Human exposure to unconventional natural gas development: A public health demonstration of periodic high exposure to chemical mixtures in ambient air

Abstract

Directional drilling and hydraulic fracturing of shale gas and oil bring industrial activity into close proximity to residences, schools, daycare centers and places where people spend their time. Multiple gas production sources can be sited near residences. Health care providers evaluating patient health need to know the chemicals present, the emissions from different sites and the intensity and frequency of the exposures. This research describes a hypothetical case study designed to provide a basic model that demonstrates the direct effect of weather on exposure patterns of particulate matter smaller than 2.5 microns (PM_2.5) and volatile organic chemicals (VOCs). Because emissions from unconventional natural gas development (UNGD) sites are variable, a short term exposure profile is proposed that determines 6-hour assessments of emissions estimates, a time scale needed to assist physicians in the evaluation of individual exposures. The hypothetical case is based on observed conditions in shale gas development in Washington County, Pennsylvania, and on estimated emissions from facilities during gas development and production. An air exposure screening model was applied to determine the ambient concentration of VOCs and PM_2.5 at different 6-hour periods of the day and night. Hourly wind speed, wind direction and cloud cover data from Pittsburgh International Airport were used to calculate the expected exposures. Fourteen months of daily observations were modeled. Higher than yearly average source terms were used to predict health impacts at periods when emissions are high. The frequency and intensity of exposures to PM_2.5 and VOCs at a residence surrounded by three UNGD facilities was determined. The findings show that peak PM_2.5 and VOC exposures occurred 83 times over the course of 14 months of well development. Among the stages of well development, the drilling, flaring and finishing, and gas production stages produced higher intensity exposures than the hydraulic fracturing stage. Over one year, compressor station emissions created 118 peak exposure levels and a gas processing plant produced 99 peak exposures over one year. The screening model identified the periods during the day and the specific weather conditions when the highest potential exposures would occur. The periodicity of occurrence of extreme exposures is similar to the episodic nature of the health complaints reported in Washington County and in the literature. This study demonstrates the need to determine the aggregate quantitative impact on health when multiple facilities are placed near residences, schools, daycare centers and other locations where people are present. It shows that understanding the influence of air stability and wind direction is essential to exposure assessment at the residential level. The model can be applied to other emissions and similar sites. Profiles such as this will assist health providers in understanding the frequency and intensity of the human exposures when diagnosing and treating patients living near unconventional natural gas development.

Keywords:

• Diagnostic tools
• dispersion air model
• exposure patterns
• health impacts
• unconventional natural gas

Acknowledgments

The authors thank the residents in Washington County, Pennsylvania, who shared information with SWPA-EHP. We gratefully acknowledge the support and assistance of the staff at SWPA-EHP, Dr. Kathy Nolan and Earl Ivan White.

Funding

This research was supported by the Heinz Foundation and the Colcom Foundation.

Supplemental material

Supporting information for this paper can be viewed at environmentalhealthproject.org.

Notes

¹ The Southwest Pennsylvania Environmental Health Project is a nonprofit public health organization established to respond to individual and community needs for access to accurate health information and health services associated with UNGD. The southwest region of the state is among the fastest growing areas for this industry because it lies over the Marcellus shale deposits.

² Two hundred and fourteen of these residences were found to have between 1 and 77 UNGD well pads at a distance of 2–5 km. Eighty-five residences had from 1 to 17 well pads located between 1–2 km. Thirty-one homes had from 1 to 7 well pads within 1k km. Two hundred and sixty residences had between 1 and 5 compressor stations 2–5 km distant. Fifteen homes had 1–2 compressor stations within 1–2 km. Five residences had one to two compressor stations less than 1 km distant. Washington County currently has two processing stations.

³ The very first stage of well pad development, access road and well pad construction, is omitted from this case study, although there are public health implications for this stage because of truck traffic, diesel exhaust emissions and particulate matter (PM) effects on air quality.