number of non-detects included in the reported data The concentration units for each HAP are indicated in the first line of the concentration data All concentrations are in mass per volume units, i.e., micrograms per cubic meter (µg/m3), nanograms per cubic meter (ng/m3), or picograms per cubic meter (pg/m3) As noted in Chapter 3, mass per volume concentration units can be readily converted into mixing ratios at assumed atmospheric conditions For example, at 20º C and one atmosphere pressure, the conversion between µg/m3concentrations and part-per-billion by volume (ppbv) mixing ratios is

or

1 ppbv = 0.0416 · MW µ g/m3

where MW is the molecular weight of the HAP in question.

The Year column in Table 4.1 indicates the period of data collection for each data source Note that, in some cases, the number of locations and number of samples were not evident from the literature In those cases, the numbers were estimated, or lower limits are shown Inconsistency was also found in the treatment of measurements below the detection limit Some studies failed

to state the detection limit, or to define the number of measurements below that limit The value assigned to non-detects (e.g., zero, half the detection limit, etc.) in calculating a mean value was also not always clearly stated Whenever possible, these inconsistencies were addressed by inferring or estimating the detection limits and number of non-detects from information in the

1 0.0416 MW ⋅ -

literature Mean values were calculated using a value of one half the estimated detection limit for the results that were listed as non-detects

The most noticeable feature of the data in Table 4.1 is the extremely wide variation in the amount of data found for individual HAPs The number of sampling locations for individual HAPs varies from zero to more than 900 sites, and the number of measurements varies from zero to more than 470,000 Of particular importance is that the number of samples is zero for

60 of the HAPs, i.e., no ambient concentration data were found These features of the HAPs data are presented in more detail in Figures 4.1 and 4.2 for the 186 HAPs for which the number

of sampling locations and number of measurements could be established (For fine mineral fibers and radionuclides, ambient concentration estimates were drawn from sources that did not allow specification of the numbers of sampling sites and measurements.) Figure 4.1 shows the frequency distribution of the HAPs by number of sampling locations The greatest frequency

is found for zero sampling locations, with the 60 HAPs in this category composing nearly one third of the HAPs list The second-largest frequency in Figure 4.1 is for 1–10 sampling locations, again indicating the scarcity of data for some HAPs Only 86 chemicals (46% of the list) show data from more than 10 locations, and only 49 (26%) show data from 50 or more locations.

Figure 4.2 shows the corresponding frequency distribution by number of measurements found, and clearly indicates the wide range in the availability of ambient data for the HAPs The 60 chemicals for which no ambient data were found constitute the largest frequency range in Figure 4.2 For a total of 83 chemicals (44% of the list), fewer than 100 measurements of each exist, and

a total of 106 chemicals (56% of the HAPs list) show fewer than 1,000 measurements each However, the second-largest frequency range includes the 36 chemicals for which between 10,000 and 100,000 measurements were found, and for nine HAPs, more than 100,000 measurements were found These observations illustrate the primary characteristic of the HAPs list from the CAAA:

it is a unique mix of some chemicals frequently measured in ambient air, and others rarely or never measured.

FIGURE 4.1 Distribution of the HAPs by number of ambient air sampling locations

010203040506070

as EPA Compendium Method TO-14.9 In contrast, no data are available for most of the HAPs in the nitrogenated organic category, and for one third of the HAPs in the oxygenated organic category This fact is particularly important because together, these two groups make up nearly half of the HAPs list (88 total HAPs).

Several reasons may exist for the scarcity of ambient measurements within some chemical categories For the nitrogenated and oxygenated organics, which collectively fall under the definition

of polar VOCs, the most likely reason is the lack of sampling and analysis methods for these compounds Due to their water solubility and reactivity, measurement of these chemicals at likely ambient levels of a few µg/m3 or less (ppbv to sub-ppbv mixing ratios) is more difficult than measurement of VOCs, and methods for many of these chemicals are still in development (see Chapter 3) That this development is occurring is confirmed by the substantially improved state of ambient data shown in Figure 4.3 for the nitrogenated and oxygenated organics, relative to the state

at the time of the initial ambient concentration survey.4,5 For example, Figure 4.3 shows that, of the 49 nitrogenated organics, 34 have no ambient concentration data, whereas, in the initial survey,4,5

39 of the nitrogenated organics had no ambient data The corresponding results for the 39 ated organics (caprolactam was dropped from the HAPs list since the initial survey) are 13 compounds with no data at this time, compared with 21 with no data in the initial survey.4,5 For comparison, no change occurred in the number of HAPs with no ambient data in the chemical categories of hydrocarbons, halogenated hydrocarbons, halogenated aromatics, inorganics, pesti-

oxygen-FIGURE 4.2 Distribution of the HAPs by number of ambient concentration measurements

0 10 20 30 40 50 60 70

cides, phthalates, and sulfates The category of aromatics (18 total HAPs) showed an improvement from four HAPs with no ambient data4,5 to two with no ambient data at this time It can be concluded that development and validation of measurement methods for polar volatile compounds in air is particularly needed before risk assessment and regulation of many of these HAPs can be adequately accomplished

For other chemical categories, the scarcity of data may have other causes Pesticide ments, for example, are generally made in agricultural areas in association with application of these chemicals Such measurements are not directly relevant to the exposure of the urban U.S population and are not included in the tabulated data Many other chemicals have been measured in the workplace but not in ambient air For example, the list designates titanium tetrachloride, elemental phosphorus, and dye intermediates such as 3,3 ′ -dimethoxybenzidine, as HAPs Although the poten- tial toxicity of these chemicals has been established, their ambient concentrations have not been measured because they have been considered unlikely to be present at significant concentrations except near very localized sources For such compounds, assessment of the potential for human exposure might best be focused in areas around known sources.

measure-Another reason for the lack of ambient air data for some HAPs is the ambiguous nature of the identification on the CAA list A good example is coke oven emissions The emission of a variety

of toxic chemicals, including sulfur compounds, benzene, other aromatics, and polycyclic aromatic compounds, from coke ovens is well documented However, it is impossible to quantify those compounds originating in ambient air from coke oven emissions in the face of other sources of the same compounds, without, for example, detailed source apportionment modeling in the area of a coke oven source As a result, measurements of coke oven emissions as a chemical group in urban areas simply do not exist

The representativeness of the HAPs data for use in health risk assessments is an important issue Clearly, both the number of measurements and the number of locations in which measurements have been made are important in this regard ( Table 4.1 ) Some HAPs, such as many of the chlorinated and aromatic hydrocarbons, have been measured tens of thousands of

FIGURE 4.3 Number of HAPs, including the number with no ambient data, for each chemical category

Sulfates Phthalates Inorganics Pesticides Oxygenated organics

27 1

18 2

8 1

49 34

39 13

15 2

23 6

4 2 1 0

times in hundreds of locations The geographic spread of these data is also wide, because of the large number of studies that have included these chemicals Thus, it can be argued that sufficient data exist to estimate typical and elevated human exposures to these chemicals However, as noted above, nearly half of the 188 HAPs have been measured fewer than 100 times, and more than half have been measured in fewer than 10 locations Such small datasets and limited geographic coverage are unlikely to represent adequately the exposure of the U.S population to those chemicals For many of the HAPs, therefore, the representativeness of the existing data is very limited More measurements of these compounds are needed for adequate health risk assessment

0.5-10.10.005-0.75

4.5 RECENT DATA FOR HIGH PRIORITY HAPs

As a final example of the ambient HAPs concentration data, Table 4.2 summarizes selected ambient data for the group of 33 high priority HAPs identified in Chapter 1 The data shown in Table 4.2 are a subset of the complete datasets compiled for these chemicals and shown in Table 4.1 Shown

in Table 4.2 are the number of study locations, number of samples, mean, range, and years of recent measurements for the 33 high priority HAPs.

The availability of data for the 33 HAPs in Table 4.2 is generally better than for the 188 HAPs

as a whole For most of these HAPs, substantial numbers of recent samples are indicated Exceptions are coke oven emissions and hydrazine, for which no ambient data exist, and ethylene oxide and quinoline, for which only a few ambient measurements were found Data are also relatively scarce for 1,3-dichloropropene, 2,3,7,8-TCDD, hexachlorobenzene, particulate mercury compounds, and polycyclic organic matter Recent ambient measurements of polychlorinated biphenyls are also scarce With these exceptions, Table 4.2 shows that ambient data exist with which to estimate population exposures for the majority of the 33 high priority HAPs Inspection of the full dataset ( Table 4.1 ) also suggests that the recent data in Table 4.2 exhibit means and ranges that are generally lower than those of earlier data This difference may indicate decreases in the emissions of these chemicals However, changes in the choice of sampling locations might also account for this difference Site selection in early urban field studies often emphasized worst-case locations such

as urban traffic centers, whereas recent studies have tended to emphasize sites that are more representative of local population distributions As a result, the recent data shown in Table 4.2 may

be useful for initial human exposure assessments for these 33 HAPs

4.6 SUMMARY

This chapter has presented an updated assessment of the ambient concentration data available for the 188 HAPs The primary observation to be made is that the HAPs list includes a large group of chemicals that have rarely or never been measured in ambient air, and another group that has been measured very frequently For example, 60 of the 188 HAPs (nearly one third of the list) have no ambient concentration data, and 83 HAPs (44% of the list) have been measured fewer than 100 times On the other hand, 45 HAPs (24% of the listed HAPs) have been determined more than 10,000 times each in ambient air These results indicate that the representativeness of existing ambient data for estimating population exposures to HAPs will vary widely among different HAPs.

A somewhat more optimistic picture emerges when considering the 33 HAPs considered to cause the greatest public health risks in urban areas Substantial numbers of ambient data are available for most of the 33 high priority HAPs, and thus, for most of these 33 HAPs, estimates

of population exposures should be feasible.

6 Menzie, C.A., Potocki, B.B, and Santodonato, J., Exposure to carcinogenic PAHs in the environment,

Environ Sci Technol., 26, 1278, 1992

7 IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans: PolynuclearAromatic Hydrocarbons, Part 1, Chemical, Environmental, and Experimental Agency for Research

on Cancer, World Health Organization, 1983

8 Rosenbaum, A.S., Stiefer, P.S and Iwamiya, R.K., Air Toxics Data Archive and AIRS Combined DataSet: Data Base Descriptions, prepared for U.S EPA, Office of Policy, Planning, and Evaluation, bySystems Applications International, SYSAPP-99/25, 1999

9 McClenny, W.A et al., Canister-based method for monitoring toxic VOCs in ambient air, J Air Waste Mgt Assoc., 41, 1308, 1991

14 U.S urban sites

1980-00 1974-84

1990-91 1989 1989

14,143

384

56

332 406

2.446 µg/m 3

6.09 (Median 0.87) 3.2

2.32 2.51 (Median 2.21)

< 0.004-102 µg/m 3

up to 105

< 0.2-16.7 0.37-16.7 0.68-13.9 (Site means 1.72-3.40)

28 71

21

33 78

Includes 735 nondetects Includes 189 nondetects

44

4 8

0.84 µg/m3

0.05 0.84

All nondetects Includes 2 nondetects All nondetectsAcetophenone

98-86-2

3 U.S locations

2 U.S urban locations

19931977-78

3 3

1990-00 1981

1990-91 1990-91 1990-91

2,926 36

22 22 8

0.273 µg/m 3

0.66 (Median 0.22) 1.1

1.1 1.1

20

20 21

Includes 2,714 nondetects Includes 4 nondetects

All nondetects All nondetects All nondetects

Allyl chloride

(3-chloro-1-propene)

107-05-1

32 U.S locations Lima, OH

5 U.S cities

1988-981990-911980-81

Includes 1,146 nondetectsAll nondetects

1,023 4

All nondetects Includes 3 nondetects

48 U.S cities

1970s1970sca.1980

ca.1970

– –

31

187

– (Median 20) (Median 0.9)

(Median 1.6)

1–100 ng/m3

(roughly 30-3000 fibers/m3)1–70

Up to 9.8 (90th %) (roughly up to 330 fibers/m3)

Up to 6.8 (90th %) (roughly up to 230 fibers/m3)

54

55

56

57

11 U.S cities

Lima, OH

1980-00 1973-87

298 4,620

349

81

2.147 µg/m 3

10.1 (Median 6.5) 1.56 2.68 (Median 1.92) 4.77 (Median 3.3) 2.56

1988-99

1980 19891990-91

8,174

42

298 81

0.249 µg/m3

0.05 (Median 0.05) 1.84 0.27

Includes 7,716 nondetectsIncludes 2 nondetectsIncludes 119 nondetects All nondetects

Biphenyl

92-52-4

16 U.S locations Columbia, SC

1993-971989

811 2

Bis (2-ethylhexyl)

phthalate (DEHP)

117-81-7

11 U.S locations New York City, NY College Station, TX Gulf Coast, TX Portland, OR

1992-9719781979-8019821984-85

984

?

14

? 10

5 ng/m3

– 1.99 0.62 0.39 (gas-phase) 0.48 (particle-phase)

< 1-310 ng/m3

10.2-16.8 0.77-3.60 – – –

28

15

1

9 32

Includes 836 nondetects

3 urban sites Rural site Number of samples estimatedBis (chloromethyl) ether

26 397

Includes 17 nondetectsIncludes 395 nondetects

0.8 ng/m3

1.3 3.1 2.7 2.3

Summer samples; All nondetects Spring samples; All nondetects Winter samples; All nondetects Spring samples; All nondetects Winter samples; All nondetects Detection limits estimated

436 15

10.7 µg/m3

0.3 (Median 0.14)

< 0.003-654 µg/m3

0.05-1.07

28 71Includes 242 nondetects

1989 1990

1990-91

39,404 5,011

Includes 20,811 nondetects Includes 1,014 nondetects Includes 3 nondetects Includes 5 nondetects

Springfield/Chicopee, MA

1973-741979-8019801977-791977-7919821987-88

0.1 ng/m3

0.1 1.05 0.063 1.30 0.15 0.036 38.4 9.5 27.4 3.1 2.0

up to 0.9 ng/m3 (α-chlordane)

up to 1.4 (γ-chlordane) 0.32-2.64

– – – –

13 non-detected samples (each isomer); urban site Rural sites

Downtown, winter

Summer samples Spring samples Winter samples Spring samples Winter samplesChlorine

1981-001976-86

1989 19901990-91

44,3551,012298

349 81

0.68 µg/m3

2.15 (Median 0.84) 0.07 0.14 0.24

Includes 38,893 nondetectsIncludes 154 nondetects All nondetects Includes 300 nondetectsAll nondetectsChlorobenzilate

510-15-6

1981-00 1973-87

1989 1990 1990-91

57,258 3,640

298

349 81

0.81 µg/m 3

2.68 (Median 0.27) 0.16 0.55 0.25

Includes 41,268 nondetects Includes 859 nondetects

Includes 250 nondetects Includes 332 nondetects All nondetects

5,645

1 349

0.44 µg/m3

2.2 0.29

Includes 4,883 nondetectsIncludes 261 nondetectsCresol/cresyllic acid

1992-971984

1,015 7

1 3

106-44-5

10 U.S locations

11 urban sites, CA

1992-971985

1,01262

0.014 µg/m3

4.6 2.1-8.6 (range of site means)

< 0.001-1.721 µg/m3

0.5-19.9

28 19Includes 964 nondetects

0.139 µg/m1.3 (Median 0.25) 0.34 (Median 0.18)

Includes 262,689 nondetectsIncludes 305 nondetects

2,4-D, salts and esters Baltimore, MD, Fresno, CA,

Riverside, CA, SL City, UT

Jordan, NY Rome, NY Salt Lake City, UT Jacksonville, FL

Springfield/Chicopee, MA

1967-68

1980

1980 19801987-88

–

1.15 ng/m3

1.54 4.0 0.27 7.5 5.5

12 7

up to 4.0 ng/m3

– – –

Includes 436 nondetects;

detected in 1 sample at SL City, UT; detection limit unknown; urban sitesOne-yr study, 16 U.S cities,

3 samples contained 2,4-D

Summer samples; Includes

46 nondetectsSpring samples; all nondetectsWinter samples; all nondetectsSpring samples; all nondetectsWinter samples; all nondetectsDetection limits estimated

Springfield/Chicopee, MA

1992-971967-68

1967-68

1978 19801979-801987-88

4.5 0.093 0.021 0.26 –

up to 14.2– – 0.04-0.66 –

–

–

28 7,82

7,82

8

8

1 11

11

11

11

All nondetects Includes 424 nondetects (4,4′); detection limits unknown; urban sitesIncludes 291 undetects (4,4′); Includes 393 nondetects (2,4′); detection limits unknown, rural sites

Rural site Spring samples Winter samples; all nondetects; detection limit unknown

Spring samples; all nondetects; detection limit unknown

Winter samples; all nondetects; detection limit unknown

1,012

? 13

?

3 ng/m3

–1.400.42

< 2-101 ng/m3

3.3-5.7 0.48-3.60 –

28

15

1 9

Includes 912 nondetects

3 urban sites Rural site

1981-001976-86

1989 19901990-91

31,805

719

298

349 81

0.676 µg/m3

4.16 (Median 0.55) 0.09 1.04 0.31

Includes 20,306 nondetectsIncludes 136 nondetects All nondetects Includes 219 nondetects All nondetects

Lima, OH

1993-94 1990

trans: includes 335 nondetects; cis: includes

339 nondetects Sum of cis- and trans- isomers All nondetects

49 50

0.75 ng/m3

2.0 3.2 2.8 2.0

Summer samples; all nondetects Spring samples; all nondetects Winter samples Spring samples; all nondetects Winter samples; all nondetects Detection limits estimatedDiethanolamine

< 0.02-13.8 µg/m–

23 23

Residential areas around waste site (unsettled wind)Upwind of waste site1,1-Dimethylhydrazine

1992-971985-861987

1,000

3 3

0.006 µg/m3

0.03 0.03

Includes 954 nondetects All nondetects All nondetectsDimethyl sulfate

59 533

Includes 37 nondetectsIncludes 346 nondetects

11 U.S cities Lima, OH

1973-001968-87

1989 199019901990-91

463,8213375298462034981

1.019 µg/m3

7.9 (Median 2.5) 1.10 2.04 (Median 1.30) 1.68 (Median 1.06) 1.1

< 0.002-3874 µg/m3

up to 1,248 0.13-5.56

Includes 96,608 nondetectsIncludes 182 nondetects

11 U.S cities

1983-991980-8619891990-911990

13,513

180

298

81 349

0.246 µg/m3

55.7 (Median 0.17) 0.04 0.24 0.15

Includes 12,402 nondetectsIncludes 28 nondetects All nondetects Includes 79 nondetectsIncludes 342 nondetects

1990-91

27,269 2,120

All nondetects

1981-00 1976-87

1989 1990 1990-91

49,908 2019

298

349 81

1.340 µg/m 1.61 (Median 0.04) 0.16 0.08 0.21

Includes 43,496 nondetects Includes 883 nondetects

Includes 250 nondetects Includes 348 nondetects All nondetects

Urban site; Also reported:

remote marine air: 0.046; and urban near- source: >1.8

1987-001976-87

1989 19901990-91

30,460644

298

349 81

0.228 µg/m3

0.170.06

< 0.16 0.21

Includes 29,757 nondetectsIncludes 437 nondetects All nondetects Includes 335 nondetects All nondetects

1989 1989

51,801

554

332 416

6.387 µg/m 3

8.93 (Median 4.71) 3.74 2.64 (Median 2.22)

< 0.003-436 µg/m 3

up to 87.1

0.12-19.2 0.53-11.0 (Site means 1.75-4.72)

28

71

33 78 Includes 1,005 nondetects Includes 75 nondetects