Reports
Dissolved-methane concentrations in well water in the Appalachian Plateau Physiographic Province of Maryland
2016, Bolton, D.W., Pham, Minh Phung T., and Drummond, D.D.
Report of Investigations 82
Abstract
Dissolved-methane concentrations were measured in 87 drinking-water wells in the Appalachian Plateau Physiographic Province of Maryland between 2012 and 2014. The objectives of the study were to measure baseline dissolved-methane concentrations in well water in the region where natural gas from the Marcellus may eventually be produced, and (to the extent possible) evaluate factors associated with the occurrence and distribution of dissolved methane. Wells were sampled for dissolved methane, ethane, propane, and ethene; most samples were also tested for n-butane and isobutane. Field-measured parameters included pH, specific conductance, dissolved oxygen, alkalinity, chloride, and total hardness. The key results of this investigation are:
- Dissolved-methane concentrations ranged from less than the reporting limit of 1.5 micrograms per liter (μg/L) to 8,550 μg/L (0.0015 to 8.55 milligrams per liter [mg/L]). Seven samples exceeded 1,000 μg/L (1 mg/L) dissolved methane.
- Dissolved-methane concentrations were below the reporting level of 1.5 μg/L (0.0015 mg/L) in more than half of all sampled wells (47 of 87 wells). Dissolved ethane was detected in six wells; no higher-chain organic compounds were detected.
- Dissolved methane was detected in 26 of the 41 wells located in valleys (about 63 percent), compared with 14 of 46 wells (about 30 percent) located in upland settings. There was a significant difference in dissolved methane concentrations between the two topographic positions (Mann-Whitney rank-sum test, p<0.001). Fifteen of the 16 wells that exceeded 20 μg/L (0.020 mg/L) methane were located in valleys.
- Dissolved methane was detected in 20 of 37 wells (about 54 percent) in areas underlain by coal, compared with 20 of 50 wells (40 percent) in areas without coal. However, there was no statistically significant difference between the groups.
- Valley wells in areas underlain by coal had the highest proportion of dissolved-methane detections (13 of 17 wells, about 76 percent), followed by valley wells in non-coal areas (13 of 24 wells; 54 percent), upland wells in coal areas (7 of 20 wells; about 35 percent), and upland wells in non-coal areas (7 of 26 wells; about 27 percent).
- Dissolved methane was detected in all nine geologic formations in which sampled wells were located.
- Monthly dissolved-methane samples collected from three wells showed considerable variation. The average percent difference from the median monthly methane concentration in each well was between 20 and 30 percent, although individual variations in each well were frequently larger. The data indicate that a single methane analysis from a well is insufficient to characterize dissolved-methane levels in individual wells.
- Isotopic data from two wells suggest a thermogenic origin for methane. However, methane-to-ethane ratios from these wells and six others suggest both a thermogenic and biogenic origin. More data are needed to better evaluate methane sources in the region.
- Dissolved-methane concentrations from 14 wells in the vicinity of the Accident Dome natural gas storage field are more closely correlated with topographic position than with the number of natural gas production or injection wells within a one-half mile radius of the water wells.
- Dissolved-methane concentrations from this study were compared to county-level studies in Pennsylvania, statewide assessments in West Virginia and New York, and a study of the Upper Delaware River Basin (Pennsylvania and New York). The vast majority of samples in each of these studies had methane concentrations that were either less than 1 milligram per liter (mg/L) or were reported as non-detections. The Maryland study is the only study in which no methane concentration exceeded 28 mg/L; however, methane concentrations above 28 mg/L are uncommon in the other studies, even in areas of extensive natural gas development. These studies show that low-level methane concentrations in well water are common throughout the region, even in areas where gas development has not occurred.
