Supplemental Material

Meteorological Conditions

Table S-1 contains a summary of the average and range of meteorological conditions monitored at the two sites. Generally, the conditions were similar at both sites and were expected based on typical southwestern Pennsylvania conditions. The historical average high and low temperatures in McDonald, PA (the location of the study) in November, December and January were -0.6 to 11, -5.0 to 4.4 and -8.9 to 2.2 °C, respectively (www.weather.com, accessed December 21, 2015). The historical average precipitation values were 0.26, 0.24 and 0.23 cm/day, respectively (www.weather.com, accessed December 21, 2015). Notably, the meteorological station at Site A was positioned at a height of approximately 26 ft (7.9 m) while the meteorological station at Site B was positioned at a height of approximately 5 ft (1.5 m). The logarithmic increase of wind speed with increasing height likely explains the disparity in wind speed values between the two sites.

Table S-1 Overview of 24-hour average meteorological conditions at Sites A and B

Site / Sampling Period / 24-hour Mean Value (Range)
Temperature (°C) / Humidity (%) / Barometric Pressure (mmHg) / Precipitation (cm/day) / Predominant Wind Direction / Wind Speed (m/s)
A / Baseline
(11/14/2011 – 11/29/2011) / 11
(0 – 22) / 77
(48 - 98) / 744
(744 - 762) / 0.48
(0.00 – 3.0) / WSW / 3.9
(0.85 - 7.3)
Fracturing
(11/30/2011 – 12/29/2011) / 3
(-3 - 12) / 80
(55 - 97) / 757
(744 - 767) / 0.28
(0.00 – 2.1) / WSW / 3.9
(1.3 - 9.8)
Flaring
(12/28/2011 – 1/6/2012) / 0
(-9 - 8) / 79
(56 - 91) / 752
(747 - 757) / 0.036
(0.00 - 0.25) / WSW / 5.9
(3.0 - 9.4)
Post Flaring
(1/7/2012 – 1/27/2012) / -1
(-9 - 8) / 82
(63 - 96) / 752
(737 - 762) / 0.22
(0.00 – 1.9) / WSW / 2.8
(1.3 - 8.4)
B / Baseline
(11/14/2011 – 11/29/2011) / 9
(0 - 17) / 76
(51 - 94) / 762
(752 - 772) / 0.56
(0.00 – 2.8) / SSW / 1.6
(0.31 - 3.4)
Fracturing
(11/30/2011 – 12/29/2011) / 2
(-4 - 12) / 82
(64 - 95) / 767
(752 - 777) / 0.28
(0.00 – 2.2) / W / 1.6
(0.31 - 4.3)
Flaring
(12/28/2011 – 1/6/2012) / 1
(-9 - 7) / 78
(57 - 90) / 759
(757 -767) / 0.064
(0.00 - 0.33) / W / 2.7
(1.5 - 4.4)
Post Flaring
(1/7/2012 – 1/27/2012) / 0
(-11 - 11) / 82
(64 - 96) / 762
(747 - 772) / 0.43
(0.00 - 2.0) / W / 2.2
(0.45 - 5.9)

Variation of TVOC Measurements with Time of Day

As explained in the body of the manuscript, in order understand potential time variation patterns in the data, the ppbRAE TVOC data were grouped by the hour of measurement and averaged for each sampling period. Figure S-1 presents the average hourly concentration of TVOCs at each site for each sampling period.

Fig. S-1 Total VOC concentration (measured by ppbRAE) average per hour for A) Site A and B) Site B.

Information Regarding Individual VOCs

Table S-2 contains a list of the individual compounds for which Summa canister 24-hour samples were analyzed in this study (EPA TO-15). Individual VOC reporting limits for these chemicals ranged between 0.13 and 0.84 ppb.

Table S-2 Individual chemical species included in analysis via USEPA Method TO-15

Chemical Analyte / Range of Reporting Limits (ppb) / Chemical Analyte / Range of Reporting Limits (ppb)
1,1,1-Trichloroethane / 0.10-0.17 / Chlorobenzene / 0.10-0.17
1,1,2,2-Tetrachloroethane / 0.10-0.17 / Chloroethane / 0.50-0.84
1,1,2-Trichloroethane / 0.10-0.17 / Chloroform / 0.10-0.17
1,1-Dichloroethane / 0.10-0.17 / Chloromethane* / 0.10-0.17
1,1-Dichloroethene / 0.10-0.17 / cis-1,2-Dichloroethene / 0.10-0.17
1,2,4-Trichlorobenzene / 0.50-0.84 / cis-1,3-Dichloropropene / 0.10-0.17
1,2,4-Trimethylbenzene / 0.10-0.17 / Cumene / 0.10-0.17
1,2-Dibromoethane / 0.10-0.17 / Cyclohexane* / 0.10-0.17
1,2-Dichlorobenzene / 0.10-0.17 / Dibromochloromethane / 0.10-0.17
1,2-Dichloroethane / 0.10-0.17 / Ethanol* / 0.50-0.84
1,2-Dichloropropane / 0.10-0.17 / Ethyl Benzene / 0.10-0.17
1,3,5-Trimethylbenzene / 0.10-0.17 / Freon 11* / 0.10-0.17
1,3-Butadiene / 0.10-0.17 / Freon 113 / 0.10-0.17
1,3-Dichlorobenzene / 0.10-0.17 / Freon 114 / 0.10-0.17
1,4-Dichlorobenzene / 0.10-0.17 / Freon 12* / 0.10-0.17
1,4-Dioxane * / 0.10-0.17 / Heptane* / 0.10-0.17
2,2,4-Trimethylpentane / 0.50-0.84 / Hexachlorobutadiene / 0.50-0.84
2-Butanone* / 0.50-0.84 / Hexane* / 0.10-0.17
2-Hexanone / 0.50-0.84 / m,p-Xylene* / 0.10-0.17
2-Propanol* / 0.50-0.84 / Methyl tert-butyl ether / 0.10-0.17
3-Chloropropene / 0.50-0.84 / Methylene Chloride* / 0.20-0.34
4-Ethyltoluene / 0.10-0.17 / o-Xylene / 0.10-0.17
4-Methyl-2-pentanone / 0.10-0.17 / Propylbenzene / 0.10-0.17
Acetone* / 0.50-0.84 / Styrene / 0.10-0.17
alpha-Chlorotoluene / 0.10-0.17 / Tetrachloroethene / 0.10-0.17
Benzene* / 0.10-0.17 / Tetrahydrofuran / 0.50-0.84
Bromodichloromethane / 0.10-0.17 / Toluene* / 0.10-0.17
Bromoform / 0.10-0.17 / trans-1,2-Dichloroethene / 0.10-0.17
Bromomethane / 0.10-0.17 / trans-1,3-Dichloropropene / 0.10-0.17
Carbon Disulfide / 0.50-0.84 / Trichloroethene / 0.10-0.17
Carbon Tetrachloride / 0.10-0.17 / Vinyl Chloride / 0.10-0.17

*Detected above the reporting limit one or more times.

Of these chemicals, 15 (as denoted by asterisks in Table S-2) were detected above the reporting limit one or more times. The RSL values listed in Table 5 were calculated using the USEPA RSL calculator available on the USEPA website (epa-prgs.ornl.gov, accessed December 24, 2015). The selections used as input to the calculator were as follows: scenario=resident, media=air, SL type=site specific (user-provided inputs), select risk output=yes, select RfD/RfCtype=subchronic. The individual chemicals selected were those marked by an asterisk in Table S-2. For m,p-xylene, p-xylene, m-xylene and xylenes were selected. Metadata were retrieved in order to determine the references for the RSL for each individual chemical.

Following the initial data entry, user-provided inputs for the chemical-specific toxicity values were automatically gathered by the program and not altered. The exposure duration for a resident was set at 1 year, and the exposure frequency was set at 90 days/year to reflect the duration of the present study. The exposure time was set at 24 hr/day. The target hazard quotient was set at 1.0, the target cancer risk was set at 1.0x10-6 and the resident lifetime was set at 70 years. Only one of the detected species (methylene chloride) was a suspected mutagen. The mutagenic parameters were set such that the exposure duration, frequency and time reflected 90 days of exposure in all exposure duration age ranges. Select results from the calculator are presented in Table S-3 along with the basis for each RSL and the subchronic target organ for non-carcinogens.

Table S-3 Summary of results from RSL calculation with RSL information

Volatile Organic Compound / Inhalation Unit Risk (m3/µg) / RfC (mg/m3) / Chronic/
Subchronic / RSL (ppb) / Basis for RSLb / Subchronic Target Organ
1,4-Dioxane / 5.00 x 10-6 / 0.72 / - / 15.8 / cancer / NA
2-Butanone / NA / 1 / subchronic / 1380 / non-cancer / Developmental
2-Propanol / NA / 7 / subchronic / 11600 / non-cancer / Neurological
Acetone / NA / 30.9 / chronic / 52600 / non-cancer / Hematopoietic
Benzene / 7.80 x 10-6 / 0.08 / - / 11.4 / cancer / NA
Chloromethane / NA / 3 / subchronic / 5910 / non-cancer / Neurological
Cyclohexane / NA / 18 / subchronic / 21200 / non-cancer / Developmental
Dichlorodifluoromethane / NA / 1 / subchronic / 821 / non-cancer / General
Ethanol / NA / - / - / - / - / -
Heptane / NA / - / - / - / - / -
Hexane / NA / 2 / subchronic / 2300 / non-cancer / Neurological
Methylene Chloride / 1.00 x 10-8 / 1.04 / - / 113 / cancer / NA
Toluene / NA / 5 / chronic / 5390 / non-cancer / Neurological
Trichlorofluoromethane / NA / 1 / subchronic / 723 / non-cancer / Neurological
m,p-Xylenec / NA / 0.1 / subchronic / 93.5 / non-cancer / Neurological

NA = not applicable (non-carcinogens do not have Inhalation Unit Risk Values, carcinogens do not have non-cancer target organs).

a Subchronic RfCs were preferable in this assessment. Chronic RfCs were used if no subchronic value was available (epa-prgs.ornl.gov, accessed December 24, 2015).

b Two RSLs are calculated for carcinogens: one based on cancer and one based on non-cancer effects. The lower RSL of the two is used (epa-prgs.ornl.gov, accessed December 24, 2015).

c Represented by either p- or m-xylene, which have identical values for RfD, RfC and RSL.

3