GSA Annual Meeting in Denver, Colorado, USA - 2016

Paper No. 299-14
Presentation Time: 4:50 PM


WENDT, Anna K., Department of Geosciences, The Pennsylvania State University, 318B EES Building, University Park, PA 16802, SOWERS, Todd, Department of Geosciences, Penn State University, University Park, PA 16802, HYNEK, Scott A., Geosciences, Penn State University, 302 Hosler Building, University Park, PA 16802, LEMON, Jacob, Trout Unlimited, Lock Haven, PA 17745 and BRANTLEY, Susan L., Earth and Environmental Systems Institute and Department of Geosciences, Pennsylvania State University, University Park, PA 16802,

In Pennsylvania in the area of high gas-well density, the most commonly cited water quality problem related to the gas industry is leakage of methane into groundwater. Methane also leaks from old and abandoned oil and gas wells into groundwater in this area. Finding such leakage is difficult due to the time and expense of sampling individual locations for groundwater. Also, methane naturally can seep into waters. Where groundwater discharges into streams, however, streams collect methane from both natural and anthropogenically-caused upward gas migration. To test if streams can be used to find leakage, we collected a set of calibration stream waters to determine the range of methane concentrations, [CH4], in different types of sites. We then sampled new sites to determine if our calibration groupings identified any leakage areas. To amplify the spatial sampling reach, we collaborated with citizen-scientist groups. Stream [CH4] in 149 sites in four counties in Pennsylvania defined skewed distributions with high numbers of low values. The median of the distribution of [CH4], 1.00, is therefore a better descriptor of the data population than the mean and standard deviation, 22.54 ± 288 ppb (n = 254). [CH4] ranges from 0.02 to 4600 ppb. For sites located within 30 meters of a wetland [CH4] equaled 3.08 ± 8.8 ppb (n = 67; range = 0.05 to 68.8 ppb). For sites within 762 meters (2500 feet) of a leaking unconventional gas well, concentrations equaled 142 ± 753 ppb (n = 37; range = 1.30 to 4600 ppb. For sites within 1800 meters downstream of a lineament known to be naturally emitting thermogenic gas, we observed average [CH4] = 8.5 ± 38.7 ppb (n = 28; range = 0.13 to 206 ppb). For sites with no known source of biogenic or thermogenic gas, [CH4] equaled 1.00 ± 1.25 ppb (n = 123; range = 0.02 to 6.27 ppb). In comparison, water in equilibrium with CH4 in air (1.87 ppm) is 0.08 ppb. Using the categories defined in the calibration dataset, we analyzed a test dataset from across PA and discovered high methane values downstream of a landfill and an abandoned gas well. Stream sampling for methane is a straightforward way for communities to find such leakage and for scientists to probe fundamental controls of methane production and emission. Stream [CH4] above 10 ppb without any nearby wetland-type environment should be considered a possible site of gas well leakage.