outdoor air, drinking water and the breath of 355 persons in NJ, in the fall of 1981. The NJ ... residents of Bayonne and Elizabeth, New Jersey, between 3 ...
PERSONAL EXPOSURES, INDOOR-OUTDOOR RELATIONSHIPS, AND BREATH LEVELS OF TOXIC AIR POLLUTANTS MEASURED FOR 355 PERSONS IN NEW JERSEY LANCE A. WALLACE* U.S. ~~ronrnen~ Protection Agency
EDO D. PELLIWRI, TY D. HARTWELL,CHARLESM. SPARACINO, LINDA S. SHELDON and HARVEY ZELON Research Triangle Institute, Research Triangle Park, NC 27709, U.S.A. (First receiced 2 November 1984 and ~n~~l~?rn
28 Februury 193.5)
A&tract--EPA’sTEAM Study has measured exposures to 20 volatile organic compounds in personal air, outdoor air, drinking water and the breath of 355 persons in NJ, in the fall of 1981.The NJ residents were selected by a probability sampling scheme to represent 128,000 inhabitants of Elizabeth and Bayonne. Participants carried a personal monitor to collect two 12-h air samples and gavea breath sample at the end of
the day. Two consecutive 12-houtdoor air samples were also collected on identical Tenax cartridges in the back yards of 90 of the participants. About 3tXIO samples were coffected,of wtdi 1000were quality control samples. Eleven compounds were often present in air. Personal exposures were consistentIy higher than outdoor concentrations for these chemicals, and were sometimes ten times the outdoor concentrations. Indoor sources appeared responsible for much of the difference. Breath concentrations also usually exceed
outdoor concentrations, and correlated more strongly with personal exposures than with outdoor concentrations. Some activities (smoking, driving, visiting dry cleaners or service stations) and occupations (chemical, paint and plastics plants) were associated with significantly elevated exposures and breath levels for certain toxic chemicals. Key word index: Exposure, indoor air, toxics, volatile organic compounds, drinking water, TEAM Study. ratio, consumer products, building materials, organic emissions.
Tenax, personal monitors, exhaled breath, occupational exposure, expired air, indoor-outdoor
INTRODUCTIOS The TEAM (Total Exposure Assessment Methodology) Study was designed to develop and demonstrate methods to measure human exposure to toxic substances in air and drinking water. A first phase to fieldtest the methods was completed in 1981 (Pellizzari et al., 1982; Sparacino et at, 1982a,b; Entz et al., 1982; Wallace et al., 1982, 1984). Methods developed or demonstrated in Phase I included a personal monitor employing Tenax cartridges; a spirometer for collecting expired air on Tenax cartridges; and a statistical design with field-tested questionnaires for the present study. Preliminary results from Phase I, which included pilot studies of 11 students at Lamar University in Beaumont, Texas, six students at the University of North Carolina, and 12 persons in New Jersey and North Carolina included: (1) About a dozen vofatite organic compounds were
*Presently a Visiting Scientist at Harvard University, School of Public Health, 665 Huntington Avenue, Boston, MA 02115, U.S.A. M
IY
d-G
often or always found in personal air samples of all subjects. (2) Exposures were highly variable, ranging from 1 to 1000 pgrn- 3 for some chemicals. (3) Breath values were correlated with personal exposures for several chemicals. The objective of the second phase (Pellizztri et of., 1984) was to estimate the distribution of exposures to target substances for the entire population of an industrial/chemical manufacturing area. A total of 20 toxic, carcinogenic, or mutagenic organic compounds were measured in the air and drinking water of 350 residents of Bayonne and Elizabeth, New Jersey, between 3 September and 23 November, 1981. The participants were selected from over 10,000 residents screened by a probability sampling technique to represent 128,000 persons (over the age of seven) who lived in the two neighboring cities, which included extensive chemical manufacturing and petrochemical refining activities. One hundred geographic areas throughout the two cities were selected for monitoring. Each participant carried a personal sampler with him during his normal daily activities for two consecutive 12-h periods. (One
1651
resident in each of the IO0 areas had an identical sampler operating in the back yard for the same two 12-h periods.) All participants also collected two drinking water samples. At the end of the 74-h sampling period, al1 participants gave a sample of exhaled breath, which was analyzed for the same compounds. Ail participants also completed a questionnaire on their occupations and activities during the sampling period. An extensive quality assurance program was carried out on all samplingjanalysis activities. (Return visits were made in the summer of 1982 and the winter of 1982-83. Subsets of the original group of 355 persons were sampled to determine seasonal differences in exposures. These results will be discussed in future papers.)
graphic column used to trap compounds of Interest (Sparadno 61 ai.. 19S41.The target compounds NC:: then anal!zcd bq flame ionization (aromatics1 Ind A Hall sistrolqtic cnnduc. tivity detector (halo-carbonsi
Blank samples and control samples spiked with all 20 target compounds were kept at the laboratory and shipped to the fieid to determine background contamination levels. recovery efficiencies and effects of transportation and storage. Duplicate air. water and breath samples were collected and analyzed at the primary laboratory (Research Triangle Institute) and QA laboratory (11-T Research Institute) to determine intralaboratory and interla~ratory precision. Deuterated benzene was load& on all duplicate cartridges to determine unambiguously the extent of losses during sampling operations. Periodic audits were carried out by EPA’s Environmental Monitoring Systems Laboratory at Research Triangle Park (EMSL-RTPI.
RESULTS Air
Personal and outdoor air samples were collected on Tenax cartridges for 12-h periods. A Dupont pump pulled air at 30mlmin-’ (4 22 / sampling volume) across the 1Scm i.d. cartridge, which contained 6cm ( - tg) of @/@mesh purified Tenax. Cartridges were analyzed by thermal desorption and cryofocussing of the organic vapors (Krost rt af.. 1982). followed by capillary gas chromatographyjmass spectrometry/computer analysis (GC/MS/COMP). Breath Breath samples were coilected by a specially-designed spirometer consisting of a humjdi~ed supply of pure air, a 40f Tedlar bag lo collect the pure air, a two-way Douglas mouthpiece, a second 40-t Tedlar bag to collect expired air, and two Nutech pumps to pull the expired air across two Tenax cartridges. After the first bag is partially filled with pure air, the subject uses the two-way mouthpiece to inhale from the bag and exhale into the second bag. The pumps pull the expired air across the two Tenax cartridges, which are then stored at -20°C. Analysis was by G~jMS/COMP. Background contamination of the bags is reduced to acceptable levels by flushing with helium at least ten times over a period of days before use. Waler Drinking water samples were collected in the morning and evening from the kitchen tap after a 20-s run in 40-mf Teflon-capped amber glass vials containing 5mg sodium thiosulfate. For analysis, a purge-and-trap technique was used (B&r and Lichtenberg, 1974). Purgable organics were swept onto a Tenax cartridge from a specially-designed allglass 2S-ml purge device connected to a short gas chromato-
About 4400 of the 5200 target households were contacted and provided information on 11,414 household residents. These data were employed in the second stage to select a sample of participants, The sample was weighted to overrepresent certain high potential exposure groups. About 58 O,,of the eligible residents (all persons 7 or older nor living in group quarters) in each city agreed to participate fully in the study (Table 1). Limited follow-up studies on nonrespondents showed no outstanding differences from respondents. About 1950 air, breath and water samples were collected and chemically analyzed {Table 2). An additional 980 quality control samples (duplicates. spikes and blanks) were analyzed. Quality assurance results
Results from 1.55blank cartridges (Table 3) showed low backgrounds (corresponding to < 2 pg m- ‘1 benzene (5 f 3pgm--‘). (Mean for except backgrounds for each batch of Tenax cartridges were subtracted from the measured amounts on field cartridges from that batch.) The results from 201 spiked control cartridges showed recovery eEciencies ranging from 85 to 1IOy,0 (Table 3). The deuterated benzene results showed consistently acceptable losses of j-1 5 4, Results from 134 pairs of duplicate personal air samples and 34 duplicate outdoor air samples analyzed
Table 1. Results of two-stage probability sampling (TEAM Study. Fall. 1981)
Stage I: Screening Households screened Households completing questionnaire Persons providing data Stage II: Monitoring Eliaible oersons Pe&ons completing data collection
Bayonne
Elizabeth
2063 1788@7”,)
3145 2638(81”,)
4687
4727
266
345
ls4t57.9 ?J
201(%.3 “,)
Personal exposures, indoor+utdoor
1653
relationships, and breath levels of toxic air pollutants
Table 2. Samples collected (TEAM Study, Fall, 1981) Personal air
Outdoor air
Breath
Drinking water
705 131 283 II19
183 32 90 305
358 37 185 580
718 70 I52 950
Field samples Duplicates Blankscontrols Total
Table 3. Blank and control data for air and breath volatiles Blanks (&cartridge k SD.) Field Lab. (N = 76) (N = 79)
Compound Vinylidene chloride Chloroform l,2-Dichloroethane I.l,l-Trichloroethane Benzene Carbon tetrachloride Trichloroethylene Bromodichloromethane Dibromochloromethane Tetrachloroethylene Chlorobenzene Bromoform Dibromochloropropane Styrene p-Dichlorobenzene Ethylbenzene o-Xylene p-Xylene o-Dichlorobenzene
l&2 22&20
