Methane in well water has been reported anecdotally over the years in the Appalachian Plateau of Maryland; however, no systematic study has been conducted regarding methane occurrence and distribution. The potential development of the natural gas reserves in the Marcellus Shale in western Maryland has raised concerns about whether such development could result in methane contamination of the water-supply aquifers in the region. Methane is not routinely tested for in well water in Maryland, since it does not have an established Primary or Secondary Maximum Contaminant Level (MCL). Because of the concern over possible methane contamination of water wells resulting from Marcellus Shale gas-development activities, in 2012 the Maryland Geological Survey (MGS) collected samples from 49 wells in Garrett County in the Appalachian Plateau region of Maryland. The purpose of this preliminary study was to measure ambient methane concentrations in water wells in the region, and to begin to gain an understanding of the occurrence and distribution of methane in water wells.

Key Findings

Untreated water samples were collected from 49 wells in the Appalachian Plateau portion of Maryland and were analyzed for dissolved methane, ethane, ethene, propane, and field parameters (pH, specific conductance, dissolved oxygen, alkalinity, chloride, and total hardness). Wells were selected on the basis of whether they were in areas with or without coal, and whether they were located in valleys or hilltop+hillside topographic settings.

The key results of this study are:

• Dissolved-methane concentrations ranged from less than 1.5 to 8,550 micrograms per liter (µg/L). Twenty-nine of the 49 samples had less than 1.5 µg/L of dissolved methane. Twenty of the 49 wells had dissolved-methane concentrations greater than 1.5 µg/L.
• Three wells exceeded 1,000 µg/L dissolved methane; all were below the recommended action level of 10,000 µg/L. These wells were located in valley settings in three different coal basins. Two of these wells had detectable levels of ethane; no other samples contained ethane. None of the 49 wells had detectable ethene or propane
• Methane detections (defined as methane concentrations ≥1.5 µg/L) were observed in wells in both the coal basins and the non-coal areas, although a greater proportion of wells in the coal basins had methane detections (11 out of 21 wells, or about 52 percent) than in the non-coal areas (9 out of 28 wells, or about 32 percent)
• Methane was detected in a greater portion of valley wells (16 out of 29 wells, or about 55 percent) than hilltop+hillside wells (4 out of 20 wells, or 20 percent)
• Valley wells in coal basins had the highest proportion of detections (9 of 13 wells, or 69 percent), followed by non-coal/valley wells (7 of 16 wells, about 44 percent), coal/hilltop+hillside (2 of 8 wells; 25 percent), and non-coal/hilltop+hillside wells (2 of 12 wells, about 17 percent).

Background

Methane is a colorless, odorless, flammable gas that can occur naturally in well water with a solubility of about 28 milligrams per liter (mg/L) (28,000 µg/L). Even though methane is not a regulated constituent in drinking water, it is recommended that methane levels above 10 mg/L (10,000 µg/L) need to be addressed to prevent asphyxiation and explosive conditions in confined spaces (Eltschlager and others, 2001). Prior to the present study, no quantitative measurements for methane have been done for well waters in Maryland on a regional basis, although methane has been detected in wells in western Maryland, using a simple qualitative test (i.e. a flame test using well water placed in a jar) (Steve Sherrard, Garrett County Health Department, oral commun., 2012).

Situated in the westernmost part of Maryland, Garrett County and parts of Allegany County lie within the Appalachian Plateau Physiographic Province, which is characterized by outcrops of sedimentary rocks of Carboniferous (Pennsylvanian and Mississippian) and Devonian periods. The gently folded strata form synclines and anticlines that are the source regions for coal and natural gas, respectively (Nutter and others, 1980). The five major coal basins in Garrett County are the Lower Youghiogheny Basin, Upper Youghiogheny Basin, Castleman Basin, Upper Potomac Basin, and Georges Creek Basin. Part of the Georges Creek Basin also extends into Allegany County.

Study Objectives

The purpose of this preliminary study was to measure ambient methane concentrations in water wells in the region, and to begin to gain an understanding of the occurrence and distribution of methane in water wells.

Methods of investigation

(1) Well selection:

Forty-nine wells were selected throughout the Appalachian Plateau Province and sampled for methane and other water-quality constituents. Wells were selected according to whether they were located in coal basins (21 wells) or non-coal regions (28 wells). Additionally, the wells were identified as being located in valleys (29 wells) or hilltop/hillside topographic settings (20 wells). Thus, the sampled wells fell into four groups: coal/valley (13 wells), coal/hilltop+hillside (8 wells), non-coal/valley (16 wells), and non-coal/hilltop+hillside (12 wells). A 50th coal/valley well was sampled but was eliminated from the analysis because the sample was later found to have gone through a water-treatment system.

For each well, the well-permit number was used to acquire well-construction data. Other well-selection criteria were as follows:

• Wells had construction documentation (i.e. well-permit application and well completion report)
• Wells had submersible pumps that were being used on a regular basis
• Samples of untreated well-water could be obtained
• Well water could be run for about 30 minutes (i.e. purging and sample collection)
• Well locations provided a reasonable spatial distribution throughout the study area
• No obvious or potential contamination sources were identified (e.g. well cap on securely; well located upgradient of septic system; well had not been recently chlorinated)

Site inspections were performed to determine suitability prior to sampling

2) Sample collection and analysis:

Water samples were collected at the pressure tank spigot or another tap source that dispenses untreated well water. During well purging, the spigot was turned on, and water was allowed to run into a bucket wherein pH, specific conductance, and dissolved oxygen probes were submerged. Field measurements were recorded at 5-minute intervals until measurements stabilized.

Once the purge was completed, untreated well-water samples were collected in two 40-milliliter glass vials using the inverted bottle technique. Hydrochloric acid drops were then added to the vials to preserve the sample to pH less than 2, re-capped, shaken by hand, and stored on ice.

The samples were brought back to the office, and arrangements were made for the private laboratory (ALS Environmental, Middletown, Pennsylvania) to pick up the samples for analysis. The constituents analyzed were dissolved methane, propane, ethane, and ethene concentrations using the headspace method (RSK-175). Towards the end of the sampling portion of the study, the laboratory analyses also included n-butane and isobutane.

Alkalinity, chloride, and total hardness were measured in the field using unfiltered water samples collected in polyethylene bottles after purging had been completed. In addition to the water-quality measurements, photographs were taken of the purging and sampling area as well as the wellhead for supplemental documentation. At each wellhead, the latitude and longitude were recorded using a handheld GPS unit. Fifty wells were sampled from June through September 2012. Forty wells were residential wells; the remainders were State park and public drinking-water supply wells.

[Prepared by Phung Pham, Maryland Geological Survey]