NGWA Workshop — Groundwater Quality and Unconventional Gas Development: Is There a Connection?: Alphabetical Content Listing

Afternoon Session

Adventures in Groundwater Monitoring: Approaches for Shale Gas Development Sites

Daniel Soeder, National Energy Technology Laboratory

Assessing the potential risks of shale gas drilling and hydraulic fracturing on local and regional groundwater resources has been challenging in the absence of substantial amounts of field data. Obtaining baseline data at drill sites for up to a year before pad construction, performing continuous monitoring of multiple aquifer zones during the drilling and hydraulic fracturing process, and continuing periodic monitoring for several years after well completion are all needed to fully understand the possible pathways for contaminants to reach shallow aquifers. Attempts to engage drillers and landowners to allow monitoring adjacent to drill sites have met with resistance on several levels. Our experience has been that landowners are often concerned the monitoring process will delay gas well completion and production, thereby impacting their royalty payments. Drillers are concerned that by allowing monitoring at one site, it may be required at all sites. Another concern expressed is that if the gas well drilling is done correctly, groundwater monitoring will produce non-detects and we will learn nothing. A third argument is that drillers are already collecting water samples from nearby domestic supply wells, and another study is not necessary. Our responses to these concerns have been: (1) monitoring will be a scientific, not a regulatory activity, and will not impact drilling schedules or royalty payments; (2) even non-detects are helpful for reducing uncertainty in engineering risk assessment models; and (3) domestic wells are commonly open-hole completions that mix water from multiple aquifer flow zones, making it impossible to pin down the origin of stray gas or groundwater contaminants. Scientific access to planned future drill sites has been obtained recently in Pennsylvania and South Dakota. Multi-year field monitoring programs are under development to collect data at both locations, in collaboration with state and university partners. The status of these plans will be presented.

Geophysical Logging of Fresh- and Saline-Water-Producing Zones in Fractured Bedrock of the Northern Appalachian Basin

John H. Williams

Geophysical logging was used to characterize fresh- and saline-water-producing zones in fractured bedrock at selected upland and valley drilling sites in the northern Appalachian Basin. The logged sites included several stratigraphic test holes drilled by the Pennsylvania Geological Survey on upland ridges, a top hole for an oil and gas well on an upland slope, a domestic water-supply well in a tributary valley, and a geothermal wellfield in a major river valley. The geophysical methods included nuclear, electric, induction, and acoustic logging; video and televiewer imaging; and specific conductance, temperature, and flow logging under ambient, pumped, and recovery conditions. Integrated analysis of the geophysical logs with drilling data provided qualitative and quantitative information on bedrock lithology, fracture orientation and transmissivity, and hydraulic head and salinity of water-producing zones. Such information is critical for monitoring and protection of the groundwater resources during energy development.

The geophysical logging delineated water-producing zones in the penetrated interbedded sandstone, siltstone, and mudstone at fractures subparallel to bedding and at steeply dipping fractures related to jointing. In the upland boreholes, the most transmissive fractured zones were penetrated between 50 to 300 feet below land surface. These zones produced freshwater and generally had hydraulic heads significantly higher than zones at intermediate depths. Minimal fracture transmissivity was penetrated in these boreholes below a depth of about 700 feet, and the few deep fractures produced small inflows of saline water. In the valley boreholes, fractured zones that produced saline water were penetrated at 175 and 250 feet below land surface. The saline-water-producing zone at the river valley site was very transmissive and extended for at least 300 feet across the wellfield.

Groundwater Monitoring for Shale Gas: Adaptation of Concepts from DNAPL Site Investigations

Beth L. Parker, Ph.D.

Shale gas development has been a substantial and growing activity for more than a decade but only limited groundwater monitoring has been done to date, nearly all of which has been sampling of domestic wells. This has caused confusion and debate about the results. There is recognition of need to move beyond this type of sampling to science- based monitoring serving multiple objectives and relevant questions. The challenge for the hydrogeological community is to figure out how such monitoring should best be done. This presentation examines some concepts based primarily on experience drawn from investigations of DNAPL sites on fractured sedimentary rock.

The main challenges for shale gas development are fractured sedimentary rock monitoring spanning a much larger depth range than that of conventional practice. The concerns about shale gas pertain to varied contaminant types from both shallow and deep sources. Shallow sources including fracking chemicals, fuels and flowback fluids and contaminants such as natural gas and salinity originating at depth below the fresh groundwater zone, most likely from the intermediate zone and less so from the deep zone where the fracking occurs. But the further challenge draws from multiple objectives for assessing performance, sentry and receptor monitoring, each with their own design and baseline data needs. There is much experience from contaminated site investigations to guide the shallow monitoring but not the deeper monitoring into the Intermediate zone. Rather than conventional monitoring wells where one well is installed in each drill hole, shale gas monitoring can derive effectiveness from use of depth- discrete, multilevel systems (MLS's) in single holes for which there are many versions to select from depending on the depth and nature of the hydrogeology (e.g FLUTe, Westbay-Schlumberger and Solinst- Waterloo and CMT systems). Experience at DNAPL sites shows that effective monitoring first requires subsurface characterization to guide vertical placement of ports and seals and aid interpretation. Experience at DNAPL sites in sedimentary rock covers the depth range down to 400 m and is further supported with experience from deep sedimentary rock investigations for radioactive waste repositories and CO2 injection sites.

How Hydraulic Fracturing Impacts Our Water: The Pennsylvania Experience

Susan Brantley

The number of unconventional Marcellus Shale wells in Pennsylvania has increased from 8 in 2005 to more than 7000 today. To understand the Pennsylvania experience, we are analyzing publicly available data. After removing permitting and reporting violations, the average percentage of wells/year with at least one notice of violation (NOV) from PA DEP has averaged almost 20%. Most violations are minor. Nonetheless, several incidents have led to fears that shale gas cannot be developed without endangering our water supply. Indeed, development of shale gas using the newest form of hydraulic fracturing has not only caused methane migration, but also spills of briny formation waters, hydraulic fracturing fluids, sediments, drilling muds, fuels, drill cuttings, and hydrostatic testing waters into surface waters and groundwaters. Shale gas activities have contaminated natural waters with benign substances such as sodium chloride—sometimes at toxic concentrations—as well as more toxic naturally occurring radioactive materials (NORMs) as well as solutes such as bromide that can combine with organic compounds and disinfectants to create toxic species. Such incidents have led many to wonder: Can we maintain clean water and still develop shale gas? To date, it remains impossible to answer this question adequately due to a dearth of publicly accessible, transparent data. However, the data that are available for Pennsylvania show that the frequency of problems is relatively small compared to the number of gas wells.

Afternoon Session (cont.)

A retrospective case study of perceived water well interference by coal bed methane fracking

Cathryn Ryan, Ph.D.

Although water well concerns are one of the most frequently cited concerns about hydraulic fracturing, few case histories are available. A 2004 case study of perceived water well interference by ‘stimulation’ of a coal bed methane well is reviewed here to highlight the challenges for both the energy industry and domestic well owners contributing to a lack of a ‘social license to operate’ in parts of Canada.

The circumstances around the perceived water well interference were evaluated by reviewing relatedenvironmental consulting reports, water well records, baseline water well testing and energy well activities in the region, and through discussions with various individuals directly involved. The timeline of events suggest the water well could have been affected by the stimulation. The energy well perforations were unusually shallow when compared to other energy wells in the region, and similar in elevation to the domestic use aquifer.

Although the evidence suggests that the energy well stimulation may have been related to the perceiveddomestic water well interference, subsequent stimulation in energy wells located even closer to the rural residence (but perforated at greater depth) did not cause observed interference in a replaced domestic water well on the same property.

Appropriate evaluation of water well complaints is clearly challenging. Dedicated monitoring systems are seldom used to evaluate hydraulic fracturing impacts in shallow groundwater. The industry reportedly deals with a large fraction of apparently invalid water well complaints. In part in an effort to minimize invalid water well complaints, they i) do not disclose details about confirmed water well complaints that are appropriately addressed and ii) do not provide any notification of energy activities in a given region. The latter can prevent the timely collection of appropriate information, with subsequent the difficulties in conducting a ‘post-mortem’ analysis and perceived water impacts whose cause(s) are difficult to assess.

Groundwater Ages and Mixing in Western Energy Basins: Implications for Monitoring

Peter B. McMahon, Ph.D.

Identifying the effects of energy development on groundwater quality can be difficult because interactions between hydrocarbon reservoirs and aquifers can be complex and involve both natural and human processes. Groundwater age and mixing data can strengthen the design of monitoring programs and interpretation of monitoring data by providing better understanding of the groundwater flow systems. Age, chemical, and isotopic tracers were used to examine age and mixing characteristics of shallow groundwater in the Piceance Basin, Colorado and Williston Basin, Montana and North Dakota. In the Piceance Basin, natural gas in tight sands of the Mesaverde Formation is extracted using hydraulically-fractured vertical wells. In the Williston Basin, oil in tight clastic and carbonate rocks of the Bakken Formation is extracted using hydraulically-fractured horizontal wells. In both basins, age-tracer data (sulfur hexafluoride, tritium, helium-3, helium-4, carbon-14) revealed complex patterns in shallow groundwater ages (< 10 to > 50,000 years). Chemical and isotopic data (chloride, bromide, sulfate, methane, strontium, δ²H-H₂O, δ¹⁸O-H₂O, δ¹³C-CH₄, ³He/⁴He, ⁸⁷Sr/⁸⁶Sr) showed that shallow groundwater mixed with water from adjacent formations and with water from much deeper formations. The age data provide information on time scales at which water quality changes in aquifers might be expected to occur. This information could be used to establish appropriate distances of monitoring wells from energy development activity and the necessary duration of monitoring. The mixing data provide information on hydraulic connections between shallow and deep formations that could be used to guide more detailed monitoring efforts and also indicate areas where additional care in drilling oil and gas wells could be warranted to avoid further enhancing that connectivity.

Smart-Monitoring to Address Risks of UCOG Development

Jon Fennell, M.Sc., Ph.D., P.Geo.

Canada is well placed to take North America forward in the unconventional oil and gas (UCOG) sector. Given the trillions of cubic feet of shale gas and associated liquids, development of these resources will generate significant economic benefit for the country and provide a clean energy source for end users. On the other hand, concern is mounting regarding the potential impacts of UCOG development and associated hydraulic fracturing activities on non-saline groundwater resources and connected systems. The Government of Alberta has recognized this concern, and responded in kind by working to enhance the provincial groundwater observation well network in active and future development areas. This presentation will showcase the multi-attribute risk analysis approach designed to map subsurface risk and surface access opportunity, and identify optimal monitoring locations to define baseline groundwater conditions and detect changes due to hydraulic fracturing activities.

Breakout Groups Meet

Breakout Session 1: What current recommendations can be made with respect to baseline sampling constituents and protocols?

Breakout Session 2: What are the long-term (3-5 years) recommendations for hydrogeologic research and monitoring?

Change to Breakout Groups

Keynote: Evolution of Groundwater Monitoring - Implications for Shale Gas

John A. Cherry, Ph.D.

Morning Session

Effective Bradenhead Pressure and Mitigation Protocols for Reducing Stray Gas Risk in Groundwater

Anthony Gorody, Ph. D., P. G.

Hydrocarbon gas migration through uncemented annuli and small channels in cemented casing poses a risk that, under the right circumstances, may allow stray hydrocarbon gases to invade shallow groundwater. These gases migrate vertically and preferentially accumulate in the annular spaces between surface and production casing long before groundwater can become significantly impacted. For this reason, bradenhead pressure monitoring in conjunction with systematic sampling and analysis are effective risk mitigation tools.

An effective bradenhead pressure monitoring program includes the following: (a) defining a minimum 9 square mile area ahead of drilling; (b) compiling and mapping depths of surface casing and underlying uncemented annuli intervals; (c) monitoring bradenhead pressures in wells throughout the area; (d) identifying wells where bradenhead pressure builds up to significant pressures within a 24-hour period following venting; (e) collecting paired bradenhead and production gas samples in all wells found to have both initially elevated bradenhead pressures and significant recurring bradenhead pressures; (f) analyzing paired gas samples for gas composition and stable isotopes; (g) planning and implementing a cement squeeze or well abandonment program to shut off the source of bradenhead gas in wells; (h) planning and implementing a drilling and mud logging program for the first new well development well drilled in the 9 square mile area; (i) collecting mud gas samples and analyzing selected mud gas samples for gas composition and stable isotopes; (j) identifying and contouring a shallow formation boundary near or below the average depth of surface casing in the area; and (k) continuing periodic bradenhead pressure and groundwater monitoring following drilling. This approach is particularly effective in areas where historic drilling practices have left annular intervals of production casing uncemented. Such protocols also facilitate identifying point sources of stray hydrocarbon gases discovered in response to water well complaints and pre-drill water well sampling.

Gas Migration Behind Cased Energy Wellbores as a Transient Process

Maurice Dusseault, Ph.D., P.Eng.

Because of autogenous shrinkage of cement and the subsequent reduction in radial contact stress between the borehole wall and the cement, a micro-annulus may develop that creates a pathway by which natural gas may rise outside the casing. This gas may become evident as sustained casing pressure (SCP) in the U.S., or in Canada as surface casing vent flow (SCVF) or as gas migration (GM), i.e., that which occurs outside the casing strings. GM may emit at ground surface as a greenhouse gas or penetrate shallow aquifers, causing groundwater contamination. Gas emissions recorded as surface casing vent releases frequently display a pulsing or periodic nature. These gas slugs are well known to those using noise logs to detect annular cement pathways. We hypothesize that these pulses or slugs are due to the formation of Taylor bubbles, which are created by the coalescence of small bubbles and their buoyant ascension behind the casing. Their displacement pressure is sufficient to overcome the capillary entry pressure posed by shallow aquifers and thus cause groundwater contamination. The periodic nature of these gas slugs is of consequence if such emissions are to be fully accounted for during monitoring by hydrogeologists in water wells.

Presence and Origin of Dissolved Gas in Groundwater in the St-Edouard Area (Quebec, Canada)

Christine Rivard

The Upper Ordovician Utica Shale, located in the St. Lawrence Lowlands, is a potential shale gas producer. The industry targeted this formation between 2007 and 2010, until a moratorium was imposed. A total of 29 wells were drilled in this shale, of which 18 were subjected to hydraulic fracture stimulation. The Geological Survey of Canada began a project in 2012 in the St-Edouard area, 65 km southwest of Quebec City, to study the potential impacts of shale gas development on near-surface aquifers. This study includes four components—geochemical, geomechanical, geophysical, and hydrogeological—so as to collect and integrate multi-source data in order to better understand the behavior of the hydrogeological system as a whole.

Results of geochemical analyses show the presence of methane in groundwater throughout the region, but with widely variable concentrations. Some samples also show the presence of propane, indicating a thermogenic component. Results of geochemical analyses of soil gas provide similar indications. Available isotopic analyses suggest that all groundwater samples have a biogenic signature, but that 20% to 40% of samples have a mixed origin (i.e., containing both biogenic and thermogenic gas). Geochemical analyses of core samples from shallow wells (50-60 meters depth) show that the near-surface bedrock contains hydrocarbons (C1-C20) that could constitute the source for both biogenic and thermogenic gas. Additional analyses are planned to identify more precisely the origin of dissolved natural gas and the possibility for fluids to migrate from deep shale units to the surface aquifers.

Regional Variability in the Gas Geochemistry of the Appalachian Basin and Implications for Groundwater Investigations

Fred Baldassare

The occurrence of stray gas (methane) in aquifer systems can be a natural condition or due to anthropogenic activity. Investigations of stray gas migration require data collection and evaluation at the site-specific level. Analyses of gas geochemistry provide evidence of gas origin and focus for the investigation that must be integrated with other data types including geological and engineering to conclude a stray gas source.

Most gases in the Paleozoic strata of the Appalachian Basin were either cracked directly from kerogen or cracked directly from oil that was generated in those source rocks. These gases were subjected to further thermal stresses, geologic upheaval, and migration. Other sources of gas include the regionally prominent coal beds of the bituminous coalfields and microbial gas in glacial drift found in the northern area of the basin. Any of these gases may have migrated to shallower formations on geologic timescale or contemporaneously due to anthropogenic activity. Gas migration of both orders may yield mixing of gases on both a regional and local scale. Understanding the complexity associated with the interplay of gases of different origins found in the shallow system is instructive for investigations of stray gas migration.

Morning Session (cont.)

Assessment of Baseline Groundwater Quality and Unconventional Development of Hydrocarbons: The Science and the Belief

Donald Siegel, Ph.D.

Opponents to unconventional hydrocarbon development in the last five years have developed a compelling narrative that potable waters may be broadly contaminated by methane and various solutes derived from the industrial practice. This narrative cleverly incorporated video, such as the documentary Gasland, investigative newspaper reports, and some peer-reviewed scientific publications based on small data sets. While local contamination may be technically “possible,” is it scientifically plausible or probable?

To what extent is this belief reality? What evidence actually shows drilling for unconventional hydrocarbons causes meaningful contamination above and beyond natural water chemical variability in the Appalachian Basin? In my talk I show there has been minimal harm to potable waters from unconventional hydrocarbon development compared to that caused by other acceptable industrial practices. I arrive at my conclusion from the basic principles of hydrogeology, physics, and chemistry, and analysis of more than 30,000 baseline samples collected where drilling already has been most intense in Pennsylvania and Ohio.

Baseline sampling, if done, should focus on halogens (Cl, Br, I), not major solutes such as sodium, trace metals such as iron and barium, nutrients, or fecal coliform bacteria, to assess possible solute contamination from oil and gas development. A forensically viable suite of VOCs and SVOCs associated with hydraulic fracturing could be explored to assess changes in methane concentrations if they occur beyond reasonable variability in native waters. This suite needs to be designed to sample those compounds that do not biodegrade too fast to be seen, and to avoid cross contamination by materials used during the drilling of production wells.

Evaluating Key Sources of Groundwater Quality Variability in Residential Water Wells for Pre-Drill Sampling

Stephen Richardson, Ph.D., PE, PEng

Reported changes in residential water well quality in areas of active shale gas operations have fueled significant concerns regarding the impact of shale gas extraction on surrounding drinking water resources. Given that water quality can vary substantially over time due to factors unrelated to shale gas extraction (e.g., intensity of residential water use, well construction, aquifer geochemistry, precipitation events, temperature, weather patterns, and road salting), the source(s) of observed variability in water quality is often left in question. To further complicate matters, recommended practices for baseline water quality sampling vary considerably between states and organizations, imparting additional variability on water quality analytical results. An improved understanding of the effect of sample methodologies and temporal variability on dissolved gas concentrations, isotopic signature, and other water quality parameters in residential water wells is needed to better differentiate sources of natural variability in residential water wells from induced variations (e.g., stray gas events).

This presentation summarizes the results of two field studies aimed at evaluating sources of variability in pre-drill groundwater data and quantifying their impact on dissolved gas concentrations and other water quality parameters from a series of private residential water supply wells in northeast Pennsylvania. The objectives of the study are twofold: (1) investigate the effects of sampling methodologies on pre-drill water well quality, and (2) quantify the degree of variability in dissolved gas concentration, isotopic signature, and general water quality parameters over an 18-month period. The goal of this work is to form a better understanding of the inherent variability in pre-drill and post-drill analytical results and develop recommendations for improved sample collection methods and data interpretation.

Preliminary Results of a Dissolved Methane Sampling Campaign in Texas

Jean-Philippe Nicot

Many constituents related to oil and gas, such as methane, exist naturally in groundwater. Recent studies of aquifers in the footprint of several gas plays across the U.S. have shown that (1) dissolved thermogenic methane may or may not be present in the shallow subsurface, and (2) shallow thermogenic methane could be natural and is not necessarily a consequence of gas production from a gas play. A team of researchers at the University of Texas at Austin Bureau of Economic Geology is currently conducting a large sampling campaign across the state of Texas to characterize shallow methane in freshwater aquifers overlying shale plays and other tight formations. Some aspects of the study and results will be presented.

The Relationship Between Redox State and Methane Concentrations: Implications for Pre- and Post-Drill Sampling

Lisa Molofsky

This study investigates the relationship between prevailing redox conditions and dissolved methane concentrations in more than 800 pre-drill residential water well samples from Susquehanna County, Pennsylvania. For each sample, redox state was determined using threshold concentrations for redox indicator parameters (e.g., nitrate, iron, manganese, and sulfate). In addition, the water type (i.e., Ca-rich vs. Na-rich) and topographic location (i.e., valley vs. upland) was determined for each sample. Based on this information, we identified a combination of environmental factors (i.e., advanced redox state, Na-rich water type, and valley setting) that are strongly related to naturally elevated methane concentrations in water wells. Given a fundamental understanding of the relationship between the microbial oxidation of methane and groundwater geochemistry, we assess the changes in water geochemistry that may be expected from a sudden new influx of methane in these water wells.

Overall Discussion of Conclusions and Next Steps

Posters

A Comparative Analysis of Water Use for Hydraulic Fracturing

Mike Nickolaus, PG

It has often been said that hydraulic fracturing uses a relatively small amount of water when compared to other uses. Clearly, high volume horizontal well hydraulic fracturing is a relatively minor user of water at the national level, accounting for less than 0.1% of water use when compared to other consumptive water uses such as irrigation and public water supply. However, the question remains as to whether or not this ratio of usage between sectors remains consistent down to the local level. In this presentation we will discuss the findings of a comparative analysis of water use in the sectors of irrigation, public water supply, and hydraulic fracturing at decreasing geographic scales in three oil and gas producing states (Texas, Pennsylvania, North Dakota) to determine whether or not the relationship between water usage in these sectors remains relatively static from national to county levels. To analyze water usage across the three sectors, data from USGS water use publications and the FracFocus chemical registry were utilized. Where unique state or regional conditions might apply, additional information from state regulatory agencies was utilized to ensure that water usage was consistent with agency figures.

A Comparison of the Methods Used for the Investigation of Dissolved Gas Concentrations

Christine Jampo

The analysis for dissolved gases in pre- and post-drill samplings of water wells in and around active drill sites has become an important and controversial topic. The results of such sampling and analysis can have a significant impact on water quality for the affected homeowner and a significant impact on costs of operation for the natural gas generator. With such high stakes for quality of life and reputations at hand, it would seem imperative that a very standardized and consistent method and analytical approach be utilized. Unfortunately, this is not the current practice. Various applications of RSK-175, a proposed ASTM method, and methods developed by several manufacturers are all used for pre-drill assessments and more significantly for stray gas investigations. Oftentimes the method, or variation of a given method, used by the enforcement agency’s lab will not return results consistent with the laboratory used by the generator’s lab. Yet each reporting lab is a certified lab for the application of dissolved gas analysis.

This paper will discuss several of the differences between the different methodologies and the variables within a given methodology that can impact the results generated from the analysis of a given sample location. Impact of sampling methodology will also be discussed along with some initiatives of the MSC around this issue.

Assessment of Drinking Water Risk from a Hypothetical Marcellus Shale Flowback Water Spill

William Rish, Ph.D.

One source of human health risk associated with shale gas development is possible exposure to chemicals present in a spill of flowback water from the horizontal hydraulic fracturing process. In the Marcellus Shale, water with sand and low concentrations of chemical additives are used to hydraulically fracture a horizontal well. These additives may include friction reducers, corrosion inhibitors, oxygen scavengers, scale inhibitors, and biocides that are blended into the water and sand mixture that is injected into a well at high pressures to fracture the shale rock. Following fracturing, a portion of the injected water flows back (i.e., flowback water) out of the well and is collected in tanks and/or impoundments. Flowback water contains salts, metals, and organic compounds from the formation and the chemicals that were introduced as additives to the influent stream. The potential exists for flowback water to be spilled during handling, transport, and storage.

In 2009, the Gas Technology Institute (GTI) published the findings of sampling and laboratory analysis of flowback water from 19 shale gas wells drilled into the Marcellus Shale in Pennsylvania and West Virginia. The chemical analysis results from the GTI study are used in this paper to characterize the chemical composition of an assumed spill of flowback water. A risk assessment is presented that quantitatively evaluates possible human health risk from a hypothetical scenario where 10,000 gallons of this flowback water is spilled on the ground, infiltrates into groundwater that is a source of drinking water, and a person located downgradient drinks the groundwater. Key uncertainties encountered when estimating risk are given explicit quantitative treatment using Monte Carlo Analysis. Chemicals significantly contributing to estimated health risks are identified, as are key uncertainties and variables to which risk estimates are sensitive.

Baseline Methane Concentrations in Drinking-Water Wells in the Appalachian Plateau Province of Maryland

David Bolton

The potential development of natural gas from the Marcellus Shale in western Maryland has raised concerns about whether such development could result in methane contamination in drinking-water wells. Because methane is not routinely analyzed in drinking water in Maryland, virtually nothing was known of its occurrence and distribution. In 2012 and 2013, the Maryland Geological Survey collected water samples from 77 water-supply wells in Garrett and Allegany Counties in the Appalachian Plateau physiographic province of western Maryland. The near-surface geology of the region is characterized by gently folded Devonian, Mississippian, and Pennsylvanian rocks. Coal is present in the synclinal basins, and natural gas is produced from and stored in the (Devonian) Oriskany Formation in parts of the anticlinal structures. Well selection was based on geology (coal vs. non-coal areas) and topographic setting (valley vs. hillside or hilltop), resulting in wells being assigned to one of four categories: coal/valley (15 wells); coal/hilltop+hillside (20 wells); non-coal/valley (17 wells); and non-coal/hilltop+hillside (25) wells. Dissolved methane was detected in about 44% of all wells (reporting level: 1.5 micrograms per liter [µg/L]). Concentrations ranged from less than 1.5 to 8550 µg/L; four of the 77 wells exceeded 1000 µg/L. Methane was detected in wells from all geologic formations. Wells located in coal/valley settings had the highest percentage of methane detections (about 73%), followed by wells in coal/hilltop+hillside settings (45%), non-coal/valley settings (41%), and non-coal/hilltop+hillside settings (28%). Monthly methane concentrations analyzed in samples from three wells generally varied by 20% to 30% from the median monthly methane concentration in each well. Isotopic analyses for ¹³C-CH₄ and ²H-CH₄ suggest a thermogenic origin for the methane. Additional methane sampling is planned in the vicinity of a natural-gas storage facility in Garrett County.

Chemical Characteristics of Saline Water from the Catskill Formation in Test Wells in Northeastern Pennsylvania

Dennis Risser

Saline water was encountered in three test wells drilled to depths of 1400 to 1664 feet in upland settings in Bradford, Tioga, and Sullivan counties in northern Pennsylvania. Geophysical logging of the test wells identified elevated specific conductance of water from fractures in the Catskill Formation at depths ranging from 914 to 1026 feet below land surface. Groundwater samples, collected by the use of a point sampler, verified inflows of saline water having total dissolved-solids concentrations of at least 2470 to 12,700 milligrams per liter. Chemical analyses showed that the saline water in the Catskill Formation had chemical characteristics similar to oil and gas well brines diluted with freshwater.

The predominant major ions in samples of saline groundwater from the test holes were sodium, chloride, and calcium. The predominant trace constituents were strontium, bromide, barium, lithium, iron, manganese, zinc, fluoride, boron, molybdenum, and arsenic. Mass ratios of chloride to bromide in saline-water samples were about 100, similar to reported values of brines from western Pennsylvania and Marcellus flowback waters.

The saline water from the Catskill Formation contained hydrocarbon gases. Methane concentrations ranged from 7.8 to 55 milligrams per liter in five samples. The isotopic ratios of ¹³C/¹²C and ²H/¹H of the methane indicated a thermogenic gas, but one which was much more depleted in ¹³C and ²H than gas produced from the Marcellus Formation.

The analyses showed that the inorganic chemical composition of saline waters found at relatively shallow depths in the Catskill Formation can be difficult to distinguish from theoretical mixtures of freshwater with deeper brines. Thus, inorganic constituents should be used with caution as indicators of contamination from shale-gas operations, even for elements such as barium, strontium, and bromide that have been suggested as highly specific indicators of Marcellus waters.

Comparison of Water Well Quality Before/After Nearby Hydraulic Fracturing of Gas Wells in Pennsylvania

Bryan Swistock

Over three million rural residents in Pennsylvania rely on private water wells for their home or farm water supply. Marcellus Shale gas well drilling is increasingly occurring in proximity to these water wells, raising questions about potential impacts on water quality or quantity. This project sought to provide an unbiased and large scale study of water quality in private water wells both before and after the drilling of Marcellus Shale gas wells nearby, and to document both the enforcement of existing regulations and the utilization of voluntary measures by homeowners to protect water supplies. During 2010-2011, more than 200 private water wells located within about one mile of active Marcellus gas well sites were tested before and after gas well drilling. Water samples were evaluated for 14 to 18 water quality parameters by state-accredited water testing labs. Approximately 40% of the water wells in this study failed at least one drinking water standard before drilling occurred. Dissolved methane gas was detected in about 20% of the water wells before gas drilling occurred although concentrations were generally below 1 mg/L. Statistical analyses did not suggest significant changes in inorganic water quality parameters after gas well drilling or hydraulic fracturing. A small number of water wells, generally within 3000 feet of the nearest gas well, had increases in sediment and metals which may have been due to disturbance from drilling. This research was limited to the study of relatively short-term changes that might occur in water wells near Marcellus gas well sites. Additional monitoring at these or other longer-term studies will be needed to provide a more thorough examination of potential water quality problems related to Marcellus gas well drilling.

Compounds Specific Isotope Analysis to Recognize and Evaluate Natural Attenuation of Methane in Groundwater

John Wilson, Ph.D

Biodegradation can be an important mechanism for the natural attenuation of methane in groundwater. Biodegradation of methane can be carried out by three classes of microorganisms: bacteria that use oxygen as the electron acceptor, bacteria that use nitrate as the electron acceptor, and archaea that form a syntrophic association with sulfate-reducing bacteria. At sites that have been impacted with methane from unconventional gas development, sulfate is likely to be the important electron acceptor. All three classes of microorganisms fractionate the stable isotopes of carbon and hydrogen in methane as the methane is metabolized. It should be possible to recognize biodegradation of methane from the associated fractionation of carbon or hydrogen isotopes. Under sulfate-reducing conditions, degradation of as little as 50% of the methane originally present in the groundwater can shift the value of δ¹³C by 10‰. For years, geochemists in the oil and gas industry have used stable isotopes to identify the source of methane in a reservoir. The same approach and the same techniques should have application to recognize and evaluate the natural attenuation of methane that has escaped the reservoir and entered shallow groundwater.

Field-Scale Monitoring — a Collaborative Approach

Greg White, P.G.

Groundwater and surface water monitoring programs are central to responsible development of unconventional oil and gas resources and are often required by regulators to assess potential long-term impacts associated with upstream oil and gas extraction. In New South Wales, Australia a monitoring program is required by state regulators to identify sensitive groundwater and surface water receptors, characterize baseline conditions, assess changes in the receiving environment, and mitigate potential impacts associated with coal bed methane extraction. However, a monitoring network over the entire potential upstream development area, covering more than 7500 square miles, would present significant challenges due to high capital costs associated with multi-depth well installation (with depths ranging between 60 feet and 3500 feet), lengthy implementation timeframes, long-term monitoring requirements and associated costs, access issues, and public relations considerations.

To overcome these challenges, a collaborative approach was employed between the natural gas proponent and the state water board. The process involved negotiations with the water board to gain access to the state’s surface water and groundwater monitoring networks within and around the proposed development area. Utilizing the state’s infrastructure provided the gas proponent with a robust purpose-built monitoring network targeting key water resources, an extensive historical and independent dataset to characterize baseline conditions, and an avenue for increased transparency with the public. The use of existing infrastructure reduced the need to install new monitoring infrastructure and provided historical monitoring data to supplement the current understanding of baseline conditions. Further, the process provided the water board with a thorough assessment of their monitoring sites and a source of ongoing data to utilize for various internal initiatives.

While this collaborative approach does not remove all challenges associated with field-scale data acquisition, it provides a framework for reducing costs and characterization timeframes, increasing transparency, and ultimately streamlining resource development in an environmentally responsible manner.

Hydraulic Fracturing for Oil and Gas Production in California — Water Use and Water Quality

Eric Nichols, PE(CA)

Hydraulic fracturing has been used in California for well stimulation since 1953. In 2013, 830 wells were hydraulically fractured. Most wells are 1000 to 4000 feet deep. The average volume of water used to hydraulically fracture each well in 2013 was 127,000 gallons (0.39 AF), and the total volume of water used in 2013 for hydraulic fracturing was 323 AF.

The oil industry is exploring the potential of the deep Monterey shale at depths of 7000 to 14,000 feet. Initial exploratory wells have used an average 10 AF of water for hydraulically fracturing each well. Based on recent estimates of Monterey oil production, the volume of water expected to be used by 2030 for all hydraulic fracturing in California, including the Monterey Formation, is less than 2500 AF, representing 0.004% of freshwater usage in California.

There have been no documented incidents of groundwater contamination in California caused by hydraulic fracturing because: (1) California has strict regulations for well construction to protect groundwater; (2) hydraulic fracturing fluid consists primarily of non-toxic materials (water, sand, guar, etc.); (3) oil producing zones are isolated from freshwater and separated from overlying aquifers by several thousand feet of sediment; and (4) the practice has primarily been used in the western San Joaquin Valley, where freshwater is absent and groundwater is often naturally saline.

The deep Monterey Formation in the San Joaquin Valley is separated from overlying aquifers by 5000 to 13,000 feet of sediment, and groundwater below a depth of 1000 feet in many areas has been found to be brackish or saline, and commonly contains natural gas.

Current regulations require groundwater monitoring or proof that no protected groundwater is present. Protected groundwater contains less than 10,000 mg/L TDS and is not in an oil- or gas-producing zone. Regulations also require public disclosure of all hydraulic fracturing fluid components, baseline groundwater sampling, and ongoing monitoring.

Hydraulic Fracturing — a South African Perspective

Fanie de Lange

The promise of natural gas to be a “game-changer” in energy-related questions has stimulated interest in central South Africa (a region which is generally known as the Karoo) fossil resources and political leadership is engaging in shale gas development. The moratorium issued on application requests by energy firms to explore economical viable gas reserves was lifted in September 2012 and public debate is gaining momentum concerning the effects of hydraulic fracturing.

Hydraulic fracturing has become a prevalent public and regulatory issue in most countries developing shale gas. One of the key issues being debated is the protection of groundwater resources in rural areas, since most of South Africa’s rural and some inland cities are dependent on groundwater for potable water supply. Much interest in the country is now directed towards the Karoo because of its potential to deliver shale gas as a future fuel source. Production of shale gas by means of hydraulic fracturing has the potential of contaminating shallow groundwater resources. A large range of chemical elements that could pollute the freshwater is possible, e.g. (a) the current groundwater and methane that is captured in the organic Ecca shale, (b) fracking fluids that will be used during the process, and (c) existing elements in the shale that will be released due to input of fracking fluids, e.g., NORMs. Water currently captured in the organic shale is not suitable for drinking by humans. Due to the unique geological structure of the Karoo, the presence of dolerite structures, a number of risk mitigation methods might be required to successfully develop hydraulically fractured wells. Holistically, the chemical and hydrogeological impacts related to wellfield development cannot be ignored in the Karoo aquifer system, as it has the potential to directly influence human and environmental health.

Hydrolysis of Hydraulic Fracturing Fluid Organic Additives and Their Interaction with Pyrite

Nizette Edwards

The Marcellus Shale is currently one of the largest resources and producers of natural gas in the U.S. Shale gas production is expected to increase in the coming decades and its extraction requires the use of hydraulic fracturing and up to 7 million gallons of water per well. Various chemical compounds are commonly added to the hydraulic fracturing fluid to increase the efficiency of natural gas extraction from the low-permeable shale. Although the reactions of these compounds in surface waters are well documented, their mobility and fate downhole is largely unknown.

The objective of this work was to study the hydrolysis of hydraulic fracturing fluid additives and their interaction with pyrite. Pyrite is a ubiquitous mineral of the Marcellus Shale lithostratigraphy. It is also known to be redox active, and is capable of catalyzing reactions with organic compounds. In this study, the reported solubility, volatility, and hydrophobicity of the most common hydraulic fracturing ingredients was used to assess which compounds are mobile enough to react downhole. The hydrolysis of these selected hydraulic fracturing chemicals was studied, and then each compound was added to an aqueous suspension of pyrite. The transformation products were monitored over time using liquid chromatography triple quadrupole mass spectrometry (LC-QQQ). Preliminary results showed that pyrite catalyzed reactions with the biocide dazomet, producing different products from those observed in the literature for surface waters. This is indicative of pyrite’s reactivity and the need for further understanding of its behavior with hydraulic fracturing fluids. The results of this study will increase the energy sector’s understanding of the fate and efficacy of the chemicals used, and will inform wastewater management schemes of methods to dispose of fracking flowback water.

Indicators of Impacts of Hydraulic Fracturing Fluids on Groundwater Quality

Dale Van Stempvoort

Environment Canada is conducting research on the impacts of shale gas development on groundwater in Canada. Following a literature review to establish the state of the science, the current focus is on laboratory activities to develop suitable indicators to probe for impacts of hydraulic fracturing fluids (FF). Information posted at fracfocus.ca was used along with other relevant information to screen and select candidate FF chemicals for testing as indicators. This includes development of new analytical methods and using batch and column experiments with simulated fracturing fluids to investigate the persistence and mobility of FF chemicals in the subsurface, and their tendencies with respect to sorption and biodegradation. The goal is to select one or more suitable tracers to study the possible migration of FF into groundwater resources. An overview of the laboratory methodology and preliminary results will be presented at the workshop.

Passive Grab Sampling for Dissolved Methane in Deep Groundwater

Kathleen A. Mihm, PG

Groundwater sampling for methane in deep wells is challenging due to difficulties related to high hydraulic pressure, degassing, and physical retrieval of representative samples. At a site in Mississippi, where groundwater is saturated with natural gas at depths of up to 700 feet below ground surface and under hydraulic pressure of up to 180 psi, two passive groundwater sampling methods were used to collect water samples for methane analyses: the Snap Sampler and the Solinst Discrete Interval Sampler.

Several rounds of groundwater samples were collected from site wells to assess the passive groundwater samplers and to characterize the deep groundwater. Methane concentrations measured for samples collected using the Snap Sampler and the Solinst Discrete Interval Sampler method closely approximated calculated methane-saturated concentrations at depth. Degassing within the Snap Sampler vials did occur upon retrieval because the vials are not pressure sealed; however, because degassed vapor and fluid is contained within the vessel, the methane concentrations measured in the Snap vials appear to be representative of downhole concentrations. Split samples of Snap Sampler vials indicated inter-laboratory variability, which was partially attributed to inconsistency in laboratory protocols. Based on experimentation, we found that several approaches were effective in obtaining deep groundwater samples for groundwater with dissolved methane at concentrations of tens to hundreds of mg/L.

The Science-Policy-Technology Nexus of Next Generation Water Quality Regulation

Heather Dawn Gingerich

The quandary of assessing human health effects of hydraulic fracturing is a textbook example of what happens when regulatory policies are driven by myopic economic interests (as opposed to a solid foundation in science) and uniformly fail to keep pace with technological advancement in extractive industries that is explored by participants of the Water School for Decision-Makers (WiSDoM), a pilot program for municipal, provincial, and federal-level politicians in Ontario, Canada. Although provincial drinking water standards are considered to be world-class, the current reliance on outdated, inappropriate, and overly restrictive “target lists” provides a government-sanctioned smokescreen for environmental pollution and threatens to destabilize the economy through preventable healthcare expenditures and flawed corporate decision-making. An alternate strategy of applying selected modern technologies developed for (1) genetic testing, (2) mineral prospecting, and (3) the Olympic Drug Squad to completely characterize water resources (as a pre-drilling baseline and during ongoing monitoring and maintenance) is described within the framework of a revised approvals process designed for the North American watershed and informed by key principles of indigenous traditional knowledge. The comprehensive dataset thus efficiently and affordably generated could be made publicly available for multiple uses including, but not limited to, integrated environmental protection policy-making at all tiers, impetus for the development of new water treatment technologies, and preventive healthcare. The digital platform could also support the emerging industry of consumer decision support tools like Baby Bear Care (the world’s first medical geology smartphone application) which will be briefly demonstrated.

Understanding Key Elements of Past Subsurface Fluid Releases to Better Inform Future Groundwater Monitoring

Pete Penoyer, Hydrogeologist

This presenter has determined that impacts to shallow groundwater from oil and gas development are best looked at in one of two ways. The first is conventional surface/near surface, primarily vadose zone releases/spills of fluids from oil and gas surface operations (tanks, gathering lines, drilling-related reserve pits, retention ponds, transportation accidents, etc.). These releases will typically behave similar to fluid releases by other industries. Largely driven downward by gravity, the groundwater industry has a long history dealing with such spills/releases through site characterization, standard monitoring practices, and remediation of any resulting groundwater plumes. The second category is subsurface releases of fluids that typically are driven upward from some depth under a pressure gradient or relative density difference of the contaminating fluid, or both. In the latter type of release the groundwater industry has much less experience in terms of understanding the numerous elements that may come into play. Often there is a lack of well drilling, installation, and operational information in addition to limited deeper subsurface geologic or seismic data that is not readily available due to the confidential and proprietary data nature of the E&P (upstream) industry.

This presentation attempts to briefly summarize what is understood about the local geology, migration pathways, primary COCs, drive mechanisms, wellbore integrity, etc. that may have been contributing factors to some alleged/documented release incidents. Those fluid releases of hydrocarbons/brine that have gained national notoriety (e.g., PA, OH, TX, WY, CO, ND) suggest the actual role played by the hydraulic fracturing process itself is largely unrelated and the focus on the induced fracture pathway through/beyond the target formation appears unwarranted (“frac outs” excepted). Only through a better understanding of such past documented subsurface fluid releases may future groundwater monitoring be best informed and focused to cost effectively detect subsurface releases to shallow groundwater.