Tioga Junction Pennsylvania's Story
 

Report
 

In January 2001, State oil and gas inspectors noted bubbles of natural gas in well water during a complaint investigation near Tioga Junction, Tioga County, north-central Pa. By 2004, the gas occurrence in ground water and accumulation in homes was a safety concern; inspectors were taking action to plug abandoned gas wells and collect gas samples. The origins of the natural-gas problems in ground water were investigated by the U.S. Geological Survey, in cooperation with the Pennsylvania Department of Environmental Protection, in wells throughout an area of about 50 mi2, using compositional and isotopic characteristics of methane and ethane in gas and water wells. This report presents the results for gas-well and waterwell samples collected from October 2004 to September 2005.

Ground water for rural-domestic supply and other uses near Tioga Junction is from two aquifer systems in and adjacent to the Tioga River valley. An unconsolidated aquifer of outwash sand and gravel of Quaternary age underlies the main river valley and extends into the valleys of tributaries. Fine-grained lacustrine sediments separate shallow and deep water-bearing zones of the outwash. Outwash-aquifer wells are seldom deeper than 100 ft. The river-valley sediments and uplands adjacent to the valley are underlain by a fractured-bedrock aquifer in siliciclastic rocks of Paleozoic age. Most bedrock-aquifer wells produce water from the Lock Haven Formation at depths of 250 ft or less.

A subsequent detailed gas-sampling effort focused on 39 water wells with the highest concentrations of combustible gas (12 representing the outwash aquifer and 27 from the bedrock aquifer) and 8 selected gas wells. Three wells producing native gas from the Oriskany Sandstone and five wells (two observation wells and three injection/withdrawal wells) with non-native gas from the gas-storage field were sampled twice. Chemical composition, stable carbon and hydrogen isotopes of methane (δ 13CCH4 and δ DCH4), and stable carbon isotopes of ethane (δ 13CC2H6) were analyzed. No samples could be collected to document the composition of microbial gas originating in the outwash deposits (outwash or “drift” gas) or of native natural gas originating solely in Devonian shale at depths shallower than the Oriskany Sandstone, although two of the storage-field observation wells sampled reportedly yielded some Devonian shale gas. Literature values for outwash or “drift” gas and Devonian shale gases were used to supplement the data collection.

Non-native gases from wells in the gas-storage field and native gases from wells producing from the Oriskany Sandstone were similar in chemical composition; methane (volume fraction ranging from 94.5 to 97.2 percent) and ethane (volume fraction ranging from 2.0 to 2.6 percent) were predominant. Isotopic composition data for storage-field gases (median δ 13CCH4 of about -44.1 per mil, δ DCH4 of -168 per mil, and δ 13CC2H6 of -32.7 per mil) were different than gases from the Oriskany Sandstone (median δ 13CCH4 of -34.6 per mil, δ DCH4 of -159 per mil, and δ 13CC2H6 of -40.4 per mil). Both Oriskany  Sandstone and storage-field gases were thermogenic. Compositions of gases from storage-field observation wells were intermediate to, and likely related to mixing of, native gases from the Oriskany Sandstone and non-native gases from the storage-field injection/withdrawal wells.

In water-well samples, methane and ethane were the only hydrocarbons detected at reportable concentrations. Methane concentrations as high as 44.8 mg/L (milligrams per liter) were measured and methane concentrations were greater than 25 mg/L in 38 percent of the 39 samples. The δ 13CCH4 values were measurable in 35 well waters and had a bimodal distribution with modes at -65 per mil (14 wells) and -40 per mil (21 wells). Gas in water samples from the 14 wells in the -65 per mil mode had a small measure of microbial gas (outwash or “drift” gas) in the isotopic signature as determined by carbon- 4 content of methane. The microbial gases were found chiefly in bedrock-aquifer well waters; 10 water wells representing upland and valley settings were along the northern flank of the Sabinsville Anticline.

Waters with microbial gases contained traces of ethane (volume fraction of 0.01 percent or less) that were too small for determination of δ 13CC2H6. Gases from the 21 water-well samples in the -40 per mil mode for δ 13CCH4 were thermogenic. The δ DCH4 and δ 13CC2H6 values for the 21 samples also showed thermogenic signatures. The thermogenic gases were found chiefly in a 17-well cluster on the axis of the Sabinsville Anticline at the eastern margin of the gas-storage field. This cluster corresponds with the cluster of wells with high concentrations of methane from the combustible-gases inventory. An observation well for the gas-storage field, TW805, was nearest to the cluster and three water wells in the cluster contained gases that nearly matched the stable carbon and hydrogen isotope composition in TW805. All the water wells had gas signatures indicating mixing of gases from different origins; however, the overall isotopic composition of methane and ethane showed that the gases in water wells at the eastern margin of the gas-storage field were principally thermogenic. The δ 13CCH4 and δ 13CC2H6 values of the majority of thermogenic gases from water wells either matched or were intermediate between the samples of storage-field gas from injection/withdrawal wells and the samples of storage field gas from observation wells.

Proximity to the axis of the Sabinsville Anticline and the eastern margin of the gas-storage field correspond to the presence of thermogenic gas in water wells. Of the water-well gases with a thermogenic signature, about half are from outwash aquifer wells and half from bedrock-aquifer wells. Of the bedrock- aquifer- well gases with a thermogenic signature, the majority are from wells drilled into bedrock beneath the Tioga River valley. Clay layers in the main Tioga River valley may play a role in keeping gas migration confined to the deep waterbearing zones of the outwash aquifer and the underlying bedrock aquifer.

Isotopic signatures have been used successfully in this study to help discern the origin of the gases in water wells near Tioga Junction. The thermogenic gas found in water wells does not match the composition of native gas from the Oriskany Sandstone. Mixing of Oriskany gases with storage-field gases has occurred, and there was also evidence for mixing of a microbial component of gas in some water wells. The possibility of three or more end-member compositions and many possible mixing scenarios for gases complicate the data interpretation. The lack of samples solely representing native shallow Devonian gas and the small number of storage-field gas samples places some limits on making firm conclusions about the origin of the methane in ground water. The weight of the evidence, however, points to storage-field gas as the likely origin of the natural gases found in water wells near Tioga Junction.

Introduction

In January 2001, State oil and gas inspectors noted bubbles of natural gas in well water during a complaint investigation near Tioga Junction, Tioga County, north-central Pa. A natural- gas storage field near Tioga Junction in the Oriskany Sandstone of Devonian age is a former native gas-production zone about 3,900 ft below land surface. Many abandoned native gas wells exist in the area. By 2004, natural gases, primarily methane, in ground water from household-supply wells were enough of an issue that inspectors were taking action to plug abandoned gas wells; however, plugging of wells did not solve the problem. Questions remained on the origin of the gases, and residents remained concerned. Although it is common for ground water in the region to contain salt brine and hydrogen sulfide, natural gas in ground water was rarely noted by earlier investigators. Historic assessments of ground-water resources only mentioned natural gas in water-supply wells converted from gas wells (Lohman, 1939; Taylor and others, 1983; Williams and others, 1998).

In response to the concerns of the residents, the Bureau of Oil and Gas Management (BOGM) of the Pennsylvania Department of Environmental Protection (PADEP) sampled water from selected household- supply wells in 2001 to 2004. Analytical results for water samples indicated concentrations of dissolved methane as high as 92 mg/L, which greatly exceeds the solubility of methane of about 28 mg/L in water exposed to one atmosphere pressure of pure methane. Concentrations of this magnitude result in outgassing of methane and provide the possibility of accumulation of methane and subsequent explosion, with the potential for substantial property damage and loss of life. The potential for outgassing of dissolved methane in ground water may be estimated using well-construction factors and the water level in the well. A description of the maximum theoretical concentration of methane dissolved in ground water and a discussion of the risks and the importance of venting a plumbing system if methane gas is present in well water are presented in the appendix at the back of the report.

The extent of methane occurrence in shallow ground water in the area near Tioga Junction was not well known. The problem was most commonly reported in water wells in the valley of the Tioga River near Tioga Junction and on the ridge west  of the valley (fig. 1). However, local residents noted that wells to the north near Lawrenceville, Pa. (fig. 1), also had high concentrations of natural gas.

The origin of the natural gas was unknown at the time of PADEP’s investigation. The natural gas could be thermogenic, formed by the thermal breakdown of organic material in sediments resulting from high temperatures caused by deep burial. Possible origins of thermogenic gas include the following: native thermogenic natural gas in the Oriskany Sandstone; nonnative thermogenic gas imported by pipeline and stored in one of three natural-gas storage fields near Tioga Junction (Tioga, West Tioga, and Meeker, at depths of 3,500 to 4,100 ft below land surface); or native gas from strata at depths greater or less than the Oriskany Sandstone. Alternatively, the natural gas could be formed in the shallow subsurface by microbial reduction of carbon dioxide or methyl-type microbial fermentation of organic debris. PADEP expressed concerns that the gas could be migrating from compromised gas wells (improperly plugged, leaking, or abandoned), leaking gas-storage fields, or new uncontrolled pathways opened by some event (possibly seismic) may have changed local permeability in the bedrock.

The U.S. Geological Survey (USGS), in cooperation with the PADEP, conducted a study to investigate the occurrence and origin of natural gas in ground water in the area of about 50 mi2 near Tioga Junction, Tioga County, in the spring and summer of 2005. This report is a summary of findings from the study.

This report describes an inventory of water wells completed in bedrock and in unconsolidated sediments to evaluate the extent of elevated concentrations of hydrocarbon gas in ground water and subsequent detailed isotopic analysis of samples from water and gas wells to investigate the origin of the gas in the ground water. For the inventory, combustible-gas concentrations and field water-quality characteristics, measured in 2005 in 91 wells in an area of about 50 mi2 near Tioga Junction, Pa. For the subsequent detailed analysis, analytical results from sampling 39 water wells for field water-quality characteristics, natural-gas hydrocarbons, other gases (oxygen, carbon dioxide, nitrogen, argon, hydrogen, and helium), stable carbon and hydrogen isotopes of methane, stable carbon isotopes of ethane and dissolved inorganic carbon (DIC), and carbon-14 of methane (9 samples) are presented and discussed. The ground-water gas compositions are compared to 17 gas-well samples (nearby storage gas and native deep gas) and literature values for microbial gas to characterize the origin of natural gas in ground water in the study area. This study focused on the question of gas origin in well water and did not determine a specific source of the stray gases, mechanism of migration, or evaluate the ground waters for other contaminants.

Description of Study Area

The study area in northeastern Tioga County north of Mansfield (fig. 1) includes the Tioga River valley and adjacent uplands about 4 mi east and west of the valley (fig. 2). The area extends from just north of the Tioga and Hammond Reservoirs to the general area of Lawrenceville, Pa., a few miles north of Tioga Junction. Two gas wells south and east of Mansfield also were studied (fig. 1). The Meeker, West Tioga, and Tioga gasstorage fields lie at a depth of about 3,900 ft below a ridge on the western side of the Tioga River valley near Tioga Junction and extend about 8 mi westward.

Ground water for rural-domestic supply and other uses near Tioga Junction is from two aquifer systems in and adjacent to the Tioga River valley. An unconsolidated aquifer of outwash sand and gravel of Quaternary age underlies the main river valley and extends into the valleys of tributaries. Fine-grained lacustrine sediments separate shallow and deep water-bearing zones of the outwash. Outwash-aquifer wells are seldom deeper than 100 ft. The river-valley sediments and uplands adjacent to the valley are underlain by a fractured-bedrock aquifer in siliciclastic rocks of Paleozoic age. Most bedrock-aquifer wells produce water from the Lock Haven Formation at depths of 250 ft or less. Structural Geology Eastern Tioga County straddles glaciated and non-glaciated areas of the Appalachian Plateaus Physiographic Province (Sevon, 2000). The geology of the study area is described in Fuller and Alden (1903), Cathcart and Myers (1934), Lytle (1963), Luce and Edmunds (1981), Williams and others (1998), and Harper (1999) and is excerpted here for the Silurian and younger-age rocks.

The Devonian and older rocks were structurally deformed from folding at the end of the Paleozoic Era into a series of synclines and anticlines with axes trending southwest-northeast. The Sabinsville Anticline (Fuller and Alden, 1903, p. 5) trends northeast from the southwest corner of Farmington Township to Tioga Junction and continues northeast (fig. 2). The anticline axis plunges to the southwest at 50 ft/mi or less. The dips on the north flank of the anticline are 150 to 350 ft/mi; the south flank has beds with dips ranging from 700 to 900 ft/mi. A fault block on the south flank of the anticline, bounded by two north-dipping faults with down-thrown beds to the south, is documented and mapped from gas-well drilling records (Cathcart and Myers, 1934, p. 16) at the top of the Lower Devonian Oriskany Sandstone (fig. 2). In the vicinity of Tioga Junction, the top of the Oriskany Sandstone is about 3,900 ft below land surface. The displacement of the faults is less in limestone beds above the Oriskany Sandstone than in the Oriskany Sandstone (Cathcart and Myers, 1934, p. 19). Additional faults are shown in figure 2 as mapped at the top of the Oriskany Sandstone by Lytle and presented in Luce and Edmunds (1981, p. 63). The fault traces are approximated as dashed lines in all the earlier work. Recent seismic and drilling data have been used to refine the structural geology of the Oriskany Sandstone (Beardsley and others, 1999, p. 289; NE Hub Partners, L.P., 1996, project application on file at Pennsylvania Bureau of Oil and Gas Management). The recent mapping shows the fault traces as solid lines; these imply the faults are located with more certainty than in the earlier work. The faults in the Oriskany Sandstone do not extend into near-surface rocks (Beardsley and others, 1999, p. 289).

Upper Silurian (Salina Group) salt deposits occur (Norris, 1978) at depths greater than 4,500 ft and are proposed targets for solution mining and development of salt caverns for gas storage (Susquehanna River Basin Commission, 1996; NE Hub Partners, L.P., 1996, project application on file at Pennsylvania Bureau of Oil and Gas Management). The Upper Silurian Tonoloway Limestone is part of the Salina Group (Laughrey, 1999, p 105) at a depth of about 4,200 to 4,300 ft. Above the Tonoloway Limestone are Lower Devonian limestones and shales that include the basal Devonian Keyser Formation. Drillers call the Keyser Formation the “Helderberg” (Harper, 1999, p. 113).

Devonian age Oriskany Sandstone (hereafter termed “Oriskany”) overlies the Keyser Formation. The Oriskany is equivalent to the Ridgeley Sandstone and drillers sometimes use the names interchangeably. The Oriskany is an almost pure quartzose sandstone (Harper, 1999, p. 114); however, it is described as calcareous. Above the Oriskany, the bedrock consists of shale, siltstone, and sandstone along with beds of limestone and calcareous rocks of Devonian age. Near the land surface, the rocks are part of the Upper Devonian Lock Haven Formation (Harper, 1999).

Unconsolidated sediments of glacial and postglacial origin cap the geologic section. Glacial ice entered the study area during the Pleistocene Epoch. The glaciers eroded the bedrock and deepened and widened the valleys. During glacial melting, the valleys were partly filled with sediments deposited by glacial ice,  meltwater, and proglacial lakes. The uplands were mantled with till. In the postglacial period, the streams deposited alluvium and organic-rich deposits. Glacial and postglacial valleyfill sediments consist of stratified drift, outwash, alluvium, swamp deposits, lacustrine deposits, and till (Fuller and Alden, 1903). The unconsolidated deposits are referred to in this report as outwash or alluvium. The outwash consists of poorly sorted to well-sorted sand, gravel, silt, and some clay deposited by glacial meltwater or glaciofluvial processes. Outwash, as indicated in this report, refers to the deposits of sand and gravel deposited as valley-trains in the Tioga River and tributary valleys in the study area (fig. 2). The alluvium is generally shallow sand and gravel or silt and fine sand found on floodplains and typically overlies the outwash deposits in the study area. The generalized extent of mapped unconsolidated deposits (Commonwealth of Pennsylvania, 1989; Williams and others, 1998) is shown in all the maps in this report. Till, consisting of clay, silt, some sand, and rock fragments deposited by the glacier, is on the hillsides and ridge tops of the study area but is unmapped in this report.

The hydrogeologic characteristics in the study area were described by Williams and others (1998) and are excerpted here. The generalized framework of the outwash aquifer system with major ground-water recharge and discharge areas is shown in figure 3. The outwash aquifers are stratified unconsolidated deposits of valley fill overlying lacustrine deposits or bedrock with till and bedrock as basal confining units. The lacustrine deposits, till, and bedrock typically have low primary permeability. Hydrogeologic sections (traces of sections shown on fig. 4) from Williams and others (1998) show sequences of glacial deposits that occur along the major stream valley in the study area (fig. 5). The shallow deposits of alluvium and some outwash generally are thin and are unconfined aquifers. In the Tioga River valley, lacustrine deposits may form a local confining unit, and confined aquifers of outwash and related deposits may be present beneath. Sources of natural recharge to the outwash aquifers consist of infiltration of precipitation on the valley floor, infiltration of unchanneled runoff from the uplands at the valley walls, ground-water inflow from the uplands, and infiltration from tributary streams (fig. 3). Mitchell Creek (fig. 5) has been shown by Williams and others (1998, p.16) to have losing reaches. Recharge to the confined outwash aquifers occurs mainly near the valley walls where surficial sand and gravel is in hydraulic connection with the outwash aquifer (fig. 3).

Most ground water flows from the outwash aquifers toward points of discharge, generally either the Tioga River, the lower reaches of principal tributaries to the Tioga River, or swampy areas. In the confined aquifers, ground-water flow discharges to the Tioga River and reaches near the mouths of the principal tributaries through either the confining-unit materials or through overlying unconfined aquifer materials where a confining unit is absent.

The outwash aquifer, unconfined or confined, in the glaciated valleys of Tioga and adjacent counties is characterized by a 20 gal/min median yield to wells (Williams and others, 1998, p. 26); however, yields can be larger for industrial and municipal supply wells. Domestic wells drilled into the outwash aquifers are completed as open-ended casing in a water-bearing zone.

The Lock Haven Formation is the principal bedrock aquifer and is recharged by precipitation. The bedrock is commonly mantled by till on hillsides and ridge tops.

Secondary permeability in the bedrock is dependent on fractures. Regional patterns of fractures or faults in the Lock Haven Formation have not been mapped; however, the southwest to northeast orientation of the channels of the Cowanesque River, Crooked Creek as it flows into Hammond Reservoir, other smaller tributary streams to the Tioga River, and the Tioga River near Mitchell Creek may be due to regional patterns of weakness or increased fracturing that allowed glacial and weathering processes to downcut preferentially in the bedrock, which led to river-channel incision and the present-day channel orientation. Reports by PADEP oil and gas inspectors of natural gas bubbling into streams at the channel bottoms (Robert Gleeson, Pennsylvania Department of Environmental Protection, oral commun., 2006) could support the concept that stream channels are zones of increased fracture density (Wyrick and Borchers, 1981) and preferential pathways for migration of natural gas from depth.

There is no evidence in the hydrogeologic literature for the Tioga  Junction area that faults at depth in the Oriskany extend to near land surface and into the Lock Haven Formation to act as fractured or preferential pathways for ground-water flow. Nevertheless, the faults are mapped for reference in figure 4.

Locally, fractures in the bedrock may be in hydraulic connection with overlying outwash aquifers. Domestic and nondomestic water-supply wells intercept water from fractures in the bedrock and are completed as open boreholes with surface casing. The median yield from domestic wells in the Lock Haven Formation is 10 gal/min (Williams and others, 1998, p. 26).

History of Gas Development and Gas Storage

The first gas well drilled into the dome in the Sabinsville Anticline in northern Tioga County near Tioga Junction produced gas from the Oriskany in 1930 (Cathcart and Myers, 1934). During the next several decades, the Oriskany gas was produced by many gas wells in the Sabinsville Anticline area.

Drilling efforts also produced dozens of dry holes, and many   drill holes were later abandoned (fig. 2). Natural gas was determined on the basis of pressure responses to reside in three independent “pools.” The pools were labeled the Boom, the Meeker, and the Elbridge (Cathcart and Myers, 1934). Southeast of Tioga Junction in a separate anticline, the Oriskany is at a depth of about 5,800 ft at the Krause gas well (fig. 1, the Krause gas well has a total depth of 5,850 ft and the Oriskany is the producing zone of the gas). Other gas-producing strata such as the Trenton and Black River Limestones of Ordovician age are much deeper than the Oriskany—approximately at a depth of 12,000 ft in Tioga County (Christopher D. Laughrey, Pennsylvania Topographic and Geologic Survey, oral commun., 2006).

In the 1960s, the native Oriskany gas in the Sabinsville Anticline near Tioga Junction was depleted, and the pools or reservoirs were converted into gas-storage fields (Lytle, 1963). These reservoirs are now called the Tioga, Meeker, and West Tioga storage fields (fig. 1), respectively, and are collectively referred to as the Tioga gas-storage field.

The gas-storage field receives gas primarily from the gas producing areas of the Gulf of Mexico region of the United States but also from other areas of the Appalachian Plateaus Physiographic Province in the United States and Canada. Gas from a pipeline is injected into the field during spring, summer, and fall for future withdrawal. During winter, gas is usually withdrawn from the Tioga storage field and transmitted to a pipeline for distribution and use. The storage fields currently (2005) are managed by two gas companies. Dominion Gas Company manages the Tioga and West Tioga fields (Thomas Rice, Dominion Gas Company, written commun., 2005). Pennsylvania Power and Light manages the Meeker field (Douglas Welsh, Pennsylvania Department of Environmental Protection, written commun., 2005).
 

As of 2005, methane gas was leaking into water wells near Tioga Junction, Tioga County, north-central Pa. State oil and gas inspectors found water from household-supply wells with concentrations of dissolved methane as high as 92 mg/L, which greatly exceeds the solubility of methane in water. Concentrations of this magnitude result in outgassing of methane and provide the possibility of accumulation of methane and subsequent explosion with the potential for substantial property damage and loss of life. The origin of the methane was investigated by the U.S. Geological Survey in cooperation with the Pennsylvania Department of Environmental Protection (PADEP) in an area of about 50 mi2, using compositional and isotopic characteristics of methane and ethane in gas and water wells. This report presents the results from October 2004 to September 2005.

Ground water for rural-domestic supply and other uses near Tioga Junction is from two aquifer systems in and adjacent to the Tioga River valley. An unconsolidated aquifer of outwash sand and gravel of Quaternary age underlies the main river valley and extends into the valleys of tributaries. Fine-grained lacustrine sediments separate shallow and deep water-bearing zones within the outwash. Outwash-aquifer wells are seldom deeper than 100 ft. The river-valley sediments and uplands adjacent to the valley are underlain by a fractured-bedrock aquifer in siliciclastic rocks of Devonian age, primarily the Lock Haven  Formation. Most bedrock-aquifer wells produce water from the Lock Haven Formation at depths of 250 ft or less.

A review of previous geologic investigations was used to establish the structural geologic framework and identify four plausible origins of the natural gas. The Sabinsville Anticline, trending southwest to northeast, is the major structural feature in the Devonian bedrock. The anticline, a structural trap for a reservoir of deep native gas in the Oriskany Sandstone (origin 1) at depths of about 3,900 ft, was explored and tapped by numerous wells from 1930-60. The gas reservoir in the vicinity of Tioga Junction, depleted of native gas, was converted to the Tioga gas-storage field for injection and withdrawal of non-native gases (origin 2). Devonian shale gas (shallow native gas) also has been reported in the area (origin 3). Gas might also originate as outwash or glacial “drift” gas generated from buried and decomposed organic material in the outwash deposits (origin 4).

During May to August 2005, a combustible-gas inventory of 91 water wells (62 in the bedrock aquifer and 27 in the outwash aquifer) yielded 49 positive readings (greater than or equal to 0.1 percent) in the headspace of water samples (36 from wells completed in bedrock and 13 from wells completed in outwash). The percentage of combustible gas as measured in the headspace of ground-water samples collected in the vicinity of Tioga Junction, when multiplied by a factor of three, gave an approximation of the concentration of methane dissolved in ground water in milligrams per liter. High combustible-gas readings corresponded with methane concentrations in water greater than 10 mg/L; the occurrence was in clusters in the central and southern parts of the study area, and the largest cluster was near the eastern boundary of the gas-storage field.Southwest- to northeast-trending bands of occurrence characterized the northern flank of the Sabinsville Anticline.

A detailed gas-sampling effort from June 2005 to August 2005 focused on 39 water wells with the highest gas concentrations representing the bedrock aquifer (27 wells) and the outwash aquifer (12 wells). Detailed gas sampling also included 8 selected gas wells that were sampled twice—three wells producing native gas from the Oriskany Sandstone and five wells (two observation wells and three injection/withdrawal wells) with non-native gas from the gas-storage field. These data were combined with data collected earlier from the same wells by PADEP. Chemical composition of gas, stable carbon and hydrogen isotopes of methane (δ 13CCH4 and δ DCH4), and stable carbon isotopes of ethane (δ 13CC2H6) were analyzed. Isotopes are reliable indicators of the origin of different gases, whether thermogenic or microbial. No samples could be collected to document the composition of “drift” gas or of native natural gas originating solely in Devonian shale at depths more shallow than the Oriskany Sandstone, although two of the storage-field observation wells sampled reportedly yield some Devonian shale gas. Literature values for outwash or “drift” and Devonian shale gases were used to supplement the data collection.

Non-native gases from wells in the gas-storage field and native gases from wells producing from the Oriskany Sandstone were similar in chemical composition; methane (volume fraction ranging from 94.5 to 97.2 percent) and ethane (volume fraction ranging from 2.0 to 2.6 percent) were predominant. Isotopic- composition data for storage-field gases (median δ 13CCH4 of about -44.1 per mil, δ DCH4 of -168 per mil, and δ 13CC2H6 of -32.7 per mil) were different than gases from the Oriskany Sandstone (median δ 13CCH4 of -34.6 per mil, δ DCH4 of -159 per mil, and δ 13CC2H6 of -40.4 per mil). The Oriskany Sandstone and the storage-field gases were thermogenic, but of different origins. The δ 13CCH4, δ DCH4, and δ 13CC2H6 differences are because the Oriskany gas is more thermally mature than the storage-field gas and is more enriched in 13C and 2H.

Compositions of gases from storage-field observation wells were intermediate to, and likely relate to mixing of, native gases from the Oriskany Sandstone and non-native gases from the storage-field injection/withdrawal wells. One depleted Oriskany gas well (TW805) that is approximately 500 ft northeast of the eastern boundary of the storage field and is used as an observation well for the storage field yielded about a 50:50 mixture of Oriskany gas and storage-field gas. This provided an initial line of evidence that there was some gas leaking and migrating from the eastern boundary of the storage field toward well TW805.

Chemical and stable-isotope characteristics of water-well samples with and without combustible gas were compared by aquifer. Bedrock-aquifer waters with gas exhibited the largest range and the largest median conductivity. The median pH of 6.8 for outwash-aquifer water without gas was the most acidic; the presence of gas seemed to be associated with a buffering effect. This relation also held for bedrock-aquifer waters with and without combustible gas, although the differences in median pH were less pronounced.

Dissolved oxygen and Eh data provided evidence of a geochemical environment in the aquifers that was not oxygenated. Waters with combustible gas had lower median Eh and dissolved oxygen compared to waters without gas, regardless of aquifer. The lack of dissolved oxygen is conducive to anaerobic microbial reduction of sulfate to hydrogen sulfide and carbon dioxide to methane. During well-water sampling, the odor of hydrogen sulfide or “sulfur smell” was common and provided evidence of sulfate reduction by bacteria under conditions of low dissolved oxygen.

Stable-isotope ratios for hydrogen and oxygen of ground water in bedrock (26 samples) and outwash (10 samples) aquifers were not different for wells with and without combustible gas and did not correlate with gas concentrations. When δ Dwater was plotted against δ 18Owater for Tioga Junction ground water and compared to typical river water in north-central Pennsylvania, the ground water composition was similar to that of nearby river waters. There is no strong evidence that the waters from one or both aquifers are isolated from recent recharge or represent Pleistocene-age water or water originating from a high-altitude or cold-environment source. The stable isotopes of water provided no additional information on gas occurrence or origin.

Methane and ethane were the only hydrocarbons detected at reportable concentrations in 37 and 32, respectively, of 37 water-well samples analyzed. The range of methane as a volume fraction in headspace gases ranged from 0.0066 to 94.4 percent; ethane ranged from 0.003 to 1.61 percent. Of the 35 wells with isotopic data for methane, 10 were completed in the outwash aquifer, and 25 were completed in the bedrock aquifer. The range in δ 13CCH4 for ground water was from -35.54 per mil (well TI593, outwash aquifer) to -75.18 per mil (well TI611, bedrock aquifer). The range of δ DCH4 in ground water was from -105.9 per mil (well TI586, bedrock aquifer) to -266.4 per mil (well TI610, bedrock aquifer). There was a bimodal distribution of δ DCH4 results with modes at about -180 and -250 per mil. The - 50 per mil mode appears related to microbial gases, and the -180 per mil mode is related to thermogenic gases. The δ 13CCH4 values in well waters had a bimodal distribution with modes at -65 per mil (14 wells) and -40 per mil (21 wells). Gas in water samples from the 14 wells appears to have had a small measure of microbial gas (outwash or “drift” gas) in the isotopic signature as determined for well water samples by 14C content of methane. The amount of 14C was low, an indication that the microbial gas represented a minor component relative to the other gases in the ground water. The microbial gases were found chiefly in bedrock-aquifer well waters; 10 wells representing upland and valley settings were located along the northern flank of the Sabinsville Anticline. Data for δ 13C of dissolved inorganic carbon give support to microbial processes contributing to the methane in these bedrock- aquifer wells; however, the data provided no new insights on the origin of the carbon in the methane.

Microbial gases typically do not contain ethane. The  waters with microbial gases contained only traces of ethane (volume fraction of 0.01 percent or less). The trace ethane concentrations were too small for determination of δ 13CC2H6. For waters with ethane, the δ 13CC2H6 ranged from -28.73 to -39.03 per mil in a unimodal distribution; the median was -32.17 per mil. Waters from the 21 wells in the -40 per mil mode for δ 13CCH4 were thermogenic. The δ DCH4 and δ 13CC2H6 values for the 21 samples also were thermogenic.

The thermogenic gases were found chiefly in a 17-well cluster on the axis of the Sabinsville Anticline at the eastern margin of the gas-storage field. This cluster corresponds with the cluster of wells with high concentrations of methane from the combustible-gases inventory. An observation well for the gas-storage field, TW805, is nearest to the cluster, and three water wells in the cluster had gases that nearly match the gas composition in TW805. This is evidence that the gas mixture in observation well TW805 was also migrating to ground-water wells.

All the water wells had gas signatures indicating mixing of gases from different origins; however, the overall isotopic composition of methane and ethane showed that the gas in water wells at the eastern margin of the gas-storage field is mostly thermogenic. The δ 13CCH4 and δ 13CC2H6 values of the majority of thermogenic gases from water wells either matched or were intermediate between the samples from storage-field gas injection/withdrawal wells and the samples from storage-field gas observation wells.

Proximity to the axis of the Sabinsville Anticline and the eastern margin of the gas-storage field corresponded to the presence of thermogenic gas in water wells. It is not known why certain wells had microbial gas and others did not. Detailed drilling records were not available for many of the inventoried water wells; hence, without expensive and invasive borehole tests, a detailed understanding of the gas occurrence in the aquifers could not be developed using the limited well data. Well depth was a factor in the observed occurrence of combustible gas for the 62 bedrock wells inventoried. As well-depth range increased from less than 50 ft to 51-150 ft to greater than 151 ft, the percentage of wells with combustible gas increased. On the basis of the known geology and the aquifer framework in the Tioga River valley, the presence of clay-rich layers formed by lacustrine deposits probably plays a role in keeping gas migration confined to the deep water-bearing zones of the outwash aquifer and the underlying bedrock aquifer. Of 11 shallow wells inventoried (with depths of 50 feet or less) in the Tioga River valley, 2 had ground waters with combustible-gas percentages greater than 0.1 percent. These 11 wells likely tap the shallow water-bearing zones of the outwash aquifer. Of 18 wells with depths greater than 50 ft inventoried in the Tioga River valley, 10 had waters with combustible-gas percentages greater than 0.1 percent.

Isotopic signatures have been used successfully in this study to help discern the origin of the gases in water wells near Tioga Junction. The thermogenic gas found in water wells does not match the composition of native gas from the Oriskany Sandstone. Mixing of Oriskany gases with storage-field gases is occurring, and there is also evidence for mixing of a microbial component of gas in some water wells. The possibility of three or more end-member compositions and many possible mixing scenarios for gases complicate the data interpretation. The limited numbers of storage-field gas samples place limits on making firm conclusions about the origin of the methane in ground water. The weight of the evidence, however, points to storagefield gas as the likely origin of the thermogenic gas found in water wells near Tioga Junction. 
 

People
 

I am hoping someone will offer themselves as a point of contact for others interested in this story....

Updates      Canary Watch       Stand Tall       Beacons and Triumphs       Image Gallery

Divide Creek Seep 2008      Timeline of Events       Site Map  (recommended)

Consortium of the Frac'd       Home

All contents of this site, unless otherwise noted are copyrighted by Lisa Bracken, 2007-2009. All rights are reserved.
Visit links outside the scope of this website at your own risk. All information is subject to change without notice.