&EPA
United States
Environmental Protection
Agency
EPA 600/R-l 1/066 I May 2011 I www.epa.gov
Proceedings of the Technical Workshops
for the Hydraulic Fracturing Study:
Chemical & Analytical Methods
U.S. Environmental Protection Agency
-------
EPA 600/R-11/066
May 2011
United States
Environmental Protection
Agency
Proceedings of the Technical Workshops
for the Hydraulic Fracturing Study:
Chemical & Analytical Methods
Office of Research and Development
US Environmental Protection Agency
Washington, DC
-------
Table of Contents
List of Figures ii
List of Tables iv
Introduction 1
Workshop Participants 3
Agenda 5
Summary and Abstracts from Theme 1: Fracture Fluid Chemistry 8
Summary of Presentations for Theme 1: Fracture Fluid Chemistry 9
Summary of Discussions Following Theme 1: Fracture Fluid Chemistry Presentations 11
Abstracts for Theme 1: Fracture Fluid Chemistry 14
Hydraulic Fracturing Fluid Considerations in Marcellus Shale Completions 15
High Rate Hydraulic Fracturing Additives in Non-Marcellus Unconventional Shales 17
Unconventional Fracturing Fluids 22
Produced Formation Water Sample Results from Shale Plays 34
Trace Metal Chemistry and Mobility in the Marcellus Shale 38
Fracture Fluid Additive and Formation Degradations 40
Evaluating Interactions of Fracturing Fluids and Degradation Products with Radionuclides Contained in Organic-
rich Solid-Phase Host Materials 43
Summary and Abstracts from Theme 2: Chemical Fingerprinting 45
Summary of Presentations for Theme 2: Chemical Fingerprinting 46
Summary of Discussions Following Theme 2: Chemical Fingerprinting Presentations 48
Abstracts for Theme 2: Fracture Design and Stimulation 52
Chemical and Isotopic Tracers of Natural Gas and Formation Waters in Fractured Shales 53
Distinguishing the Source of Natural Gas Accumulations with a Combined Gas and Co-produced Formation Water
Geochemical Approach 60
The Origin of Some Natural Gases in Permian through Devonian Age Systems in the Appalachian Basin & the
Relationship to Incidents of Stray Gas Migration 64
Tracking Fracture Fluid Movement with Chemical and Gamma-emitting Tracers with Verification by Microseismic
Recording 70
Designing a Water Quality Program for Shale Exploration 76
Elucidating Water Contamination by Fracturing Fluids and Formation Waters from Gas Wells: Integrating Isotopic
and Geochemical Tracers 81
Summary and Abstracts from Theme 3: Field and Analytical Challenges 84
Summary of Presentations from Theme 3: Field and Analytical Challenges 85
Summary of Discussions Following Theme 3: Field and Analytical Challenges Presentations 86
Abstracts for Theme 3: Field and Analytical Challenges 89
Quality Assurance, Quality Control and Method Performance Considerations for Chemical Testing of
Environmental Samples Impacted by Hydraulic Fracturing Fluids 90
Determination of Total Organic Carbons in Difficult Sample Matrices Utilizing the Supercritical Water-Oxidation
TOC Procedure 98
Practical Quantitation, Method Detection Limits, Interferences and Dilution Challenges 103
Challenges for Precise Radium Analysis in Brine 109
Revisiting the Major Discussion Points of the Technical Presentation Sessions Ill
Summary of Discussions Following Workshop and Theme Lead Summaries 112
Glossary of Terms 114
-------
List of Figures
Figure 1. Effect of capillarity on water saturation (after Holditch, 1979) 33
Figure 2. Effect of water imbibtion on relative permeability changes (after Keelan, 1975) 33
Figure 3. Cross-section of Illinois Basin from Schlegel et al. (in press) 53
Figure 4. Methane to ethane + propane ratios versus carbon isotopes of methane for various
gas 54
Figure 5. Carbon isotopes of carbon dioxide versus methane for various gas sources 55
Figure 6. Chloride vs. bromide concentrations (log-scale) of formation waters associated with
various gas sources 56
Figure 7. Stable isotopes of formation waters associated with natural gas accumulations 56
Figure 8. Carbon isotopes of dissolved inorganic carbon (DIC) versus alkalinity concentrations of
formation waters associated with natural gas accumulations 57
Figure 9. Isotope geochemistry for Marcellus Shale gas have expanded our view to reveal more
thermally mature thermogenic methane than fit within the standard isotope crossplot, and that
also reveal isotope reversals (Cl, C2) throughout the stratigraphic section. Similarly, stable
carbon and hydrogen isotope compositions reveal a broader range for early thermogenic
methane 66
Figure 10 provides a small dataset of a much broader database documenting isotope
compositions of thermogenic gases in areas across the Central and Northern Appalachian basin.
In addition, recent geochemistry for Marcellus shale gas (unpublished data, Baldassare, 2011)
and deeper natural gas resources in areas of the Appalachian basin reveal a complicated
thermogenic history that cannot be explained by current hydrocarbon generation models
(Burruss, Laughrey, 2010) 67
Figure 11 provides a small dataset of a much broader database documenting isotope
compositions of thermogenic gases in areas across the Central and Northern Appalachian basin.
In addition, recent geochemistry for Marcellus shale gas (unpublished data, Baldassare, 2011)
and deeper natural gas resources in areas of the Appalachian basin reveal a complicated
thermogenic history that cannot be explained by current hydrocarbon generation models
(Burruss, Laughrey, 2010) 67
Figure 12. A sequential or zipper fraced well pair in Tier 2 of the Barnett was drilled in a NNW
direction in an area with possible faulting in the Ellenberger (immediately below the Barnett)
with general primary frac direction of NE/SW. Well T1H has 8 stages; well T2h has 10 stages.
Each stage was approximately 200,000 Ib 100 mesh sand (1000 Ib/ft); 7400 bbls water ( 40
bbls/ft of lateral); 6 perf clusters per frac stage; 1 bpm/perf; design frac rate was 50 bpm 72
Figure 13 shows an unexplained pumping pressure break at about 60 minutes 72
Figure 14. Near identical primary and secondary frac behavior on the T1H stage 1. The breakout
pressure shift was hidden in the ramp up but occurs at about 45 minutes 73
Figure 15 for T1H (left) and T2H (right) are the mass balances of fluid tracer flow coupled with
information on frac breakdown from proppant tracer. The information from fluid tracing shows
CFT concentration measurements in the backflow against volume produced (left side of each
figure) are useful for estimating communication of the frac stage with the wellbore and how
-------
that communication changes over time or volume produced. The left side of each figure is the
amount of initial fluid tracer recovered from each traced stage and is a reflection of the amount
of traced load fluid actually recovered. The shaded area in the center is the contribution of each
stage to the flow as calculated from the mass balance 74
-------
List of Tables
Table 1. Physical data for four shales 17
Table 2. Selected Metal Concentrations in Marcellus Shale 38
Table 3. Parameters Tested in Surface Water and Residential Wells 77
Table 4. Selected Flowback Constituents 78
Table 5. Comparison of Select Parameters from Utica Shale Stimulation Results 79
Table 6. Selected Indicator Parameters 79
Table 7. Types of QC Samples Expected in Mature QA Programs 94
Table 8. Samples evaluated for SCWO-TOC 101
Table 9. Results of SCWO-TOC analyses 101
Table 10. Typical Methods 104
Table 11. 500 Series Methods 104
Table 12. 600 Series Methods 105
Table 13. SW846 Series 108
Table 14. Inorganic Parameters 108
IV
-------
Introduction
The Hydraulic Fracturing Study
In its Fiscal Year 2010 budget report, the U.S. House of Representatives Appropriation
Conference Committee identified the need for a study of the potential impacts of hydraulic
fracturing (HF) on drinking water resources. The Committee directed EPA scientists to
undertake a study of HF to better understand any potential impacts of hydraulic fracturing on
drinking water and ground water. The EPA produced a draft study plan, which focuses on the
key stages of the HF water lifecycle: water acquisition, chemical mixing, well injection, flowback
and produced water, and wastewater treatment and waste disposal. This plan was submitted to
the Science Advisory Board (SAB) in February 2011 and the peer review of the plan was held on
March 7 and 8, 2011. At the time these technical workshop proceedings were developed, the
SAB had not given its official review to EPA.
EPA has included stakeholder concerns in the planning process of the study from its inception,
engaging stakeholders in a dialogue about the study through a series of webinars and facilitated
public meetings held between May and September 2010. EPA also held four technical
workshops in February and March 2011 to explore the following focus areas: Chemical &
Analytical Methods (February 24-25), Well Construction & Operations (March 10-11), Fate &
Transport (March 28-29) and Water Resource Management (March 29-30).
The technical workshops centered around three goals: (1) inform EPA of the current technology
and practices being used in hydraulic fracturing, (2) identify research related to the potential
impacts of hydraulic fracturing on drinking water resources, and (3) provide an opportunity for
EPA scientists to interact with technical experts. EPA invited technical experts from the oil and
natural gas industry, consulting firms, laboratories, state and federal agencies, and
environmental organizations to participate in the workshops. EPA will use the information
presented in this document to inform research that effectively evaluates the relationship
between HF and drinking water.
An initial report of results from the EPA's study is expected by late 2012, with an additional
report expected in 2014.
About the Proceedings
These proceedings provide information related to the 25 presentations given on chemical and
analytical methods at the Technical Workshop for the U.S. EPA Hydraulic Fracturing Study held
on February 24-25, 2011. This workshop consisted of three sessions or themes: Theme 1-
Fracturing Fluid Chemistry; Theme 2-Fingerprinting; and Theme 3-Field and Analytical
Challenges. The proceedings include abstracts of the presentations and a summary of the
discussions that took place during the workshop. The presentations from this workshop are not
part of the proceedings document, but may be accessed at http://epa.gov/hydraulicfracturing.
-------
This is the first of four technical workshops on topics relating to the EPA Hydraulic Fracturing
Study. The other three workshops are: Well Construction and Operations (Mar. 10-11), Fate
and Transport (Mar. 28-29), and Water Resources Management (Mar. 29-30). Proceedings will
be available separately for the other three workshops.
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
Editorial Team for the Proceedings
The attendees at the Chemical & Analytical Methods workshop were selected based on
information submitted to EPA during the attendee nomination process. Presenters, a workshop
lead, and theme leads were selected from the pool of attendees, once again, based on the
information submitted to EPA during the attendee nomination process. The workshop lead,
Wilma Subra of Subra Company, assisted EPA in finalizing details for the workshop and served
as the lead editor of the proceedings document. The theme leadsCal Cooper of Apache
Corporation and Tracy Bank of University at Buffalo for Theme 1, Jennifer Mclntosh of
University of Arizona for Theme 2, and Kesavalu Bagawandoss of Accutest Labs for Theme 3
served as editors for their respective themes.
-------
Workshop Participants
Name
Affiliation
Michael
Kesavalu
Fred
Tracy
Doug
Chrystal
Uni
Jeanne
Susan
Nancy
Cal
Robin
Brian
Robin
Jill
Dennis
Dom
Mark
Thomas
Ruth
Bill
Maria
Jennie
D V Satya
Richard
Fred
Richard
Dawn
George
Carl
Matthew
Oliver
Michael
Rick
Jennifer
Keith
Arthur
Bagawandoss
Baldassare
Bank
Beak
Beasley
Blake
Briskin
Burden
Coleman
Cooper
Costas
D'Amico
Danesi
Dean
Degner
Diguilio
Engle
Fikslin
Forman
Godsey
Gomez-Taylor
Gunderson
Gupta
Hammack
Hauchman
Hodge
Kaback
King
Kirby
Land is
Lawal
Lawson
McCurdy
Mclntosh
McLeroy
Pennsylvania State University
Accutest Laboratories
Echelon Applied Geoscience Consulting
University at Buffalo
US Environmental Protection Agency
US Environmental Protection Agency
MajiTox for Gastem USA
US Environmental Protection Agency
US Environmental Protection Agency
Chesapeake Energy Corporation
Apache Corporation
US Environmental Protection Agency
US Environmental Protection Agency
US Environmental Protection Agency
US Environmental Protection Agency
Range Resources
US Environmental Protection Agency
US Geological Survey
Delaware River Basin Commission
Environmental Standards, Inc.
Geo Logic Environmental Services, LLC
US Environmental Protection Agency
US Environmental Protection Agency
B J Services Company - Baker Hughes
US Department of Energy, National Energy Technology
Laboratory
US Environmental Protection Agency
ConocoPhillips
AMEC Geomatrix, Inc.
Apache Corporation
Bucknell University
US Environmental Protection Agency
Aquionics Inc
ExxonMobil
Chesapeake Energy Corporation
University of Arizona
Texas A&M University
-------
Richard
Matt
Greg
Stephen
Robert
lleana
Terry
Elisabeth
Brian
Harry
Subodh
Mark
Wilma
Jerry
Zoltan
Kevin
David
Denise
Jack
Avner
Rock
Frank
Rick
Nathan
Name
McMillin
Miller
Oberley
Osborn
Puls
Rhodes
Romanko
Rowan
Schumacher
Schurr III
Singh
Stebbins
Subra
Swearingen
Szabo
Tanner
Thai
Tuck, P.E.
Tuschall
Vengosh
Vitale
Walles
Wilkin
Wiser
Affiliation
US Environmental Protection Agency
Schlumberger
US Environmental Protection Agency
Duke University
US Environmental Protection Agency
Shell Global Solutions US Inc.,
TestAmerica Laboratories, Inc.
US Geological Survey
US Environmental Protection Agency
CNX Gas/CONSOL Energy
Sinte Gleska University
CNX Gas Company LLC
Subra Company
Baker Hughes
US Geological Survey
Pioneer Natural Resources USA Inc.
Environmental Standards, Inc.
Halliburton Energy Services, inc.
TestAmerica Laboratories Inc
Duke University
Environmental Standards, Inc.
Devon Energy
US Environmental Protection Agency
US Environmental Protection Agency
-------
A -6
M
Agenda
Technical Workshops for the Hydraulic Fracturing Study
Chemical & Analytical Methods February 24-25, 2011
US EPA Conference Center
One Potomac Yard (South Building)
2777 S. Crystal Drive
Arlington, VA 22202 Room S-4370 and 4380
February 24, 2011
7:30 am Registration
8:00 am Welcome
Fred Hauchman, Director, Office of Science Policy, EPA Office of Research & Development
Wilma Subra, Subra Company, Workshop Lead
Pat Field, Facilitator, Consensus Building Institute
Theme 1: Fracture Fluid Chemistry
8:15 am Technical Presentation Session 1: Fracture Fluid Formulations and Purposes
HF Fluid Considerations in Marcellus Shale Completions, Dennis Degner, Range Resources
High Rate HF in Non-Marcellus Unconventional Shale, Rick McCurdy, Chesapeake Energy
Cross-link and Linear Gel Composition, Richard Hodge, ConocoPhillips
Old vs. New Trends in HF Chemicals, Denise Tuck, Halliburton
Unconventional Fracturing Fluids: What, Where, Why?, Satya Gupta, BJ Services- Baker
Hughes
9:45 am Break
9:55 am Technical Presentation Session 2: Fracture Fluid Interaction with Host Materials
Inorganic Chemistry of Produced Water in the Appalachian Basin, Elizabeth Rowan, USGS
Produced Formation Water Sample Results from Shale Plays, Nancy Coleman, Chesapeake
Energy
Trace Metal Geochemistry and Mobility in the Marcellus Formation, Tracy Bank, University
at Buffalo
Fracture Fluid Additives and Formation Degradations, George King, Apache Corporation
-------
11:10 am Break
11:20 am Technical Presentation Session 3: Fracturing Fluid Degradation Products
What Can You Expect or Not-Chemical Breakdown and Degradation, Denise Tuck,
Halliburton
Evaluation of Radionuclides as Degradation Products of Host Materials in Response to
Altered Fluid Chemistry Environment, Zoltan Szabo, USGS
12:30 pm Lunch
Theme 2: Fingerprinting
1:30 pm Technical Presentation Session 4: Determining Ambient Groundwater Conditions
Chemical and Isotopic Tracers of Natural Gas and Formation Waters in Fractured Shales,
Jennifer Mclntosh, University of Arizona
Distinguishing the Source of Natural Gas Accumulations with a Combined Gas and Co-
Produced Formation Water Geochemical Approach: A Case Study from the Appalachian
Basin, Stephen Osborn, Duke University
The Origin of Some Natural Gases in Permian through Devonian Age Systems in the
Northern Appalachian Basin and the Relationship to Incidents of Stray Gas Migration, Fred
Baldassare, Echelon Applied Geoscience Consulting
2:45 pm Break
2:55 pm Technical Presentations 5: Introduced Tracers
Tracing Fracturing Fluid Movement with Chemical and Gamma-Emitting Tracers with
Verification by Microseismic Recording, George King, Apache Corporation
Designing a Water Quality Program for Shale Exploration, Uni Blake, Gastem USA
3:55 pm Break
4:05 pm Technical Presentations 6: Tracing Fracturing Fluids in the Environment
Elucidating Water Contamination by Fracturing Fluids and Formation Waters from Gas
Wells: Integrating Isotopic and Geochemical Fingerprints, Avner Vengosh, Duke University
Isotopic Tracing of Groundwater Contamination: Techniques, Applications, and Practical
Considerations, Michael Lawson, ExxonMobil
Inorganic Geochemistry of Pennsylvania Marcellus Flowback Waters, Carl Kirby, Bucknell
University
-------
5:20 Revisit the Major Discussion Points of the Technical Presentation Sessions
Wilma Subra, Subra Company, Workshop Lead
Cal Cooper, Apache Corporation, and Tracy Bank, University of Buffalo, Theme Leads -
Fracture Fluid Chemistry
Jennifer Mclntosh, University of Arizona, Theme Lead - Fingerprinting
6:00 pm Adjourn for the Day
February 25, 2011
Theme 3: Field and Analytical Challenges
8:00 am Technical Presentation Session 7:
Sampling Issues- Representativeness, Handling, Preservation
Representativeness of Flowback Fluid Samples: Impacts of Post-Withdrawal Evolution and
Temporal Variability of Composition, Mark Engle, USGS
Practical Quantitation and Method Reporting Limits
Quality Assurance, Quality Control and Method Performance Considerations for Chemical
Testing: Samples Impacted by Hydraulic Fracturing Fluids, David Thai, Environmental
Standards, Inc.
Determination of Total Organic Carbons (TOC) in Difficult Sample Matrices Utilizaing the
Supercritical Water-Oxidation TOC Prodedure, Keith McLeroy, Texas A&M University
Practical Quantitation Limits, Method Detection Limits, Interferences and Dilution
Challenges in Fracturing Fluid Analyses, Kesavalu M. Bagawandoss, Accutest Labs
9:30 am Break
9:40 am Technical Presentation Session 8: Interference and Dilution Challenges Including
Radionuclides
Radiochemical Analytical Challenges with Hydraulic Fracturing Fluids, Terry Romanko,
TestAmerica Labs
Analytical Challenges of Radionuclides in Brines, Zoltan Szabo, USGS
10:55 am Revisit the Major Discussion Points of the Technical Presentation Session
Wilma Subra, Subra Company, Workshop Lead
Kesavalu M. Bagawandoss, Accutest Labs, Theme Lead - Field and Analytical Challenges
12:00 pm Closing Discussions
Jeanne Briskin, Hydraulic Fracturing Research Task Force Leader, Office of Science Policy,
EPA Office of Research & Development
Wilma Subra, Subra Company, Workshop Lead
-------
Summary and Abstracts from Theme 1: Fracture Fluid
Chemistry
-------
Summary of Presentations for Theme 1: Fracture Fluid Chemistry
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
Technical Presentations
The first set of technical presentations in this theme addressed the formulation and use of
fracture fluids.
Dennis Degner, Range Resources, discussed the purposes of and need for the most common
chemical additives used in Marcellus Shale fracture treatments. Fracture fluids in the Marcellus,
according to Mr. Degner, are typically "slickwater" designed fracture treatments, which are
water-based fracture fluids. Mr. Degner identified the most common fracture fluid additives as
friction reducers, biocides, and scale inhibitors. These chemicals are added in small
concentrations to the fracturing fluid and are monitored throughout the fracturing operations
to ensure that the correct concentrations are maintained. Mr. Degner believes monitoring of
chemical concentrations is important when produced water is recycled and reused for
fracturing.
Rick McCurdy, Chesapeake Energy, discussed the most common chemicals used in fracture fluid
in four additional shale plays (Fayetteville, Barnett, Eagle Ford, and Haynesville). Mr. McCurdy
discussed the functions of potassium chloride substitute, quaternary amine surfactant,
hydrochloric acid (HCI), acid inhibitor, iron control agent, gel, crosslinkers, and breakers in
fracturing fluid. Mr. McCurdy described Chesapeake Energy's Green Frac program, which was
implemented to evaluate potential environmentally-friendly alternatives to some additives, as
well as the potential for eliminating non-critical additives. Specific chemicals added to
fracturing fluids were identified, but the range of concentrations of these chemicals was not
discussed.
Richard Hodge, Conoco Phillips, discussed the composition of crosslinked and linear gels. Mr.
Hodge stated that gelling agents and crosslinkers in fracture fluids increase the molecular
weight of polymers in the fluids to increase fluid viscosity for improved transportation and
delivery of the proppants into the induced fractures. Breakers are then added into these
fracture fluids to "break up" (reduce the viscosity of) the crosslinked polymers at the bottom of
the wellbore to promote return of the fracture fluids to the surface.
DeniseTuck, Halliburton, discussed technological advances in HF. Historically, there had been a
large number of oil-based operations, but these decreased as the industry focused more on
water-based operations, according to Ms. Tuck. The newest technology used by Halliburton is
an advanced dry polymer blending. Halliburton uses a scoring index to rank their additives
based on health, safety, and environmental impact. Ms. Tuck stated that new fracturing fluid
designs used by Halliburton involve chemicals sourced from the food industry.
-------
Satya Gupta, BJ Services-Baker Hughes, discussed the use of unconventional fracture fluids in
tight gas formations with adverse capillary effects. In these conditions, water-based fracturing
fluids could be replaced by alcohol-based fluids, hydrocarbon-based fluids, or liquid carbon-
based fluids, or enhanced by surfactants.
The second set of technical presentations addressed fracture fluid interaction with host
materials.
Elizabeth Rowan, United States Geological Survey (USGS), discussed the chemical composition
of formation waters in the Marcellus Shale. Her research found that total dissolved solid (IDS)
values in produced water from the Marcellus are similar to that of the formation waters in
older formations, as well as some Middle-Upper Devonian sandstones; the waters were found
to be high in sodium chloride and calcium chloride, and low in sulfate and carbonate. The
correlation between radium concentration and IDS suggests that salinity could possibly be used
as a predictive tool in the Appalachian Basin. Dr. Rowan observed sodium/bromide and
calcium/bromide ratios that suggest evaporated seawater may be the dominant salinity source.
Nancy Coleman, Chesapeake Energy, discussed a review of produced formation water data
from four shale formations (Marcellus, Barnett, Fayetteville, and Haynesville) to support
Chesapeake Energy's baseline water sampling program. The review found that produced
formation waters appear to be variable within and between formations, but that some
generalizations can be drawn. The concentrations of TDS and divalent cations, such as barium
and strontium, appear to increase with time after fracturing and appear to correlate with each
other. Ms. Coleman believes radium-226 and radium-228 are the most useful parameters for
gaining information about radionuclides. Ms. Coleman also outlined the most common volatile
organic compounds (VOCs) and semi-volatile organic compounds (SVOCs) in produced
formation water and noted that their concentrations are can vary but are typically low.
Tracy Bank, University at Buffalo, presented a study investigating the trace metal geochemistry
of the Marcellus Shale through core samples, well cuttings, and outcrop samples. The study
found high concentrations of zinc throughout the Marcellus and no significant differences
between the concentrations of most metals in core versus outcrop samples. The study further
described the correlation of uranium with total organic compounds throughout the formation
(using the new TOP SIMS element mapping technique) and also suggested that there may be
some positive correlation between uranium and total organic carbon. Also, extraction studies
showed that metals, including uranium, could be mobilized from the shale in fluid-rock
interactions.
George King, Apache Corporation, discussed the chemical and physical reactions that can occur
in the open wellbore, induced fractures, natural fractures, and the surrounding matrix. These
reactions occur as a result of the chemical interactions between fracture fluids and the geologic
target formations during the HF process. One of the limiting factors on these reactions is the
10
-------
low permeability of shale, which, according to Mr. King, limits the access of fracture fluids to
the formation materials. He also summarized the decomposition reactions of the most common
fracture fluid additives, and noted another reaction variable to consider is re-precipitation of
compounds over time.
The third set of technical presentations addressed fracture fluid degradation products.
Denise Tuck, Halliburton, discussed broad classes of compounds that can derive from native
reservoir fluid, fracture fluid, or raw water. Depending on downhole conditions and water/fluid
composition, Ms. Tuck stated that the following classes of chemicals can be found: salts,
metals, hydrocarbons, sulfides/sulfates, carbonates/bicarbonates, biocides, pesticides, and
naturally occurring radioactive material (NORM).
Zoltan Szabo, USGS, introduced a number of techniques for evaluating the presence of
radionuclides in produced waters including: isotope ratios, degradation experiments,
microfocused X-ray fluorescence, and fission-track autoradiography. According to Dr. Szabo,
further studies should be be based on existing information of NORM distribution in solids to
assess what NORM might be expected in produced water. Due to high variability between rock
formations, Dr. Sazbo believes there may not be one single technique that will work in every
situation for identifying NORM.
Summary of Discussions Following Theme 1: Fracture Fluid Chemistry
Presentations
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
Cross/inked systems and unconventional fluids. The presenters noted that the purpose of using
crosslinked gels is to increase the viscosity of fracturing fluids containing a heavier, coarser
proppant while minimizing the amount of gel used. At the end of a fracture job using a
crosslinked system, participants discussed the use of breakers to reduce the size of the
polymers, thereby reducing the viscosity of the fluids and promoting the return of fracture
fluids to the surface. The presenters stated that foams are created immediately prior to
injection and continuously from two lines (one containing the foaming agent and one
containing the fluid or gas) located at the wellhead. The presenters also stated that diesel has
not been widely used as a fracturing fluid in the United States since the 1970s. A participant
mentioned that analyses for diesel range hydrocarbons (DRH) may not be an accurate
assessment of hydraulic fracturing impacts to groundwater because DRH is often found in
flowback water though they are not part of the fracturing fluid.
High pH conditions in the wellbore. Participants discussed the pH conditions in the wellbore and
strategies of managing pH. Raising the pH of HF fluids is accomplished in various ways according
11
-------
to participants, such as the addition of sodium hydroxide or potassium hydroxide; potassium
hydroxide, in particular, is often used in conjunction with borate fluids. In addition, the
presenters stated that when brines of different composition are mixed for injection, reactions
occur that can lead to the precipitation of compounds from the fluids.
Viscoelastic surfactant-based (VES) systems. The presenters noted that a majority of VES system
use is in Canada. One presenter stated that very few jobs (<5%) in the United States use this
type of HF system.
Recycling. A number of workshop participants asked questions and expressed concern about
the repeated reuse of produced water. The presenters emphasized that the fracture fluid
composition is monitored and fewer chemicals (such as friction reducers and scale inhibitors)
might be added based on what monitoring identifies as already present in the (recycled) fluid
from the initial injection. One participant noted that much less fluid is recovered than is
injected into the well. The fate of the water not recovered depends on the formation. Many
companies test produced water from their wells 30, 60, and 90 days after injection and
throughout the life of the well, according to one participant. High salinity sample matrix and
lack of appropriate methodology were mentioned as analytical limitations to analyzing
flowback solutions for metals. In the case of testing for biocides in the flowback water,
participants claimed the success of the initial treatment is assessed and follow-up treatment of
the well is performed as necessary. The presenters stated that there are limiting factors to the
reuse of flowback water, such as the changing downhole conditions that may affect viability of
recycled flowback water or the (high) cost of treatment systems to appropriately treat the
water for reuse.
Similarity of fluid chemistries. The presenters noted that the typical fracture fluid chemistries
used across the industry are very similar with minor differences due to proprietary additives.
Presenters noted that when unconventional fracture fluids are employed, the composition of
these fluids is usually disclosed. Several participants indicated that the specific fluid
composition used in a particular fracture job is highly dependent on the geologic formation and
its compositional properties and also on location-specific physical properties such as depth,
pressure, temperature and geomechanical properties.
USGS study. Participants asked specific questions about the USGS study described in Dr.
Rowan's presentation. In this study, radium concentrations were measured using gamma ray
spectrometry. Radium was present in formation waters below its solubility limit and is not
affected by changes in temperature and pressure (which can accompany extraction of flowback
water). Thorium has been investigated in previous studies but was only found at levels below
the detection level and was not included in the study as a possible tracer. From a comparison of
calcium bromide versus sodium bromide concentrations, USGS found that a large component of
the salt in the Marcellus Shale is likely derived from evaporated ancient seawater rather than
from halide dissolution. Only the VOCs and SVOCs mentioned in Dr. Rowan's presentation were
targeted for analysis.
12
-------
Chesapeake Energy's Produced Water Sampling program. A participant asked about glycols with
regard to the Chesapeake review. Although glycols were investigated as part of the review,
Chesapeake is not satisfied with the analytical methods used and was unable to draw
conclusions about the occurrence of glycols in the produced formation waters from the target
shale formations studied. Total as well as dissolved arsenic was investigated by Chesapeake and
found to be below the detection limits for most samples.
University at Buffalo study. A participant asked about the HCI extractions described in Dr.
Bank's presentation. Dr. Bank explained that the determination of the time periods for HCI
extractions of uranium was estimated based on industry practice. The HCI extractions
demonstrated different metal extractabilities between core and outcrop samples. Differences
among organic carbon compounds in the shale were not investigated in Dr. Bank's study.
Radium inflowback water and as a possible tracer. Participants suggested that treatment of
radium-containing flowback water could, due to the half-life of radium, cause a buildup of
radium in the treatment sludge and pose a solid waste disposal issue. Some participants
suggested that radium could be used as a potential tracer because the isotopic ratio of radium
species in the Marcellus is different from the ratio in other aquifers. Other participants noted
that there could be problems distinguishing the source of a radium tracer (e.g., from the
Marcellus or from other subsurface sources) in the Marcellus Shale (which also has different
radium isotopic ratios than those found in the shallower black shale horizons). In addition,
radium can be mobile in groundwater. Participants noted that is important to have baseline and
background information on formation waters so that the composition of flowback water (a
composite of fracture fluids, native formation fluids, and any products of the interaction of the
two) can be better understood. Participants indicated that uranium is less useful as a tracer due
to its much longer half-life. Analysis of well cuttings can indicate the chemicals, elements, or
compounds, in addition to uranium and radium, present in the formation. Participants stated
that uranium and radium decay products such as polonium and radon can be distributed in the
rock matrix and can be found in solution in formation fluids or flowback water.
Degradation products. Participants noted that some materials (e.g., endocrine disrupters and
carcinogens) are not present in flowback water and analytical tests could be conducted to
confirm this and reassure the public. Participants indicated that petroleum distillates found in
flowback water come from some carrier fluids in the fracturing fluids as well as from some
additives. Treatment of petroleum distillates can break down the hydrocarbons (e.g., mineral
oil) used as carriers in fracture fluids.
13
-------
Abstracts for Theme 1: Fracture Fluid Chemistry
Abstracts were submitted to U.S. EPA by the presenters for use in this proceedings document.
Not all presenters submitted abstracts of their presentations.
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
14
-------
Hydraulic Fracturing Fluid Considerations in Marcellus Shale
Completions
Dennis L. Degner
Range Resources Appalachia, LLC
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Hydraulic fracturing (completion) has become a key operation when developing shale gas
reservoirs. Like all industries and even families, the natural gas industry uses certain chemicals
as a normal part of its daily business. Chemicals used during the fracturing process are a vital
component to a successful completion. The chemicals used help reduce surface treating
pressures, aid in placement of the propping agent (sand) within the deep, downhole formation,
and help maintain fluid properties that meet design specifications. The chemicals most
commonly used in Marcellus Shale fracture treatments are friction reducer, biocide, and scale
inhibitor. These chemicals are mixed in very low concentrations with water (referred to as a
slickwater frac) and make up < 1% of the total job volume.
A friction reducer (FR) is one of the common chemicals used in Marcellus hydraulic fracturing
operations. The friction reducer is necessary to reduce the frictional effects (extra pressure)
that occur as water is pumped down the long pipe (wellbore) during the hydraulic fracture
treatment. In Marcellus completions, there is often 10,000 ft of pipe cemented into the ground.
Without using a friction reducer, the surface pressure would be higher than desired to reach
the required pump rate during the treatment. Friction reducers are commonly a polyacrylamide
polymer that is added at low concentrations (± 0.5 gal/Mgal). Once FR is added to the base
fracturing fluid the viscosity remains close to that of water (< 5 cps) while the frictional effects
are greatly reduced. FR often comes from the manufacturer in dry powder form, but is most
commonly pumped as a liquid by mixing with a mineral oil base fluid to stabilize the material
until it's injected into the water stream and pumped downhole. Historically they have worked
best with fresh water fluids, but recent advancements in technology have made it possible to
pump salt tolerant versions when re-using produced salty water in Marcellus completions. Flow
loop testing was conducted in the development of these new chemicals to aid in selection
based on fluid salinity. When using produced water in the completion, pre-job water testing is
done to verify which FR is required to ensure proper application. A polyacrylamide is commonly
used in many industries and can be found in children's toys, and used for soil stabilization in
addition to several other uses.
Hydraulic fracture treatments are pumped with various water sources ranging from ground,
surface, and produced water environments. Once the source water is identified, samples are
captured and tested to determine the amount of bacteria present. The source water and
ambient temperature can be strong drivers in the amount of bacteria present, similar to the
drivers controlling growth of pond algae. Based on the level of bacteria already present in the
15
-------
water a biocide chemical will be added to the water at a proper loading rate to prevent
microbiological influenced corrosion of the downhole pipe, formation souring (H2S), or other
safety/production concerns. Biocide is also added to prevent microbial growth from occurring
downhole which could restrict flow from the created hydraulic fracture network. Due to the
high biomass content already existing in the water sources used for Marcellus Shale fracturing,
an organic biocide is required to achieve the necessary sterilization. A similar chemical
approach can be seen in medical equipment sterilization and hand sanitation. Similar to the
friction reducer, the biocide is added in a liquid form to the water pumped downhole during the
hydraulic fracturing process. Once the well is completed, post-job samples are taken, and
testing performed to assess effectiveness of the treatment and for future design optimization.
Additional sampling and culture testing occur during the life of a well to ensure bacteria levels
remain low.
Based on minerals present in the various fresh and produced frac water sources, there is a
potential to create scale during production. Extensive water analysis performed prior to
fracturing can detect the tendencies to form scale and identify types most likely to form based
on the minerals present in the water and in the Marcellus Shale. Two common scales that can
generate in this environment are calcium carbonate (CaC03) and barium sulfate (BaS04). Both
are expensive to clean within the pipe if they occur, so the purpose of this chemical is to inhibit
their formation. CaC03 can be simple to remove with remedial chemical treatment while BaS04
requires mechanical removal. To prevent scale from generating, a solvent based scale inhibitor
is added into the water pumped downhole. As the volume of produced water used for
fracturing increases, the need for adequate scale control will also increase due to minerals
dissolving from the shale reservoir into the produced water. Scale inhibitors are added in low
concentrations (± 0.1 gal/Mgal) with low level residuals observed after fracturing that help
prevent scale from occurring. Similar applications can be seen in municipal water treatments,
de-icing agents and household cleaners.
Lastly, a surfactant chemical is pumped on a low percentage of Marcellus Shale completions.
The addition of a surfactant reduces surface tension on the fracture face making it easier to
recover produced fluids. Testing results have varied resulting in basin specific application.
Examples of surfactants can be found in soaps and foaming agents.
Each chemical is electronically monitored and manually strapped for accuracy during the
treatment ensuring proper loading. With site specific testing and new chemical development
with service companies, industry is able to optimize chemical application while reducing
environmental impact in hydraulic fracturing. Best practices with chemical use are shared with
industry partners, state agencies and trade groups for transparency and responsible chemical
use in Marcellus Shale completions.
16
-------
High Rate Hydraulic Fracturing Additives in Non-Marcellus
Unconventional Shales
Rick McCurdy
Chesapeake Energy Corporation
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
In a sister publication presented during this workshop, another operator discussed the use of
hydraulic fracturing additives in the Marcellus Shale. To avoid unnecessary repetition, this
paper will focus on additives used in four other major U.S. unconventional shale plays - the
Barnett Shale located in North Texas, the Eagle Ford Shale located in South Texas, the
Fayetteville Shale in Arkansas and the Haynesville Shale that lies beneath East Texas and
Northwestern Louisiana. The additives used in each play will be discussed, along with
comments regarding why they are used, the general chemistries involved with each and the
normal usage range in fracturing fluids. Finally, a discussion of the recent trends in the selection
of additives will be discussed.
Non-Marcellus Unconventional Shales
Of the four unconventional shales listed above, all are traditional gas plays with the exception
of the Eagle Ford. Depending on location, a well in the Eagle Ford can produce gas, gas and
condensate or gas and oil. The table below illustrates some comparable physical data for the
four shales.
Table 1. Physical data for four shales
Shale Play Fayetteville Barnett Eagle Ford Haynesville
Average Depth
From Surface (ft)
Bottom Hole Temp
(F)
Bottom Hole
Pressure (psi)
4,500
130
2,000
7,400
190
2,900
9,000
260
6,200
11,500
320
10,000
Of the four, the Fayetteville is the "shallowest" with an average depth of slightly less than one
(1) mile beneath the surface of the Earth. On the opposite end is the Haynesville which, on
average, can be found more than two (2) miles beneath the surface. As such, these formations
are separated from drinking water aquifers by thousands of feet of multiple geologic strata.
Often, temperature and pressure increase with depth and that is seen in the data above. The
Fayetteville has the lowest bottom hole temperature and pressure and the Haynesville the
17
-------
highest. Temperature has to be taken into account when selecting additives and concentrations
for hydraulic fracturing applications.
Typical Additives
In many articles published in various media over the last few years, the number of additives
used in hydraulic fracturing is often grotesquely over-stated. In a recent documentary, a
statement was made claiming that hydraulic fracturing fluids could contain as many as "596
toxic compounds". Some hydraulic fracturing activities may use as many as 11 additives, but
many times the procedures contain far fewer than that. The 11 that may possibly be used are:
friction reducer, biocide, scale inhibitor, potassium chloride (KCI) substitute, surfactant,
hydrochloric acid, acid inhibitor, iron control agent, gel, cross-linker and breaker (linear and
cross-linked gels, along with a few additives that are occasionally required in these fluid
systems, will be discussed by another author and will not be discussed in this paper). Following
are the purpose, typical chemistry and use concentrations for each additive.
Friction Reducers
This is the product responsible for the term "slick water". Friction reducers are used to reduce
interfacial tension between the fluid and the contact surface of the steel pipe the fluid is being
pumped through and to maintain laminar flow while pumping. Maintaining laminar, or non-
turbulent, flow is critical to achieving the fluid injection rates necessary for hydraulic fracturing
of unconventional formations, including shales. The active ingredient in friction reducers is
typically a medium to long chain polyacrylamide. When injected into fresh water, the
polyacrylamide hydrates and uncoils and prevents turbulent vortices in the moving water. An
average usage rate for most friction reducers is 500 to 1,000 parts per million (ppm) and they
are injected throughout the entire fluid. On a total fluid basis, they typically account for 0.05% -
0.1% of the total fluid volume pumped.
Biocides
The source water that makes up 99+% of the average hydraulic fracturing stimulation often
contains varying concentrations of bacteria that have the potential of causing problems with
the additives being used, the overall integrity of the wellbore and surface production
equipment and can impart undesirable contaminants to the produced gas stream. Many
bacteria can degrade the gels used for building fluid viscosity and have the potential to
aggressively attack the metal equipment used both downhole and on the surface for producing
natural gas and liquids from the well. Biocides are added to the source water to sanitize the
fluid and greatly reduce the concentration of bacteria.
There are several chemistries utilized as biocides in hydraulic fracturing. Historically,
glutaraldehyde and blends of glutaraldehyde have been used. Glutaraldehyde has been in use
in several industries for many years and has a long history as a disinfectant for medical and
dental equipment. Since the late 1990's, a product commonly referred to as THPS (tetrakis
hydroxymethyl phosphonium sulfate) has been used in many areas because it degrades in the
environment more rapidly that glutaraldehyde. Another product that is highly biodegradable is
DBNPA (2,2-dibromo,3-nitriloproprionamide). There are some operators who use sodium
18
-------
hypochlorite (the active ingredient in household bleach) for sanitation of source waters.
Regardless of the chemistry being used, it is important to note that all biocide usage is
regulated by the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA) and accordingly,
each product must be registered for use by the U.S. Environmental Protection Agency (EPA) and
by each State in which it will be applied. Biocides are commonly applied throughout the whole
hydraulic fracturing fluid at a rate of 75 to 500 ppm and represent 0.075% - 0.05% of the total
fluid volume.
Scale Inhibitors
Water used during the hydraulic fracture stimulation often has a potential of producing a
mineral scale when coming in contact with naturally occurring water in the producing
formation. Additionally, physical changes imparted on produced water (temperature, pressure,
etc.) during initial production can cause mineral solids to precipitate from the fluid. To prevent
this, Operators use a scale inhibitor injected throughout the hydraulic fracturing fluid. Owing to
their compatibility with other fracturing additives, the most commonly used chemistries are
carboxylic acid and acrylic acid polymers. Scale inhibitors are usually used at relatively low
dosages throughout the frac fluid (75 to 120 ppm) and make up 0.0075% - 0.012% of the total
fluid volume.
KC1 (Potassium Chloride) Substitute
Most of the unconventional shales contain varying concentrations of water sensitive clays.
These clays swell when contacted by fresh water and can potentially shut off flow paths. To
prevent this, Operators historically mixed powdered potassium chloride (KCI) in their fresh
water at the surface prior to pumping the water downhole. The KCI in the water prevented
absorption by the water sensitive clays. With the advent of high volume stimulations of
unconventional reservoirs, the process of manually mixing powdered KCI in surface tanks
became cumbersome and research was conducted to identify substitutes for powdered KCI.
Laboratory testing indicated quaternary amines could prevent adsorption of fresh water by
coating the clay particles present in the shale. The most popular chemistry for this application
was a product referred to as TMAC (tetramethylammonium chloride).
In 2009, Chesapeake Energy initiated a program focused on looking at the overall
environmental footprint of hydraulic fracturing additives to determine if improvements were
possible. One of the first targets of this effort was TMAC. Through extensive lab and field
testing, a new product, choline chloride, was identified as a very suitable replacement. Choline
chloride is commonly used as a nutritional supplement in animal feed, especially as a growth
aid for young chickens and it is also used as a nutritional supplement for humans. Chesapeake
now uses choline chloride in most areas where KCI substitutes are necessary in hydraulic
fracturing activities. Average use concentration for choline chloride is 500 to 2,000 ppm or
0.05% - 0.2% of the total fluid pumped.
Surfactants
Surfactants have been historically used in hydraulic fracturing to reduce interfacial tension
between fluid and the shale and between different phases of fluid. In the first instance, the
19
-------
desire is more robust initial water production from the well and in the second, the goal is the
elimination of emulsions in shales containing oil and water. The term "surfactant" covers a
multitude of products and those used in hydraulic fracturing can be as simple as a laurel sulfate
(similar to the ones used in household shampoos) to complex fluoro- and nano-surfactants.
Where used, surfactants are typically used at 500 to 1,000 ppm throughout the fluid which
would account for 0.05% - 0.1% of the total fluid volume.
Hydrochloric acid
Hydrochloric acid (HCI) is not used throughout the fracturing fluid, but rather usually is used to
lead the fluid for each stage. This is done to clear the production casing string of any debris and
to dissolve near-wellbore acid soluble minerals present in the shale. Dissolving these minerals
promotes additional flow paths for gas and/or oil to make its way to the wellbore and thus
improving the production from the well. Calcite (CaC03), a mineral commonly found in
unconventional shales, rapidly reacts with, and is dissolved by, hydrochloric acid. Once this
reaction is complete, the only remaining by-products are a soluble calcium salt, water and
carbon dioxide.
The hydrochloric acid used in hydraulic fracturing operations is usually 7.5% or 15% active. The
amount used is dependent on the number of perforations (openings to the shale reservoir) per
well and the mineral composition of the shale. For the major shale plays mentioned earlier,
typical usage range is 0.08% - 2.1% of the total fluid pumped (as volume of 15% HCI). However,
when looking at the active component, this would actually equal to 0.012% - 0.31% of the total
fluid pumped.
Acid Corrosion Inhibitor
While it is desirable for hydrochloric acid remove debris from the production casing, Operators
do not want the acid to degrade the integrity of the casing itself. To prevent this, an acid
corrosion inhibitor is utilized throughout the entire acid volume. Acid inhibitors tend to be
complex products as they are tasked with a protecting the steel casing without reducing the
acid's ability to dissolve iron oxides (mill scale) that is usually present on the surface of the pipe.
Acid inhibitors tend to contain amines, amides and/or amido-amines and often contain formic
acid as an intensifier for higher temperature applications. Usage rates can vary from 2,000 to
5,000 ppm in the acid only which equates to 0.0004% - 0.0043% of the total fluid volume.
Iron Control
As hydrochloric acid reacts with acid soluble minerals, such as calcite, in the formation and
"spends" there is a possibility for any iron oxide mill scale dissolved from the casing to
precipitate. This precipitation could potentially block flow channels in the reservoir so iron
control agents are used to prevent this from occurring. Iron control products are common
organic acids such as citric acid, acetic acid, thioglycolic acid and EDTA
(ethylenediaminetetraacetic acid) and incorporate the dissolved iron ion into their structures
and prevent it from precipitating. Typical usage in nearly all applications is 5,000 ppm of the
acid volume which equals 0.0004% - 0.011% of the total fluid volume.
20
-------
Recent Trends in Additive Selection
Chesapeake Energy, the American Petroleum Institute and the American Natural Gas Alliance
all feel strongly that the risk of contamination to groundwater from hydraulic fracture
stimulation of deep shale unconventional gas is extremely miniscule. However, we do realize
that there are employees who routinely work around hydraulic fracturing additives and while
safety is paramount in our industry, there is always the potential for an accidental surface spill.
It was with these two concerns in mind that we forged our Green Frac efforts.
As described above, Chesapeake Energy's Green Frac program was initiated in 2009 to
determine if it was possible to improve the overall environmental "footprint" of the additives
used in our hydraulic fracturing operations. Finding a substitute for the friction reducer, TMAC,
was an early, and successful, target of our efforts. Moving forward from that point, Chesapeake
has utilized our in-house laboratory, Geophysical, Engineering and Chemical assets to
thoroughly evaluate the necessity of every additive in every play. A primary goal was to
eliminate any additive that was not absolutely critical to successful completion and operation of
our wells. For those we deemed critical, we sought out materials that posed lower risk to
personnel and to the environment in the event of an accidental surface discharge. To date, we
have either eliminated, have found more desirable substitutes or are in the process of
successfully testing substitutes for the majority of additives historically used in hydraulic
fracturing of unconventional shales.
Conclusions
Contrary to what is often published in the media, the typical fluid used to hydraulically
fracture an unconventional shale well does not contain "596 toxic compounds". Most
frac stimulations contain fewer than eleven additives, including the corrosion inhibitor
and iron control agent that is always included in the hydrochloric acid.
Many of the additives we use are not 100% active. An example is hydrochloric acid
which is typically used in a field strength of either 7.5% or 15% active ingredient. If we
look at the activity of the various additives, you quickly see that the average hydraulic
fracturing fluid contains less that 1% by volume of chemical components. 99% of the
liquid volume is water.
Chesapeake Energy is very proud of our Industry leading program, Green Frac.
Through this program, we have eliminated 20% of the additives used in our fracturing
operations. In addition, we have identified and moved to more environmentally friendly
substitutes (or have products successfully finishing field tests) for over half of the
remaining additives.
21
-------
Unconventional Fracturing Fluids
D.V. Satya Gupta
Baker Hughes
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Abstract
Many tight gas formations are water-wet and under-saturated where the initial water
saturation in the reservoir is less than the capillary equilibrium irreducible water saturation. The
use of water-based conventional fracturing fluids causes water to be trapped in the near-
wellbore region, thereby significantly impairing the ability of gas to flow. Formations with sub-
irreducible water saturation can be stimulated with fluids that minimize the interfacial tension
(such as surfactant gels), minimize the amount of water used in the fluid (such as energized or
foamed fluids), dehydrate the formation (such as alcohol-based fluids) or completely eliminate
water (such as hydrocarbon-based or liquid carbon dioxide-based fluids). Since the rheology
and proppant-carrying properties of these fluids vary, the uses of these fluids are different and
will be discussed in detail in the paper. The paper will also present guidelines, based on
formation properties, to indicate the need for considering unconventional fluids. Some of the
new trends in the development of unconventional fluids are also presented.
Introduction
As the industry moves to extracting gas from tighter and tighter formations, particularly
formations such as shales or coalbeds where production is controlled by desorption of the gas
rather than matrix flow, fluids that are non-damaging to the proppant pack and formation are
becoming increasingly important. Wells with adverse capillary effects due to sub-irreducible
water or hydrocarbon saturation also require different fluids to minimize those effects or
mitigate effects caused by drilling with the wrong fluid. Several unconventional fluids have been
developed and successfully used for these unconventional formations in the last decade.
Adverse saturation in the formation can contribute to productivity impairment. Production has
been successfully achieved in formations with matrix permeability as low as 10~3 millidarcies
(mD). However, adverse capillary forces, which result in high in situ saturation of trapped water
or liquid hydrocarbons even in very low-permeability formations, make economic production
difficult. Low-permeability formations are typically tolerant of only minimal saturation damage
due to the sensitivity to capillary retention effects, and rock-to-fluid and fluid-to-fluid
compatibility issues. In these wells, the damage from drilling and completion can be overcome
by a properly designed frac treatment, which can penetrate beyond the zone of induced
invasion and damage.
22
-------
Fluid Retention
The major cause of productivity impairment in gas wells during drilling, completion and
fracturing or workover operations is fluid retention effects. These can include the permanent
retention of water or hydrocarbon based fluids or the trapping of hydrocarbon condensate
fluids retrograded in the formation during gas production. Bennion and his collaborators have
labeled these phenomena aqueous and hydrocarbon phase trapping (Bennion etal., 1994,
1996). Capillary forces in the formation are the reason for fluid retention. Capillary pressure
forces are the difference in pressure between the wetting (typically water in gas reservoirs) and
non-wetting (gas) phases in the matrix. The imbibition effect has been observed as a
particularly severe problem in reservoirs where sub-irreducible water saturation exists. Sub-
irreducible water saturation may have been created by a combination of factors, including
dehydration, desiccation, compaction, mixed wettability, significant height above the free
water level in oil reservoirs due to drainage, and diagenetic effects occurring during geologic
time. Laboratory capillary pressure measurements supply good approximations of the
irreducible water saturation that would normally be expected, but actual reservoir water
saturation can be substantially lower, i.e., a sub-irreducible level. The high capillary pressure
associated with low-permeability microporous reservoirs is illustrated in Figure 1. Measured
capillary pressure values for four rocks with permeability from 0.001 to 1.0 mD are presented
to illustrate the greater imbibition effects of water in lower-permeability formations. The
capillary pressure of the 0.001-mD core at 40% water saturation is 325 psi greater than that of
the 0.01-mD core at initial saturation. This illustrates the higher capillary pressure available in
tighter reservoirs to imbibe and trap aqueous liquids due to capillary imbibition. Injecting
water-based fracturing fluids into a high-capillarity reservoir results in the creation of a zone of
high water saturation in the near-wellbore or near-fracture face area. The relative permeability
curves in Figure 2 show how increasing water saturation above the irreducible water saturation
results in a dramatic decrease in gas relative permeability.
Gas production results in the affected zone reverting to the irreducible water saturation
dictated by the capillary effects of the system and not the sub-irreducible saturation that
existed before. The net effect is that the critical producing area of the well retains the increased
water saturation, a lowered relative permeability to gas and therefore lower productivity.
Several diagnostic techniques are available to estimate these effects (Gupta, 2009). These
correlations can be used to estimate compatibility of the formation to water-based fracturing
fluids. These are just guidelines, and exceptions abound, particularly for over-pressured
reservoirs where the capillary imbibition effects can be overcome in a relatively short time
frame (Bennion etal., 1996).
Fracturing Fluids
Conventional fracturing fluids include water-based and polymer-containing fluids, hydrocarbon-
based fluids, energized fluids and foams. These are not covered in this paper. Unconventional
fracturing fluids include non-polymer-containing fluids such as viscoelastic surfactant fluids,
methanol-containing fluids, liquid C02-based fluids and liquefied petroleum gas-based fluids.
The most cost-effective solution is to fracture the formation with the simplest of fluids. Low-
23
-------
viscosity water or hydrocarbon with the fewest additives would be the simplest fluids.
However, these have very low proppant transport properties, very little leak-off control and, if
pumped at high rates, will result in unacceptable friction. Various additives can control friction,
but if the formation has adverse saturation effects, even in tight gas formations with very little
leak-off, desired stimulation may not be achieved. Using salts in the fluids can control
compatibility with clay containing formations. Depending on pumping conditions, i.e., the shear
regime the fluid would experience, there may be need for shear-tolerant or shear-recoverable
fluids. For higher-temperature applications, these can be achieved by the use of organometallic
or borate crosslinked water-based fluids and crosslinked oil-based fluids. If the gas formations
are under-pressured, the fluids can be energized with N2 or C02 or foamed with N2 or C02 or a
combination of the two. The foam fluids also provide good leak-off control. If compatibility with
water may be an issue due to wetting issues, the use of viscoelastic surfactant fluids can be
considered. They also do not damage the proppant pack and can also be energized or foamed.
If incompatibility is due to capillary and unloading issues, methanol-containing fluid can be
considered. If the incompatibility is severe, then crosslinked methanol-based fluid, liquid C02-
based fluid or LPG may be the answer.
Viscoelastic Surfactant Fluids
Viscoelastic surfactant (VES) gel systems have been described in the patent literature for
friction reduction and as well treatment fluids (Teot, 1981). Its use in everyday life has been
around for some time. Its use in fracturing fluids is relatively a new phenomenon, but the
patent literature has exploded in this area in the last few years.
Principally, these fluids use surfactants in combination with inorganic salts or other surfactants
to create ordered structures, which result in increased viscosity and elasticity. These fluids have
very high zero-shear viscosity without undue increase in high-shear viscosity. Thus, they tend to
be shear-degradable fluids. As explained by Asadi et al. (2002), zero-shear viscosity has been
found to be an essential parameter in evaluating proppant transport. Therefore, these fluids
can transport proppant with lower loading and without the comparable viscosity requirements
of conventional fluids.
The technology of VES systems can be broken down into several categories based on the
structure the system creates: worm-like micelles, lamellar structures or vesicles.
As the concentration of surfactant increases in water, micelles start to form. Further increasing
the concentration exceeds the critical micelle concentration (CMC) for the surfactant in water;
these molecules start interacting with each other. These interactions are based on ionic forces
and can be amplified by adding electrolytes (salts) or other ionic surfactants. Depending on the
ionic charges and the size and shapes of the surfactants and these counter ions, ordered
structures start to form, which increases viscosity and elasticity. The reverse mechanism is true
for breaking these systems. The structures can be disrupted by adding other surfactants, ionic
additives and hydrocarbons (from the formation or mutual solvents or other solvents) or can be
diluted by additional formation water. The most common commercial systems use cationic
surfactants with inorganic salts (Teot et al., 1988) or with anionic surfactants (Zhang, 2002).
24
-------
Anionic surfactants with inorganic salts are also common (di Lullo etai, 2002). Zwitterionic and
amphoteric surfactants in combination with inorganic salts have been used (Dahanayake etal.,
2004).
The common VES fluids have a temperature limit in the range of 160 to 200 °F without foaming.
High-temperature stabilizers have been known to increase the temperature limit to 250 °F. Not
all of these fluids are compatible with C02. They have been shown to be economical
replacements for conventional borate fluids for tight gas applications (Rieb, 2007). At least with
one of these fluids, the flowback water from these treatments can be recycled (Gupta and
Tudor, 2005, Gupta and Hlidek, 2009). This particular fluid uses a cationic surfactant neutralized
with an anionic surfactant. The flowback water, in gas wells, tends to return some of the
cationic surfactant and most of the anionic surfactant. The flowback water is typically collected
for 24 hours into a tank. Initially, the fluid was filtered to remove any formation fines. Based on
experience, it was found that allowing the fines to settle was sufficient to remove the fines.
After settling, the middle 75% of the flowback water was transferred to a frac tank and the rest
of the required water for the fracturing treatment was made up with fresh water. Using
analytical or viscoelastic measurements, additional surfactants were used to reconstitute the
fluid. Russell (2001) reported the procedure and well production results from using the recycled
fluid in field study in Canada showing no effect of recycling on well production.
These VES fluids are operationally very simple as only one or two additives are added on the fly
without any need to hydrate polymers. They do not require any biocides because they do not
contain any biopolymers. They do not require additional flowback surfactants because they
have inherently low surface and interfacial tension. No additional clay control additives are
needed: They contain either salts or cationic surfactants, which have properties similar to KCI
substitutes. The surfactants have molecular weights of hundreds, as opposed to the guar
polymer with millions.
Viscosity is broken by altering the surfactant properties, by adding other hydrocarbons or by
altering the salinity or pH. The regain permeability with these types of systems approaches
100%. Because of the wetting tendencies of the surfactants in some of the VES systems, they
are useful even in formations with sub-irreducible water saturation and liquid-trapping issues,
despite being aqueous-based.
Viscoelastic Surfactant Foams
A natural extension of VES fluid technology is the VES foams (Zhang, et a/., 2002). These foams
can be formed with N2 or C02. As mentioned before, not all VES systems are compatible with
C02. No additional foamers are needed with these systems. The foam viscosity can be adjusted
by adjusting foam quality and the viscosity of the base VES system. They have been successfully
used in gas formations to 250 °F (Gupta and Leshchyshyn, 2005a). In formations with potential
to form water blocks, these fluids are particularly suited because the leak-off fluid still contains
the surfactants, which reduce surface tension in the matrix, overcoming capillary forces and
helping in recovery of the fluid. These fluids have been shown to be suited for fracturing
25
-------
coalbed methane wells that contain water because the foams control leak-off into the cleats
without damage from polymer residue.
With the advent of ultra-lightweight proppants (ULWP), an extension of this technology has
been very successful in under-pressured tight gas fields. A liquid suspension of the ULWP in a
viscoelastic gel can be added to a stream of nitrogen or C02 in the field to form a very high
quality (> 85 quality or volume percent) mist as a fracturing fluid, resulting in a partial
monolayer frac treatment. This technology has been utilized very successfully in dry, low-
pressure, tight formations in shales and coalbed methane wells in the US and Canada.
Emulsion of Carbon Dioxide with Aqueous Methanol Base Fluid
Certain formations have potential to retain even limited water used in foams and VES foams of
over 70 quality. These fluids may damage these sensitive formations because of sub-irreducible
water saturation and liquid trapping. In these formations, replacing 40% of the water phase
used in conventional C02 foams (emulsions) with methanol can minimize the amount of water.
Gupta et al. (2007) showed that a 40% methanol aqueous system yielded the highest viscosity
of aqueous methanol mixtures, has a freeze point close to -40 °C (the lowest operating limit for
fracturing equipment in the field) and surface tension around 30 dynes/cm. These emulsions
use surfactants, which are methanol-compatible foamers, in the place of conventional foamers.
Typical C02 quality approaches 85, which has resulted in high regained permeability and rapid
clean up and production results in several Canadian gas formations (Gupta et al., 2007).
Non-Aqueous Methanol Fluids
In formations with severe liquid (aqueous and hydrocarbon) trapping problems, non-aqueous
methanol fluid may be a solution. Over the years, several authors have identified the
advantages of alcohol-based fluids (McLeod and Coulter, 1966; Smith, 1973; Tiner et al., 1974;
Thompson et al., 1992; Hossaini et al., 1989; and Hernandez, et al., 1994). These advantages
include, but are not limited to, low freezing point, low surface tension, high water solubility,
high vapor pressure and formation compatibility. Methanol is also the fluid of choice for
formations with irreducible water and/or hydrocarbon saturation (Bennion et al., 1994, 1996b).
Three concerns with methanol all relate to safety: low flash point, high vapor density and flame
invisibility. With special precautions, as previous authors have identified (Thompson et al.,
1992; and Hernandez et al., 1994), methanol can be safely used in the field.
Several approaches to increasing the viscosity of methanol have been described in the
literature (Thompson etal., 1992; Hossiani etal., 1989; Boothe and Martin, 1977; Crema and
Aim, 1985; and Gupta et al., 1997). These range from foaming methanol to gelling with
synthetic polymers (e.g., polyacrylamide and polyethylene oxide) and modified guar. Attempts
were also made to crosslink gelled methanol with metal crosslinkers. However, Ely (1994)
described limitations that restrict the use of gelled non-aqueous methanol: solubility of these
polymers in both aqueous and non-aqueous methanol, ability to crosslink, ability to break the
polymer, and temperature limit.
26
-------
The most recent development (Gupta, etal., 1997; and Mzik, 1993 and 1994) describes a
modified guar dissolved in anhydrous methanol crosslinked with a borate complexer and
broken by an oxidizing breaker. This system has been successfully used in the field. In under-
pressured wells, it has been energized with N2. There has been an interest in a C02-energized
methanol fluid for similar formations with severely under-pressured wells. Hence a new
polymer that is soluble in non-aqueous methanol and compatible with carbon dioxide was
identified. These non-aqueous base gels can be crosslinked with borate at pseudo-high pH
(non-aqueous fluids do not have pH) or with zirconium crosslinker at pseudo-low pH for C02
compatibility. The special version of this HPG can hydrate in 100% methanol to give viscosity to
the base gel and is also compatible with C02 without precipitation (Gupta etal., 2003).
These fluids can be completely broken with special breakers, resulting in very high regained
permeability in the proppant pack and in very sensitive formations. These fluids should be
selectively used in gas formations with special safety considerations due to flammability of
methanol. These new-generation fluids also do not require any water for hydration or for
breaking.
Liquid COz-Based Fluids
Fluids based on liquid C02 are the real unconventional fluids. The concept and applications of
these fluids require outside-the-box thinking. These fluids have been very successfully used in
tight gas applications in Canada and several US formations. Their chemistry and physics have
been extensively published, as summarized in a paper by Gupta and Bobier (1998) and
described in brief here. The family of these fluids consists of pure liquid C02 and a binary fluid
consisting of a mixture of liquid C02 and N2 to reduce costs.
Conventional fracturing fluids rely on viscoelastic properties to inhibit leak-off from the fracture
into the reservoir. Filter cake deposition from long-chain polymers or high filtrate viscosity in oil
gel systems provides fluid loss control to establish adequate fracture width. Sufficient fracture
width is required to allow proppant placement in the fracture. Liquid C02 has very low viscosity
(Gupta and Bobier, 1998) and thus does not have the viscosity or filter cake properties to
establish fracture widths when pumped at typical rates (i.e., 20 BPM).
Low-viscosity, low-temperature fluids have higher leak-off rates than conventional fluids, but
this characteristic has an upper limit that is dependent on reservoir parameters. The use of a
gas or liquefied gas makes the fluid compressible. Pumped at high pressure and low
temperature, the fluid volumetrically expands when exposed to lower pressure and higher
temperature in the formation. The fluid is not in steady state, and thus positive transient effects
occur. This thermal expansion effect inhibits leak-off near the fracture face and promotes the
development of fracture width. A combination of thermal expansion, relative permeability
effects and possible turbulence through small reservoir pore throats curtails leak-off of liquid
C02 from the fracture into the reservoir.
Several papers have described the unique nature of liquid C02 and liquid C02/N2 mixtures as
fracturing fluids (Lillies, 1982; Tudor et a/., 1994; Mazza, 1997; and Gupta and Bobier, 1998). In
27
-------
these systems, the proppant is placed in the formation without causing damage of any kind,
and without adding any other carrier fluid, viscosifier or other chemicals. As was described
previously, "the use of a reservoir friendly substance like liquid C02" (and inert N2) "offers
unique advantages through the elimination of capillary fluid retention and clay swelling"
(Mazza, 1997).
These low-viscosity fluids are not an obvious choice of fracturing fluid. However, a large
number of jobs have been successfully performed with them (Gupta and Bobier, 1998). One of
the major limitations of this technology has been their high treatment cost. Although
stimulation treatments using the low-viscosity liquid C02 system have been successful, the high
rates required to place these jobs and the associated frictional losses raised horsepower
requirements.
Liquid COz-Based Foam Fluid
Several attempts have been made to increase the viscosity of C02-based fluids while trying to
maintain the conductivity and formation compatibility of these fluids with very little
operational success (Bullen et a/., 1987). The liquid C02-based foam fluid consists of a foam of
N2 gas in liquid C02 as the external phase stabilized by a special foamer soluble in liquid or
supercritical C02 (Gupta, 2003). The main advantage of this fluid is the additional viscosity
gained by the foam over liquid C02. The use of 75 volume percent of N2 also makes the fluid
very cost-effective and applicable to project frac applications where multiple jobs can be
performed in a single day. The fluid has also found niche application in coalbed fracturing in
Canada on dry coalbeds where any water introduced into the formation damages the cleats.
Gelled Liquefied Petroleum Gas
Recently, Liquefied Petroleum Gas (LPG) has been discussed in a patent application as a
hydraulic fracturing fluid (Loree and Mesher, 2007). The application proposes that LPG can be
viscosified and proppant added to the fluid much like conventional fracturing fluid. The
application further describes a unique and novel process that safely handles LPG and meters
proppant into the gelled LPG stream for fracturing treatments. LPG gases are a mixture of
petroleum and natural gases existing in a liquid state at ambient temperatures and moderate
pressure (less than 200 psi). Unlike conventional hydrocarbon-based fracturing fluids, the
common LPG gases, propane and butane, are tightly fractionated products with over 90%
purity. There are many advantages in using liquefied petroleum gases for hydraulic fracturing if
it can be done safely. The properties of density, viscosity and surface tension with complete
solubility in formation hydrocarbons are very beneficial. Recovery of the LPG very nearly
approaches 100%, clean up is very rapid (often within 24 hours), phase trapping is virtually
eliminated, and LPG properties allow for extended shut-in times without detriment.
Additionally, direct flowback to an available pipeline can be readily achieved. The result is a
potential cost-effective stimulation with effective fracture lengths, excellent post-treatment
production and the potential for zero flare clean-up.
28
-------
New Developments
New developments in the area of unconventional fluids have been in increasing the
temperature of use of viscoelastic fluids, the use of associative polymers that associate with
surfactants that can be used as straight fluid or foams (Gupta and Carman, 2011) and fluids
based on produced water that are also based on associative polymers.
Conclusions
Several unconventional fracturing fluids are described in this paper to minimize or eliminate
phase trapping issues associated with stimulation of tight gas wells. Proper selection of the fluid
depends on the severity of the issues and economics.
Acknowledgement
The author wishes to thank all his colleagues at Baker Hughes and the predecessor companies
who have been mainly instrumental in the development and field utilization of several of the
unconventional fracturing fluids described in this paper.
29
-------
References
Asadi, M., Conway, M.W., and Barree, R.D., 2002, Zero Shear Viscosity Determination of
Fracturing Fluids: An Essential Parameter In Proppant Transport Characterizations,
paper SPE 73755 presented at SPE International Symposium and Exhibition on
Formation Damage Control, Lafayette, Louisiana, 20-21 February.
Bennion, D.B., Cimolai, M.P., Bietz, F.R., Thomas, F.B., 1994, Reductions in the productivity of oil
and gas reservoirs due to aqueous phase trapping,/CPr, November.
Bennion, D.B., Thomas, F.B., and Bietz, R.F., 1996a, Low Permeability Gas Reservoirs: Problems,
Opportunities and Solutions for Drilling, Completion, Stimulation and Production, paper
SPE 35577, presented at SPE Gas Technology Conference, Calgary, AB, Canada, 28 April-
IMay.
Bennion, D.B., Thomas, F.B., Bietz, R.F., and Bennion, D.W., 1996b, Water and Hydrocarbon
Phase Trapping in Porous Media, Diagnosis, Prevention and Treatment, JCPT,
December.
Boothe, J.E., and Martin, F.D., 1977, Thickened Alcohol Well Treating Compositions, U.S. Patent
No. 4,012,327.
Bullen, R.S., Mzik, J., Richard, J.P, 1987, Novel compositions suitable for Treating Deep Wells,
U.S. Patent No. 4,701,270.
Crema, S.C., and Aim, R.R. 1985, Foaming of Anhydrous Methanol For Well Stimulation, paper
SPE 13565, presented at SPE Oil Field and Geothermal Chemistry Symposium, Pheonix,
Arizona, 9-11 March.
Dahanayake, M.S., Yang, J., Niu, J.H.Y., Derian, P.J., Li, R., and Dino, D., 2004, Viscoelastic
surfactant fluids and related methods of use, U.S. Patent No. 6,831,108.
di Lullo Arias, G.F., Rae, P., and Ahmad, A.J.K.,2002, Viscous fluid applicable for treating
subterranean formations, U.S. Patent No. 6,491,099.
Ely, J.W., 1994, Stimulation Engineering Handbook, PennWell Books, Tulsa, OK.
Gupta, D.V.S., 2003, Field Application of Unconventional Foam Technology: Extension of Liquid
COzTechnology, paper SPE 84119, presented at SPE Annual Technical Conference and
Exhibition, Denver, Colorado, 5-8 October.
Gupta, D.V.S., 2009, Unconventional Fracturing Fluids for Tight Gas Reservoirs, paper SPE
119424, presented at SPE Hydraulic Fracturing Technology Conference, The Woodlands,
Texas, 19-21 January.
Gupta, D.V.S., and Bobier, D.M., 1998, The History and Success of Liquid C02 and C02/N2
Fracturing System, paper SPE 40016, presented at SPE Gas Technology Conference,
Calgary, AB, Canada, 15-18 March.
Gupta, D.V.S., and Carman, P.S., 2011, Associative Polymer System Extends the Temperature
Range of Surfactant Gel Frac Fluids, paper SPE 141260, presented at SPE International
Symposium on Oilfield Chemistry, The Woodlands, Texas, 11-13 April, 2011.
Gupta, D.V.S., and Hlidek, B.T., 2009, Frac Fluid Recycling and Water Conservation: A Case
Study, paper SPE 119478, presented at SPE Hydraulic Fracturing Technology Conference,
The Woodlands, Texas, 19-21 January.
30
-------
Gupta, D.V.S., and Leshchyshyn, T.T., 2005a, Surfactant Gel Foam/Emulsion: History and Field
Application in Western Canadian Sedimentary Basin, paper SPE 97211, presented at SPE
Annual Technical Conference and Exhibition, Dallas, Texas, 9-12 October.
Gupta, D.V.S., and Leshchyshyn, T.T., 2005b, C02 Energized Hydrocarbon Fracturing Fluid:
History & Field Application in Tight Gas Wells in the Rock Creek Gas Formation, paper
SPE 95061, presented at SPE Latin American and Carribean Petroleum Engineering
Conference, Rio de Janiero, Brazil, 20-23 June.
Gupta, D.V.S., Niechwiadowicz, G., and Jerat, A.C, 2003, C02 Compatible Non-Aqueous
Methanol Fracturing Fluid, paper SPE 84579, presented at SPE Annual Technical
Conference and Exhibition, Denver, Colorado, 5-8 October.
Gupta, D.V.S., Pierce, R.G., and Litt, N.D., 1997, Nonaqueous Gelled Alcohol Fracturing Fluid,
paper SPE 37229, presented at SPE International Symposium on Oil Field Chemistry,
Houston, Texas, 18-21 February.
Gupta, D.V.S., and Tudor, E.H., 2005, Method for fracturing subterranean formations, U.S.
Patent No. 6,875,728.
Gupta, D.V.S., Hlidek, B.T., Hill, E.S.W., and Dinsa, H.S., 2007, Fracturing Fluid for Low-
Permeability Gas Reservoirs: Emulsion of Carbon Dioxide With Aqueous Methanol Base
Fluid: Chemistry and Applications, paper SPE 106304, presented at SPE Hydraulic
Fracturing Technology Conference, College Station, Texas, 29-31 January.
Harris, P.C., 1988, Fracturing-Fluid Additives,/Pr40:10, 1277, October.
Harris, P.C., and Heath, S.J., 1996, Rheology of Borate Crosslinked Foams, SPEPF, 113, May.
Hernandez, J.M., Fernandez, C.T., and Sciana, N.M., 1994, Methanol as Fracture Fluid in Gas
Wells, paper SPE 27007, presented at SPE Latin America Carribean Petroleum
Engineering Conference, Buenos Aires, Argentina, 27-29 April.
Holditch, S.A., 1979, Factors Affecting Water Blocking and Gas Flow from Hydraulically
Fractured Gas Wells, JPT 31:12, 1515, December.
Hossaini, M., Jabes, W., Grisdale, J., and Hossaini, M., 1989, Fracturing with Crosslinked Gelled
Methanol: A New Approach to Well Stimulation, JCPT 28:5, 49, September-October.
Keelan, D.K., 1975, Rock Properties and their effect on Gas Flow and Recovery, presented at the
22nd Annual Southwestern Petroleum Short Course, Texas Tech Univeristy, Lubbock,
Texas, April 17-18.
Lillies, A.T., 1982, Sand Fracturing with Liquid Carbon Dioxide, paper CIM 82-33-23, presented
at the Annual Technical Meeting of the Petroleum Society of CIM, Calgary, AB, Canada,
6-9 June.
Loree, D.N. and Mesher, ST., 2007, Liquified Petroleum Gas Fracturing System, US Patent
Application 2007204991.
Mazza, R.L., 1997, Liquid C02 improves Fracturing, Hart's Oil and Gas World, 22, February 1997.
McLeod, H.O., and Coulter, A.W, 1966, The Use of Alcohol in Gas Well Stimulation, paper SPE
1663, presented at SPE Eastern Regional Meeting, Columbus, Ohio, 10-11 November.
Mzik, J., 1993, High Viscosity Crosslinked Gelled Alcohol, Canadian Patent Application No.
2,108,194.
Mzik, J., 1994, High Viscosity Alcohol Gel For Shallow Wells, Canadian Patent Application No.
2,119,682.
31
-------
Phillips, A.M., and Mack, D.J., 1991, Fracturing process using a viscosity stabilized energizing
phase, U.S. Patent No. 5,002,125.
Rieb, B., 2007, Private Communication.
Russell, D., 2001, Recycling of Water & Minimizing Fresh Water Usage, presented at Shallow
Gas Forum, PTAC, Calgary, Canada, June.
Settari, A., Bachman, R.C., Morrison, D.C., 1986, Numerical Simulation of Liquid CChHydraulic
Fracturing, paper CIM 86-37-67, presented at the Annual Technical Meeting of the
Petroleum Society of CIM, Calgary, AB, Canada, 6-11 June.
Smith, C.F., 1973, Gas Well Fracturing using Gelled Non-Aqueous Fluids, paper SPE 4678,
presented at the Fall Meeting of the Society of Petroleum Engineers of AIME, Las Vegas,
Nevada, 30 September-3 October.
Teot, A.S., Rose, G.D., and Stevens, G.A., 1981, Friction reduction using a viscoelastic surfactant,
U.S. Patent No. 4,615,825.
Teot, A.S., Ramaiah, M., and Coffey, M.D., 1988, Aqueous Wellbore Service Fluids, U.S. Patent
No. 4,725,372.
Thompson, J.E. Sr., McBain, C., Gregory, G., and Gerbrandt, D., 1992, New Continuous-Mix
Process for Gelling Anhydrous Methanol Minimizes Technology, JPT, 832, July.
Tiner, R.L., Stahl, E.J., and Malone, W.T., 1974, Developments in Fluids to Reduce Potential
Damage from Fracturing Treatments, paper SPE 4790, presented at SPE Symposium on
Formation Damage Control, New Orleans, Louisiana, 30 January-2 February.
Tudor, R., Vozniak, C, Banks, M.L., and Peters W., 1994, Technical Advances of Liquid C02
fracturing, paper CIM 94-36S, presented at the Annual Technical Meeting of the
Petroleum Society of CIM, Calgary, Alberta, Canada, 12-15 June.
Veatch, R.W. Jr., 1983, Overview of Current Hydraulic Fracturing Design and Treatment
Technology,/Pr35, 853, April.
Zhang, K., 2002, Fluids for fracturing subterranean formations, U.S. Patent No. 6,468,945.
Zhang, K., Pierce, R., Litt, N.D., and Gupta, D.V.S., 2002, Foam-fluid for fracturing subterranean
formations, U.S. Patent No. 6,410,489.
32
-------
Figure 1. Effect of capillarity on water saturation (after Holditch, 1979)
600
500
400
£ 300
Q.
re
O
200
100
0.2 0.4 0.6 0.8
Water Saturation, fraction
Figure 2. Effect of water imbibtion on relative permeability changes (after Keelan, 1975)
Relative permeability
decreases rapidly as Water
Saturation increases.
0 20 40 60 80 100
Water Saturation, Percent Pore Space
33
-------
Produced Formation Water Sample Results from Shale Plays
Nancy Pees Coleman, Ph.D.
Environmental Consultants and Chesapeake Energy Corporation
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Chesapeake Energy Corporation (Chesapeake) and Environmental Consultants reviewed non-
potable produced formation water sample results available from the literature and other
sources. The data review was specifically directed at the chemical characteristics of source
water used for hydraulic fracturing fluids, hydraulic fracturing fluids, produced formation water
from natural gas wells following hydraulic fracturing activities (i.e., less than 30 or 90 days
following hydraulic fracturing), and produced formation water from operating natural gas
wells.. The largest amount of data represents sampling conducted of produced formation water
from the Marcellus Shale, including the GTI study (Hayes, 2009) and the USGS Produced Water
database (USGS, 2010). Additional data was available from private sources. This data set
included limited data for three other shale formations, the Barnett, Fayetteville and
Haynesville. Some of the data evaluated is subject to attorney-client privilege (herein after
"privileged data").
The zones in the four shale plays from which natural gas is being produced vary in average
depth below the surface from 4,500 feet in the Fayetteville to 11,500 feet in the Haynesville.
The average depth from the surface to the natural gas producing zones in each of the four shale
plays are thousands of feet below the geologic zones which produce potable groundwater.
Furthermore, temperatures and pressures found at these depths vary from 130°F and 3,000
pounds per square inch (psi), respectively, in the Fayetteville to 320°F and 10,000 psi,
respectively, in the Haynesville. These temperatures and pressures have marked effects on the
sampling and field analysis of produced formation waters and create potential safety issues for
sampling personnel.
The primary purpose of the produced formation water data review was to support
Chesapeake's baseline water sampling program for potable water. Of major interest was the
identification of potential sentinel chemicals and/or radiochemicals that could be included in
the baseline potable water sampling program. Further, the review was intended to assist in
decision making regarding disposal or treatment for reuse or recycling of produced formation
waters.
Data available from the GTI study included general chemistry parameters and metals as well as
volatile organic compounds and semi-volatile organic compounds from 70 wells in the
Marcellus. The data evaluation included a review of chemicals currently being used in
Chesapeake wells during hydraulic fracturing. Initially, Chesapeake had chosen the parameters
34
-------
for analyses based on their potential mobility in groundwater water systems, toxicity, and the
availability of analytical methods. Eventually, the complete list found at 40 CFR Part 264,
Appendix IX, and commonly associated with groundwater monitoring, was selected. The
complete list was specified because many of the chemicals in the treatment fluids are
proprietary or not disclosed in information available to Chesapeake. All of the analyses were
conducted using EPA analytical methods and analyses were performed by certified laboratories.
The majority of analyses were conducted by one laboratory.
Some additional data was available in the privileged data for special volatile organic
compounds, primarily glycols, and radiochemistry data. The glycols had been included because
of the frequency of use in hydraulic fracturing fluids and in order to evaluate their usefulness as
sentinel species.
Both the authors of the GTI study and the sources for the privileged data indicated that there
are significant issues regarding sampling of produced formation water. Natural gas is contained
under high pressure in specialized equipment that is not designed for producing high quality
environmental samples. The sample matrix itself presents challenges such as foaming and
changes in surface tension.
Analytical techniques are also impacted by the presence of elevated concentrations of total
dissolved solids and chloride. The inorganic and wet chemistry methods were most affected by
the presence of high total dissolved solids. Radiochemistry methods are particularly affected by
the elevated concentrations of barium and total dissolved solids.
As anticipated total dissolved solids and the divalent cations, including barium and strontium,
are elevated in produced formation water samples. The concentrations noted in the data
evaluated are consistent with that gleaned from the literature (USGS, 2010). Concentrations of
these parameters appear to increase with time following hydraulic fracture activities and
remain at high concentrations in the produced formation water from operating wells in each
shale formation. The concentrations of barium and strontium appear to correlate to the
concentrations of total dissolved solids. Chloride represented the most abundant anion. These
data are consistent with the generalized data available in the USGS database.
The literature supports the interaction and release of arsenic from host rock in the presence of
refined hydrocarbons products, therefore, a specific review of the arsenic data was conducted.
In general, arsenic concentrations were not found in many of the samples above method
detection limits. The GTI study found some detectable concentrations of arsenic with produced
formation water sampled on day five following hydraulic fracturing having the highest
concentration, 124 u.g/L. Data available from analysis of produced formation water from the
Marcellus prior to disposal confirmed the general absence of arsenic. There were four detected
concentrations from 87 fluids samples. The highest detected concentration was 4.2 u.g/L.
Arsenic was not detectable in the limited data available for the Barnett, Fayetteville and
Haynesville formations; detection limits ranged from 1 to 10 u.g/L.
35
-------
In consultation with health physicists from SAIC and IEM, Chesapeake has determined that
radium 226 and 228 represent the most useful parameters to provide meaningful information
regarding radionuclides. This decision is based on the fact that radium 226 and 228 are more
soluble in water, are more potentially mobile, and represent 80 percent of the potential
ingested dose of radiation in drinking water. Based on experience with radionuclide analyses in
produced formation water, it appears that gamma spectroscopy (EPA Method 901.1) and
isotopic analyses of the select decay series provide the most accurate and usable information. It
is important for the activity results to be presented accompanied by the standard deviation and
minimum detected concentrations in order to put the data in context, particularly for non-
detected activity levels.
Certain volatile organic compounds, namely glycols, were also evaluated. Glycols were selected
because of their frequency of use in treatment fluids used in natural gas production. EPA
Method 8015 has been found to be insufficiently robust to overcome the matrix issues which
are attendant to the analysis of produced formation water. The range of detection limits seen
in the available data sets ranged from 10,000 to 50,000 u.g/L, which does not provide
meaningful results. EPA Method 8270 has some utility for larger glycols, such as glycol ethers;
however for the smaller, more soluble, ethylene and diethylene glycols, the extraction methods
are not useful. Sample results for produced formation water prior to disposal had similar issues
with elevated detection limits even when total dissolved solids levels were lower. In these
samples, ethylene glycol was found in 14 of 87 samples. Glycols may prove to be problematic
for use as sentinel compounds because of their other uses in natural gas production, such as
winterization of equipment similar to their use in automobiles, recreational vehicles, etc.
Measured concentrations of volatile organic compounds were available for the Marcellus from
the GTI study and more limited data had been provided for the Barnett, Fayetteville and
Haynesville formations. The most consistently detected volatile organic compounds were
benzene, toluene, ethyl benzene and xylenes (BTEX). This finding is consistent with data
available in the literature and would be anticipated as these compounds are closely associated
with hydrocarbon producing geologies from which the produced formation water comes. There
were no trends evident for the BTEX data. The appearance of BTEX was somewhat inconsistent
and when present was internally inconsistent. That is, the highest concentrations of benzene
were not in the same samples as the highest concentrations of the other three. In the iterative
sampling, there was no consistent timing for occurrence of the highest concentration. This may
well be due to the differences in water solubility in the presence of elevated total dissolved
solids for the individual BTEX components. The data is suggestive that there are differences in
BTEX levels between and within shale formations. There were few other volatile organic
compounds found in produced formation water; however, they were found on an inconsistent
basis.
The semi-volatile organic compounds (SVOCs) were generally not present in detectable
concentrations. The most frequently detected SVOC was pyridine in both the GTI Study and in
the other data available. In the GTI study data set, six of 70 samples had detectable levels of
pyridine in the hydraulic fracture fluids. The author speculated that presence of pyridine was
36
-------
due to its use as a precursor in the manufacture of one of the hydraulic fracturing additives.
Certain service companies have alkyl pyridines available as corrosion inhibitors (Weatherford,
2011). In the additional data evaluated, only two of eleven sampled hydraulic fracturing fluids
had detectable concentrations of pyridine. In one of these samples, the concentration was
related to the reuse of produced formation water as makeup water for the hydraulic fracturing
fluid. Pyridine has been reported as naturally occurring in oil shales (ATSDR, 1992). The
additional data sets may indicate support mobilization of the naturally-occurring pyridine since
it is soluble in water or may represent a degradation product. There appear to be differences
between the shale plays; however, the data set is small and therefore, does not support the
development of generalizations.
The iterative (time series) sample results presented by GTI and in other available data does not
reveal a consistent trend of increasing or decreasing concentration of volatile organic
compounds or SVOCs with increasing times after hydraulic fracturing.
The major conclusions from the review of produced formation water data are:
Produced formation water appears to be highly variable within and between shale plays.
Analytical techniques used for chemical and radiochemical characterization of the
produced formation water must be robust to the matrix interferences presented by
total dissolved solids.
Few volatile organic compounds and SVOCs are consistently found in produced
formation waters. Benzene, ethyl benzene, toluene and xylenes are expected to be
present in varying concentrations in natural gas. The concentrations of volatile organic
compounds and SVOCs are generally low and tend to preclude their use as sentinel
chemicals.
The most reliable sentinel compounds appear to be total dissolved solids, chloride and
divalent cations.
References
ATSDR, 1992. Toxicological Profile for Pyridine, Atlanta, GA, September.
Hayes, T, 2009. Sampling and Analysis of Water Streams Associated with the Development of
the Marcellus Shale Gas, Final Report, prepared for Marcellus Shale Coalition, Gas
Technology Institute, Des Plaines, IL, December 31.
USGS, 2010. USGS Produced Water Database, queried at
energy.cr.usgs.gov/prov/prodwat/intro.html.
Weatherford International Ltd., 2011. Corrosion Inhibitors, available on-line.
37
-------
Trace Metal Chemistry and Mobility in the Marcellus Shale
Tracy L. Bank
University at Buffalo
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Drilling and "fracing" of the Marcellus shale causes fluid-rock interactions that have the
potential to mobilize metals that are naturally enriched in the shale. While the concentrations
of these metals are low, their mobilization from the solid, through redox reactions and leaching,
is cause for further study.
In a series of studies, the trace metal geochemistry of the Marcellus Shale and the extractability
of select metals through fluid-rock interactions have been determined. The studies were
completed on 16 shale samples including outcrop, well core, and well cuttings from western
New York and Pennsylvania. Instrumental neutron activation analysis (INAA) was used to
determine the concentrations of metals in the shale samples prior to any chemical treatment.
The analysis indicates that the Marcellus Shale is enriched in barium, uranium, chromium, and
zinc, among other metals (Table 2). For comparison, the maximum contaminant level (MCL) in
the United States for barium, uranium, chromium, and zinc are 2 ppm, 30 ppb, 50 ppb, and 5
ppm, respectively. It is likely that the barium concentrations reported here are elevated due to
contamination from drilling mud which commonly contains barium. This would explain the very
high concentration of barium in cuttings and the increased concentration in core compared to
outcrop samples (which would not be contaminated).
Table 2. Selected Metal Concentrations in Marcellus Shale
Metal
Barium
Uranium
Chromium
Zinc
Outcrop Average
(range) n = 3
670 ppm
(500-900 ppm)
30 ppm
(11-53 ppm)
70 ppm
(53 -100 ppm)
515 ppm
(*BDL-660)
Core Average
(range) n= 9
1000 ppm
(*BDL-2000ppm)
28 ppm
(10-50 ppm)
86 ppm
(70-100 ppm)
526.3 ppm
(BDL-2290 ppm)
Cutting Average
(range) n=4
1925 ppm
(900-3600 ppm)
17 ppm
(9-34 ppm)
98 ppm
(70-120 ppm)
403 ppm
(BDL-530ppm)
Sample Average
(range) n=16
1314 ppm
(BDL-3600 ppm)
26 ppm
(9-53 ppm)
86 ppm
(53-120 ppm)
496 ppm
(BDL-2290)
*BDL = Below Detection Limit (Ba = 100 ppm, U = 0.5 ppm, Cr = 10 ppm, Zn = 50 ppm)
The concentration of metals in the shale generally increases as the concentration of total
organic carbon increases. This is likely because conditions that favor the preservation of organic
matter during shale formation also favor the deposition of metals in a reduced state. This is of
interest because natural gas developers naturally target regions with the higher concentration
of organic matter which are also the most metal-rich.
38
-------
To determine the extent that these metals could be mobilized during reactions that occur
between drilling and/or fracing fluids and the shale, several chemical extractions have been
performed. Batch extraction studies were completed using a measured mass of ground and
sieved shale and a known volume of chemical extractant. The batch systems were either shaken
or stirred for the duration of the extraction (which varied from just a few minutes to 24 hours).
Metal concentrations in the shale were measured prior to the extraction test and following the
extraction. Also, in some cases metal concentrations in the batch fluid were measured at the
end of the extraction study. The extraction studies included: 1) a dilute sodium bicarbonate
solution which could remove easily exchangeable ions from the rock, 2) a hydrochloric acid
extraction to dissolve carbonate minerals and strongly sorbed metals, 3) a hydrogen peroxide
treatment to oxidize organic matter and associated metal, and 4) a sodium dithionite solution
which would remove oxide and hydroxide minerals and their associated metals. The results of
the sodium bicarbonate study indicated that little to no metal could be solubilized from the
shale during a 24 hour reaction. The results of the sodium dithionite study were also negligible
because there are so few oxide minerals in the shale. The hydrogen peroxide treatment
indicated that some metals were associated with the organic matter in the shale. An average of
20% of the zinc in the samples was solubilized with an extractable range from 3-80%.
Additionally, up to 35% of the uranium was solubilized, but a greater amount of uranium could
be removed from the outcrop samples compared to the core. Up to 30% of the chromium was
removed, but an average of only 8.5% was removed from the shale.
The results of the hydrochloric acid extraction provided the largest amount of data and
indicated potential for metals to be solubilized and mobilized from the shale. HCI is used prior
to fracing a well to dissolve carbonate cement and increase flow pathways for the natural gas.
In this study, finely ground shale was reacted with 3 M and then 1.5 M HCI for a total of two
hours. The amount of metal solubilized in the batch solution was measured using inductively
coupled plasma mass spectrometry (ICPMS). Additionally, the solid sample was re-analyzed
following the extraction test. Interestingly, the results of this study varied depending on
whether the sample was outcrop or core; outcrop samples contain up to 70% HCI extractable
uranium while core samples contain only about 25% HCI extractable uranium. Zinc and
chromium were both moderately extractable in both sample types, up to 60% and 20%
respectively. Up to 40% of the barium was removed during the extraction but results vary from
sample to sample and are probably affected by contamination from drilling muds.
This study shows that metals may be mobilized from the shale into wastewater produced
during natural gas development. Additionally, the results of this study indicate that many of the
heavy metals in the shale are potentially leachable and may present a disposal issue. Drill
cuttings that are removed during drilling of thousands of feet of well hole need to be properly
disposed. These cuttings contain reduced metals that will oxidize over time if exposed to air and
water. Proper disposal of these cuttings needs to be considered.
39
-------
Fracture Fluid Additive and Formation Degradations
George E. King
Apache Corporation
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
The focus of this presentation is on reactions that describe the degradation of fracturing fluids
and formations during the hydraulic fracturing process and the clean-up period of 2 to 6 weeks
following the fracturing application. A description of the primary chemical reaction controls,
namely permeability and area-to-volume ratio, precedes a discussion of the better known and
more common degradation reactions. Although shale formations will be highlighted because of
media attention, other formations will be covered as well.
Reactions in wells are subject to most normal catalytic and restriction influences, but also have
a set of specific limiters that are found in few other places in chemical industry. Reaction
influences of temperature and pressure are reasonably predictable, but other reaction controls
such as reaction rate are strongly influenced by the area and mixing constraints described by
location of the reaction, the area-to-volume ratio and the behavior and stability of the by-
products. The reaction zones include:
Open wellbore - a zone of high energy but small area in which reactions are possible but
segregation of sequenced reactants can be enforced by turbulence during their one-
time passage through the wellbore where area-to-volume ratio is sub 1:1.
Hydraulic fracture - mostly singular in a vertical well and usually narrow (~3mm to
25mm) slit where chemical reactions that depend on diffusion are limited by the pump
rate and reactants may be quickly lost to the natural fractures or matrix through leak-off
generated by pressure differential and controlled by the permeability to those side
connections. Area-to-volume ratio is 8 to 50 for normal fracture widths.
Natural fractures - very narrow (~0.01mm to 1mm) but numerous slits that may or may
not be continuous in even a small area. Area-to-Volume ratios range from 100:1 to over
1000:1.
Matrix or interstitial porosity - a potentially highly reactive zone because of the ratio of
the very large surface area-to-volume of fluid ratio (order of 20,000:1 to 30,000:1).
Reactive target potential in the previously described zones include rock and minerals, metal,
cement and a typically short-list of chemical additives used in the wells to facilitate control of
natural mineral reactions, transport of solids (e.g., cuttings and proppant), and permit physical
reactions (i.e., fracturing) that can create, widen and/or stabilize a flow path of improved
permeability to allow the formation fluids to flow to the wellbore.
The primary down-hole limit to any chemical or physical reaction is access. Permeability (a
measure of ability to flow a fluid through a rock) is the fundamental restriction to fluid flow.
40
-------
Once the potential reactant has reached the zone of reaction, the area-to-volume ratio is a
primary influence on most reactions. Diffusion, the ability to get reactants to the site and move
products away, is controlled by area-to-volume ratio, the permeability of routes to and from
the reaction and the type and behavior of by-products themselves.
Degradation reactions involving well construction and operation issues include pipe stability
and cement stability, which are mostly chemical concerns. Although physical reactions such as
pipe collapse, burst, tension failures and erosion are known to occur, their behavior has been
well described and adequate controls appear to be in place based on the population of 50+ year
old wells still producing and passing mechanical integrity tests.
Cementing stability and degradation have been addressed in the petroleum engineering
literature by more than eleven thousand general papers and over six hundred that comment
directly on cement degradation reactions and blocking mechanisms. This discussion has
covered mineral and organic acids, C02, H2S, thermal, low pH waters, sulfate effects, cyclic
pressure behavior, long term performance and other issues. Reactions that deteriorate cement
are easily demonstrated in the laboratory in beakers but are reduced exponentially when issues
of access are considered. All reactions depend on access and the low permeability of the
formations surrounding the numerous formation barrier seal areas, coupled with the extremely
low permeability of the cement itself make significant deterioration reactions unlikely if not
practically impossible. Added to this are instability and quick precipitation of common reaction
by-products that form an impenetrable barrier on most reaction surfaces. Additives that reduce
cement degradation include those for sulfate resistant cement, special thermal compositions
for very high temperature wells, a variety of additives that reduce cement permeability,
improve strength and chemical and physical treatment to improve bonding. Problems with
cement almost universally are traced to poor application during well construction.
Pipe reactions are mostly corrosion and erosion-corrosion. These reactions are also well known
with over two thousand papers in petroleum engineering literature of direct application of
wells and pipelines. These reactions are most common in long term production with acid gases
(C02 and H2S) and the short term contact of mineral and organic acids during acid stimulation.
Because of the short duration of the frac and the very small amount of weak mineral acid used
as a breakdown stage, the effect of corrosion or erosion-corrosion during a fracturing job is
negligible.
Formation degradation during fracturing is possible, but sharply limited by access. Higher
permeability conventional formations, sandstones and carbonates with permeability over about
1 mD, are largely inert to reactions with waters used in fracturing except for the limited
swelling reactions of smectite and a few mixed layer clays. Reactions in shale are still being
researched, but access is still the dominant control. Examples of reactions in the accessible
zone of fracture between proppant, shale formation and waters are being researched and
results of two recent papers are presented.
41
-------
Chemical additive reactions resulting in degradation or disappearance of the chemicals are
presented. These reactions include gel breaking, adsorption, absorption, capillary trapping and
precipitation. Examples of each of these mechanisms will be discussed. Nearly fourteen
thousand papers in the petroleum engineering literature deal with subjects of sorption, gel
breaking and capillary effects.
42
-------
Evaluating Interactions of Fracturing Fluids and Degradation
Products with Radionuclides Contained in Organic-rich Solid-
Phase Host Materials
Zoltan Szabo
U.S. Geological Survey
W.Trenton, NJ 08628
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
In situ contaminants (arsenic, radionuclides, or others) can be released from organic-rich gas-
bearing shale host rock when fracturing fluids are injected. This potential release raises
questions regarding the nature of wastes created and pumped to surface, and those remaining
in the formation. The rate of entry of various radionuclides into return flow liquid matrixes can
be studied in the laboratory and matched with corresponding observations in the field to
estimate the reactions and their rates, and provide understanding of the solid material
degradation processes.
Radionuclides can serve as tracers for rock-matrix degradation processes in both field- and
laboratory-based experiments. Uranium (U) and thorium (Th) have multiple isotopes and
multiple direct and indirect progeny that also have many isotopes. Concentrations of many of
these isotopes, their ratios, or ratios of progeny can be determined and monitored in water or
brine to estimate changes in U and Th concentrations in rock. Laboratory characterizations can
be utilized to determine sources of U in the solids matrix and to show changes in the
distribution of U or progeny radionuclides before and after manipulation (for example,
simulated fracking of core materials tested by a variety of laboratory approaches). The
combination of simultaneous or near simultaneous solid phase and liquid phase observation of
the fate of these materials in the presence of laboratory manipulations with the capacity to
make similar observations in the field make for a powerful tool for the study of the rock-matrix
degradation processes. The characterization of U occurrence in black shales and its fate has
already received attention and will provide useful analogues for design of such experiments.
A variety of tools are available for determination of radionuclide occurrences in samples of host
rock. Solid phase analysis of the relative abundance of a suite of radionuclides (238U, 226Ra, 228Ra
and 210Pb) may include low energy gamma-ray spectrometry that can provide simultaneous,
non-destructive determinations in representative samples and important baseline
concentration information. The drawbacks to using low energy gamma-ray spectrometry are a
large sample size is required (150-250 g) and the actual distribution within the sample cannot
be determined. Developed fission-track radiographs or alpha-autoradiographs can be observed
under a microscope to reveal the sources of U or other alpha-emitting radionuclides in rock
43
-------
samples. Fission-track radiographs are generated by irradiation of a polished thin section with a
thermalized neutron flux that causes fission of 235U in the sample. Fission fragments recoiling
from the sample surface are detected by covering the sample with a detector material (low U
muscovite). Fission-fragments paths are farther developed by etching the dector material with
hydrofluoric acid. The developed fission tracks are observed under a microscope. Although Th is
a possible source of fission fragments, the sensitivity of Th to thermal neutron fission is much
less than that for an equal amount of U. Autoradiographs are images of the sources of emitted
alpha particles, but do not have as high sensitivity and resolution as fission track images. An
alpha-sensitive photographic emulsion is spread directly over the thin section and the image of
the alpha tracks is observed directly under the petrographic microscope. The alpha
autoradiograph captures images of radionuclides undergoing active alpha decay as opposed to
images induced by the 235U fission noted in fission tracks.
The use of element mapping utilizing a variety of X-ray analysis techniques can be employed for
determination of radionuclide (and other trace element) occurrences in small-scale detail in
samples of host rock. Scanning electron microscopy energy dispersive x-ray (SEM/EDX) allows
for the capture of high resolution high magnification images as well as dispersive X-ray mapping
of the occurrence of the element at points within the image where the X-ray beam is focused.
Synchrotron-based spatially resolved micro-X-ray fluorescence (u.-SXRF) allows for small-scale
element concentration mapping on individual core samples, whereas synchrotron-based X-ray
absorption near edge spectroscopy (XANES or U.XANES) can be used to deduce elemental
valence state and Extended X-ray Absorption Fine Structure (EXAFS or U.EXAFS) can allow for
analysis of atom coordination (speciation). The most recent efforts have been to obtain images
with U.-SXRF mapping on the same spatial point on the rock sample with the X-ray beam while
the leaching experiment is in progress, with chemical effluents collected and analyzed from a
split sample undergoing the same procedure. These techniques can be done over small areas
within a highly polished sample (approximately 0.3 mm2 area with a step size of 10 u.m and a
dwell time of 2s per pixel). The dispersed nature of the U in organic-rich fine-grained rocks may
pose difficulty for signal resolution. The intense fluorescence peak of iron in samples can cause
interferences. Beamline time would need to be obtained at the National Synchrotron Light
Source (NSLS) at Brookhaven National Laboratory or the Stanford Synchrotron Radiation
Lightsource at Stanford University. Images of changes in radionuclide distribution and
speciation in the core during such experiments could be captured and compared to changes in
radionuclide concentrations in leachate from laboratory-based experiments. Radionuclide
concentrations in samples collected at varying stages of fracture fluid flowback in the field
could be compared to the results of radionuclide distribution determined in the laboratory for
leachate and rock. This series of experiments and field-based observations promises to shed
light on the mechanisms responsible for the liberation of radionuclides from gas-bearing shale
formations by the fluid fracturing process and could guide the understanding the process of
rock matrix degradation.
44
-------
Summary and Abstracts from Theme 2: Chemical
Fingerprinting
45
-------
Summary of Presentations for Theme 2: Chemical Fingerprinting
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
Technical Presentations
This first set of presentations in Theme 2 related to the determination of background ground
water and natural gas conditions.
Michael Lawson, ExxonMobil, discussed the use of isotopic tracers and other biogeochemical
techniques to understand the sources of contamination that can affect ground water. Major ion
chemistry, according to Mr. Lawson, can be used to identify contamination of ground water but
may not be able to determine the source of the contamination. Mr. Lawson stated that isotopic
tracers can be used along with major ion chemistry to identify the source of contamination and
also quantify the contributions of different sources.
Jennifer Mclntosh, University of Arizona, presented a comparison of the isotopic and chemical
composition of the Devonian New Albany Shale, Pennsylvanian coal beds, and shallow glacial
drift aquifers in the Illinois Basin. Dr. Mclntosh stated that microbial and thermogenic natural
gas from the deeper formations (shale and coal beds) can be distinguished from microbial
methane from the glacial aquifers by gas composition and carbon isotopes of methane and
carbon dioxide. Associated formation waters in shale and coal beds may be distinguished from
ground water in glacial drift aquifers by elemental and isotope chemistry, according to Dr.
Mclntosh.
Stephen Osborn, Duke University, discussed fingerprinting of formation water and gas in the
northern Appalachian Basin using carbon isotopes. According to Dr. Osborn, the gas in the
Appalachian Basin is primarily thermogenic, although there is some indication of mixing
between a larger thermogenic reservoir and a smaller biogenic reservoir shown by deviations of
linearity in the carbon isotope plots. Dr. Osborn emphasized the need for a multiple analysis
approach to understanding the origin of natural gas and formation water.
Fred Baldassare, Echelon Applied Geoscience Consulting, presented a discussion on a number
of different sources of stray gas migration in the Appalachian Basin. Sources of stray gas include
abandoned and operating coal mines, naturally occurring gas seeps, natural gas pipelines,
landfills, and buried organic matter. He emphasized that it is difficult to pinpoint the source of
natural gas migrations and that identifying the source requires the synthesis of several different
data types. Mr. Baldassare said he has investigated hundreds of stray gas cases in Pennsylvania
and found no incidents of gas migration related to HF operations.
46
-------
The second set of technical presentations addressed introduced tracers.
George King, Apache Corporation, discussed the use of chemical and gamma-emitting tracers in
conjunction with microseismic monitoring to optimize fracture operations. Gamma-emitting
tracers coat the proppant in a zero-wash coating (non-removable) and can be used to identify
where the proppant left the wellbore out to a distance of about 12 inches. The half-life of the
gamma-emitting tracers ranges between 64 and 84 days depending on the tracer type.
Microseismic data allow estimation of the shape of the fractures (as plotted by acoustic
monitoring of the shear fracturing events), and the stimulated rock volume (SRV) is roughly
outlined by the densest volume of these events. These techniques provide information for the
accurate spacing and separation of wells as well as the development of fractures along the
wellbore. Microseismic techniques can also assist in targeting the placement of fracs and, when
combined with production logging techniques, can identify the major sources of gas for better
fracture placement.
Uni Blake, Majitox for Gastem USA, presented a study design of a water quality monitoring
program at the Ross Well site in New York that consisted of baseline water testing, fracture
fluid testing, and water quality monitoring. A number of challenges were encountered by the
scientists including variability in baseline water quality between residential wells, the variability
of different additives used in different fracture jobs, and the subsurface fate and transport of
chemicals and compounds. Water samples were collected for 12 months after fracturing in a
two-mile radius around the Ross Well site. More study is needed; a next step is determining a
potential organic confirmation tracer.
The third set of technical presentations addressed tracing fracturing fluids in the environment.
Avner Vengosh, Duke University, discussed the use of several isotope-fingerprinting methods
coupled with a geochemical evaluation of the possible impacts of fracturing fluids and
formation waters on the quality of water resources in affected areas. Dr. Vengosh suggested
that strontium concentrations and isotopes may be used to identify contributions of formation
water in ground water samples. He emphasized the importance of establishing a baseline and
using a combination of tracers.
Carl Kirby, Bucknell University, provided a survey of the inorganic geochemistry of Pennsylvania
Marcellus Shale flowback waters. Flowback water chemistry (salinity in particular) was shown
to change significantly as a function of time. Dr. Kirby suggested that the strontium signature of
the Marcellus Shale flowback could serve as a tracer, and he warned against relying on stream
conductivity as evidence of formation water or flowback water discharge to a stream. Dr. Kirby
noted difficulty in acquiring samples for his study. He also introduced the Bucknell Marcellus
Shale Initiative (MSI) database.
47
-------
Summary of Discussions Following Theme 2: Chemical Fingerprinting
Presentations
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
The importance of baseline/background data and multiple tracers. A participant noted that
many sites do not have baseline water quality data, and local geologic variations complicate the
collection of representative background data. One participant suggested that gas companies
should take the responsibility to collect and provide baseline/background information. Other
participants emphasized the need for multiple tracers. One participant noted that any
assessment of potential contamination requires multiple tracers and detailed analytical work,
and HF-related testing is no exception. Other participants noted that a multiple tracer approach
will be invaluable in situations where stray or migrated gas might have originated from multiple
sources. The use of strontium isotopes as tracers has been investigated and participants
claimed that the appropriateness of their use, as with all tracers, depends on whether different
formations have different and distinct isotopic ratios. Participants indicated that phenols have
been used in the petroleum sector but there is no indication in the literature that they might be
appropriate tracers in a HF context.
Analyzing tracers. In the studies discussed in the presentations, samples were collected at the
wellhead where water and gas are already separated. A presenter stated that samples from
shallow aquifers consisted of methane dissolved in water. Generally, in the experience of the
participants, the amount of the tracer recovered is proportional to the amount of fracture fluid
recovered. Participants claimed that chemical tracers cannot provide information on the
location of fractures through which water has moved (though microseismic survey data can
help provide this information). Participants also mentioned that chemical variability within
formations needs to be taken into account when trying to determine contamination pathways.
Participants noted that concentrations of volatiles or total organic solids can be high in samples
collected from trucks or tanks in which the water has been sitting for some time.
Gamma-emitting tracers and microseismic recording. Participants noted that iridium, selenium,
and antimony are the most commonly used gamma-emitting tracers due to their reasonably
short half-lives and rarity in shales and other formations.
The New York water quality study. Ms. Blake clarified that most sampling was done within three
months of fracturing, with very few samples taken on the same day as a fracture job.
Participants noted that high methane content is possible in drinking water wells due to
naturally occurring methane seeps, and their belief that high methane levels in drinking water
are not necessarily an indication of fracturing activities.
48
-------
Well drilling and operations. Participants explained that well construction and operation are in
part characterized and optimized through the use of tracers and microseismic data before and
during drilling. In shales with extensive natural fracture systems, participants noted that less
energy and pressure is required to hydraulically fracture the formation than is required for
formations that do not have significant natural fracture systems. The success rate of re-
fracturing a shale play is 85% compared to 30% in non-shale rock, according to one participant.
Some participants emphasized that with good fracture treatment design, fractures do not
propagate upward for a significant distance (toward sources of drinking water). Some
participants stated that problems related to well-construction and the possible surface spillage
of fluids are the two main issues regarding the potential for pollution from HF. Participants also
discussed mechanical integrity tests (MITs), which are periodic tests of well operations
prescribed by state regulators. Participants indicated that, historically, western states have had
more comprehensive state regulations than Pennsylvania and New York, though Pennsylvania
recently implemented some new aspects of HF regulations.
Water fate andflowback. Participants stated that, in areas with low permeability where leakoff
is observed, injected water is most likely exploiting natural fractures and fissures that are
common in shale formations. In this situation, participants explained that the capillary blocking
pressure can increase and potentially result in damage to the formation.
Recycling. A participant asked about the effects of recycling water on tracers and fingerprinting.
Participants mentioned that recycling and mixing could pose challenges by altering water
composition. However, several participants noted that regardless of the source of the recycled
water, its chemical signature is likely to be distinct from fresh water, which should enable it to
be traced.
Migration and contamination scenarios. A participant noted that the mobility of gas is greater
than that of water and asked if the presence of a water-based tracer could ever be detected
without gas intrusion into the ground water. Other participants described a spill or other
surface accident that could cause ground water impacts. However, participants indicated that,
for most subsurface migration scenarios, the arrival of the gas at a particular monitoring point
would precede the arrival of the water. A participant noted that a leak related to a
compromised well might also be anticipated by observed pressure changes in the well.
Injection pressure in the Marcellus Shale. A participant asked about the typical injection
pressure in the Marcellus Shale and how the pressure dissipates after HF is completed. A
participant stated that the average injection pressure (measured at the surface) is in the range
of 8,000 psi and does not exceed 9,000-10,000 psi. The participant continued to explain that
instantaneous shut-in pressure is 3,000-3,500 psi, which dissipates slowly over time; in
addition, the participants noted that pressure will not dissipate completely, but rather stabilize
at a level higher than the natural level.
Abandoned wells. A participant asked about abandoned wells and the possibility of fracture
communication with these wells. Participants noted that this is always a concern, especially if
49
-------
there is any kind of wellbore within a few hundred feet of the injection well. Pressure effects
can be communicated between wells across as much as 250-300 meters, according to one
participant. This participant stated that the potential for well communication is one reason why
well operational records are so important. The participant went on to explain that in Texas and
Oklahoma, the records are fairly good regarding a list of abandoned wells, and an industry tax
contributes to the plugging of abandoned wells. In addition, the participant noted that permits
in Texas are often denied if there are any wells too close to the proposed injection/HF well. On
the other hand, a participant stated there may be 100,000 abandoned wells unaccounted for in
northwestern Pennsylvania. However, participants noted that it is important to distinguish
between shallow and deep abandoned wells. Marcellus Shale HF wells should not be affected
by properly abandoned shallow wells according to participants.
Cooperation among industry, academia, and regulatory agencies. Participants expressed
appreciation that representatives from industry, academia, commercial labs, and the
government all attended the workshop. Participants called for more cooperation among the
stakeholders and suggested that industry give universities more access for sample collection
and analysis. One participant noted that, outside the industry, people do not know many details
about the long-term effects of HF.
Analytical techniques. A participant noted that conductivity pens are very inexpensive
compared to chemical analysis. Other participants pointed out that conductivity may not be a
good indication of the source of contamination since there can be multiple sources of salinity
(e.g., road salt in surface water) other than subsurface brines, and that chemical analysis is not
that expensive compared to how critical it is. A participant noted that chemical analysis may
cost $1,000, which is worth the investment over the life of operations, research or
investigations. Participants argued that chemical analysis should not be complicated or
overwhelming; collectively, the various analyses provide critical information for managing
operations.
Drilling muds. A participant asked about the composition and use of drilling muds. Other
participants explained that mud basics are fairly simple, but that details and specifics can get
complex. Basic mud is a mix of fresh water with some added minerals and chemicals. One
participant explained that mud composition depends on the downhole temperature and
pressure, the type of formation, the need to lift cuttings, the drill bit, the size of the hole, and
other factors. The participant continued to explain that the characteristics of drilling muds can
change over time as drilling progresses and more information is collected. Participants noted
that the use of oil-based muds is limited due to concerns about environmental consequences if
they are spilled at the surface. Participants stated that there are also very expensive specialized
muds, but, for most shale HF drilling purposes, simple mud systems are sufficient. According to
participants, functions of mud are to keep the well under control, allow pressure to be read,
and circulate fluids through the well bore. When a vertical well turns horizontal, participants
explained that mud and fluids can become more complicated and more expensive. Managing
mud composition and characteristics, claimed participants, is a day-by-day or hour-by-hour job.
50
-------
Compressed air can also be used instead of mud, even when drilling to depths of 5,000-6,000
ft, as noted by participants.
HFfluids. A participant explained that there are essentially 12 different types of fracture fluids
currently in use; the fluids used in the Eagle Ford Shale are typically hybrid jobs: slickwater or
base gels, followed by crosslinked borate gels. The Eagle Ford formation is deeper and hotter
than the Marcellus according to participants.
"Bad" wells and methane hazards. One participant asked about the causes of wells going bad.
Other participants noted that saying the well is "bad" could mean several things. One
participant explained that if the well does not flow in commercial quantities, then the company
did not correctly select the location or the stimulationthat is "going bad" in the economic
sense. However, well blowouts are almost universally due to human error, according to
participants. Participants stated their belief that there are very few areas where drilling cannot
occur and drilling has occurred in some very difficult areas; however, any large industry will
have accidents. One participant emphasized the importance of identifying potential methane
hazards when drilling.
The role of the caprock. A participant asked about the mechanisms that would stop fracture
fluid from moving to the surface in different situations. Other participants explained that there
will always be a caprock or seal present above gas zones; the seal is necessary to create the gas
zone in the first place.
EPA's study. A participant asked how EPA plans to use the results of the technical workshops to
revise the draft study plan. The purpose of the workshops is for EPA to collect more information
on specific topics and thematic areas. The workshops will provide EPA with more up-to-date
information to inform the study as it develops.
51
-------
Abstracts for Theme 2: Fracture Design and Stimulation
Abstracts were submitted to U.S. EPA by the presenters for use in this proceedings document.
Not all presenters submitted abstracts of their presentations.
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
52
-------
Chemical and Isotopic Tracers of Natural Gas and Formation
Waters in Fractured Shales
Jennifer Mclntosh1, Melissa Schlegel1, Brittney Bates1
"University of Arizona, Department of Hydrology and Water Resources
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
Fingerprinting the chemical and isotopic composition of formation waters and natural gas
associated with organic-rich shales is critical for evaluating potential environmental impacts of
hydraulic fracturing for gas production. This paper summarizes recent results from Schlegel et
al. (in press) comparing the chemical and isotopic composition of formation waters and natural
gas in an organic-rich shale (Devonian New Albany Shale) in the Illinois Basin to other gas
accumulations in overlying Pennsylvanian coalbeds and shallow glacial drift aquifers (Figure 3)
to determine the best analytical tools for distinguishing gas and fluid sources. Previous data are
included from Coleman et al. (1988), Strgpoc et al. (2007), and Mclntosh et al. (2002).
(Rough Creek
] Fault System
Quaternary glacial deposits
(shallow aquifers)
Pennsylvanian coals (coals) ^x^
A'
Figure 3. Cross-section of Illinois Basin from Schlegel et al. (in press)
Geologic Background
The New Albany Shale is an
organic-rich (black) shale along
the eastern margin of the Illinois
Basin and grades into a grey-
green shale along the western
margin (Barrows and Cluff, 1984;
Hassenmueller, 1993). The shale
contains predominantly type-ll
kerogen (sapropelic-marine), with
up to 16 wt% total organic carbon
(TOC) and has low thermal
maturity (R0 <0.6%; e.g. Barrows and Cluff, 1984) except in the south near the Shawneetown-
Rough Creek fault system where the shale reaches R0 values >\% (Fig. 3b; Cluff and Byrnes,
1991). Previous studies have shown that the shale contains both microbial and thermogenic
gas. Thermogenic gas is distributed throughout the basin, while microbial gas is predominantly
located along the northeastern and eastern margins of the basin where meteoric water
infiltration likely stimulated microbial methanogenesis by decreasing formation water salinity
and transporting in near-surface microbial communities into paleo-pasteurized sediments
(Mclntosh et al., 2002; Schlegel et al., in press). New Albany Shale wells in areas of
predominantly microbial methane typically contain copious amounts of formation waters,
which must be removed for gas production.
53
-------
Pennsylvanian coalbeds are composed of type-Ill kerogen (typical of terrestrial organic matter;
Hatch et al., 1991), and have low thermal maturity (R0 <0.6%) except in the south where
tectonic activity locally increased the geothermal gradient and produced higher coal maturity
(R0 ~0.7 to 0.8%; Fig. 3a; Drobniak et al., 2004; Hower et al., 2005). Pennsylvanian coals contain
thermogenic gas in the central and southern basin with little to no co-produced formation
waters. The coals contain microbial gas, associated with variable salinity formation waters,
across the northern and northeastern basin margins (Strgpoc et al., 2007).
Glacial drift sediments containing detrital organic matter overlie much of the northern portion
of the Illinois Basin; these sediments are up to 120m thick and constitute an aquifer with high
quality drinking water (Swann, 1968). The glacial sediments contain up to 17.2 wt% TOC, which
is relatively thermally immature (Glessner and Roy, 2009). Many water supply wells screened in
glacial drift aquifers contain dissolved methane that is microbial in origin, and sourced from
biodegradation of in-situ organic matter and/or in underlying shallow Paleozoic sediments
(Colemanetal., 1988).
Chemical and Isotopic Fingerprint of Natural Gas
In general, gas accumulations in the Devonian New Albany Shale, Pennsylvanian coalbeds, and
shallow glacial drift aquifers have different hydrocarbon compositions (methane (Ci) to ethane
(C2) and propane (C3) ratios; Figure 4). Dissolved gas in the shallow aquifers is predominantly
comprised of CH4 with some C2 (in
addition to atmospheric-derived gases,
such as N2 and Ar), and no detectable
C3+. Natural gas in the shale has much
lower Ci/C2+C3 ratios (12-5931),
compared to the shallow aquifers
(>106), due to the presence of
thermogenic gas in addition to
1000000
100000 -
10000
£. 1000
I
o
100 -
-80
-75
Figure 4. Methane to ethane + propane ratios versus carbon isotopes
of methane for various gas
microbial methane. Coals have
Ci/C2+C3 ratios (188-163,361) that plot
roughly between the shallow aquifers
and shale.
-70 -65 -60 -55 -50 -45 -40
513CCH4 (%« PDB)
The carbon isotope values of methane
accumulations are also distinct
between the three organic-rich
formations. Shallow aquifers have very low 513C-CH4 values (-90 to -68%o), consistent with early
stage methanogenesis where the C02 reservoir has not been significantly depleted. In contrast,
513C values of CH4 from the Devonian shale are much higher (-57 to -45%o) likely due to mixing
with isotopically-enriched thermogenic gas, and significant depletion of the C02 reservoir
during microbial methanogenesis via C02 reduction (Schlegel etal., in press). Importantly,
microbial CH4 from the Devonian shales has similar 513C values to thermogenic CH4 (-55.9 to -
52.7%o), suggesting that carbon isotope of CH4 alone cannot distinguish mechanisms of shale
54
-------
gas generation. In contrast, 513C values of thermogenic and microbial CH4 in the Pennsylvanian
coals are distinct (Strgpoc et al., 2007). In addition, 513C values of microbial CH4 (-67 to -51%o)
in the coals plot between the shallow aquifers and Devonian shale with little overlap (Figure 4).
Microbial methanogenesis produces (and consumes) C02 in addition to CH4, with high 513C-C02
values. Carbon isotope fractionation factors (a,co2-cH4) for methanogenesis via C02 reduction
typically range from 1.05 to 1.09 (Figure 5), while a,co2-cH4 values for acetate fermentation
typically range from 1.03 to 1.05 (Whiticar et al., 1986). The majority of C02 and CH4 from the
Devonian shale, Pennsylvanian coals, and shallow aquifers plot along a carbon isotope
fractionation line of ~1.05 to 1.09, except for a few coal samples influenced by methane
oxidation and sulfate reduction
(Strapoc et al., 2007), and one
outlier glacial drift sample.
Thermogenic gas in the Devonian
shale was not analyzed for 513C-
C02 as it contains very little C02
(<0.1 mole %). Thermogenic gas in
the Pennsylvanian coals has much
-80 -70 -60
513CCH4(%oPDB)
-50
-40
lower 5 C-C02 values than
microbial gas, within the range of
C3-type organic matter.
Figure 5. Carbon isotopes of carbon dioxide versus methane for various gas
Gas samples show a progression along the carbon isotope fractionation line (Figure 5) from
more negative 513C values of C02 and CH4 in the shallow aquifers to more positive values in the
shale, with coal samples plotting in between. This may be due to mixing between a
thermogenic and microbial methane source, or the progressive depletion of the carbon (C02)
reservoir, causing increased 513CCo2 and 513CCH4- Mixing between microbial and thermogenic
methane would be seen in an increasing trend of 513CCH4 versus C2+%. Though the 513CCH4 of
microbial and thermogenic methane for the shale have similar ranges (up to -47.1%o) (Mclntosh
and Martini, 2008), a slight increasing trend is observed in the shale from -53%o to -47%o for
samples with <2% C2+, however no such correlation is observed for the shallow aquifers or the
coals. These trends suggest that though some thermogenic methane is mixing with microbial
methane, the overall progression of carbon isotopes along the fractionation line may be due to
the progressive depletion of the carbon (C02) reservoir underdosed system conditions (Jones
etal., 2008).
Together, the 513C values of C02 and CH4, in combination with gas composition (including mole
% C02, CH4, C2, C3+), clearly distinguish microbial and thermogenic gas from the Devonian shale,
Pennsylvanian coals, and shallow aquifers.
55
-------
Chemical and Isotopic
Fingerprint of Formation
Waters
Formation waters associated with
natural gas in coalbeds and fractured
shales in the Illinois Basin are enriched
in Cl and Br (>1.8 mM and >8.7 u.M,
respectively), as well as other solutes,
relative to shallow aquifers (Figure 6).
These saline fluids are likely remnant
Paleozoic brines sourced from
evaporation of seawater, which have
been subsequently diluted by meteoric
recharge and modified by water-rock-
microbial reactions (Mclntosh et al.,
10000.000
1000.000
100.000
J. 10.000
i.ooo
o.ioo
31
b?
0.010
0.0001
Shallow Aquifers
D Shallow aquifers
A Penn. Coals
O Dev. Shale
0.0010
1.0000 10.0000
0.0100 0.1000
Br (mM)
Figure 6. Chloride vs. bromide concentrations (log-scale) of formation
waters associated with various gas sources
2002). In contrast, shallow aquifers contain dilute (Cl <0.8 mM), recently recharged meteoric
waters (Mclntosh and Walter, 2006).
Groundwaters from shallow aquifers have 5180 and 5D values, which plot along the global
meteoric water line (GMWL) (-8.6 to -6.6%o, and -57 to -40%o, respectively; Figure 7). Saline
18,
fluids in the Devonian shale are more enriched in 0 and
I
co
a?
-20
-40
-60
-80
Shallow Aquifers
D Shallow aquifers
A Penn. Coals
O Dev. Shale
-10
-8
-6
-4
-2
818OH2o (%»VSMOW)
Figure 7. Stable isotopes of formation waters associated with natural gas
accumulations
coal versus water from the shale or shallow aquifers.
2H (-7.5 to -0.14%o, and -46 to -8%o,
respectively) than shallow aquifers,
and plot to the right of the GMWL,
along a mixing line between Illinois
Basin brines and meteoric water.
Stable isotopes of water may be
useful for distinguishing between
fluids sourced from the Devonian
shale versus shallow aquifers.
Formation waters in Pennsylvanian
coals have 5180 and 5D values (-7.0
to -1.9%o, and -48 to -13%,,
respectively) that overlap the
shallow aquifers and Devonian shale,
making it difficult to distinguish
between water sourced from the
Alkalinity concentrations and carbon stable isotopes provide an additional tracer of fluid
sources. Formation waters associated with microbial methanogenesis in the Devonian shale
and Pennsylvanian coals typically have high 513C values of dissolved inorganic carbon (DIG;
>10%o), and variable alkalinities (<5 to 37 meq/kg). Shallow aquifers have lower alkalinities (<10
meq/kg), and low 513C-DIC values (<0%o). Formation waters associated with thermogenic gas in
56
-------
the Devonian shale have low alkalinities (up to 2.4 meq/kg), and low 513C-DIC values (most near
0%o)(Mclntoshetal, 2002).
In short, formation waters sourced
from the Devonian shale are saline
(with high Cl, Br, and other solute
concentrations), enriched in 180 and
2H, and have high 513C-DIC and
alkalinity values, compared to dilute
meteoric waters in shallow aquifers.
Formation waters from
Pennsylvanian coals have similar
solute concentrations (Cl, Br,
alkalinity) and 513C-DIC values as
Devonian shale fluids. In addition,
their 5180 and 5D values overlap the
Devonian shale and shallow aquifer
samples, suggesting that it may be
-20
Alkalinity (meq/kg)
Figure 8. Carbon isotopes of dissolved inorganic carbon (DIG) versus
alkalinity concentrations of formation waters associated with natural gas
accumulations
difficult to distinguish fluids sourced from Pennsylvanian coals versus fluids sourced from the
Devonian shale.
Summary
Natural gas (microbial and thermogenic) from the Devonian shale and Pennsylvanian coals can
be distinguished from microbial methane sourced from shallow glacial drift aquifers based on
gas composition and 513C values of CH4 and C02. Formation waters from the Devonian shale can
be distinguished from dilute groundwater in shallow aquifers using major ion chemistry (Cl, Br,
other solutes) and stable isotopes (180,2H, 13C). It may not be possible to distinguish between
formation waters sourced from Pennsylvanian coals versus the Devonian shale based on major
ion chemistry and stable isotope compositions alone. These results are important for evaluation
of potential environmental impacts of hydraulic fracturing of shales, such as the migration of
brines and natural gas into shallow drinking water resources.
Acknowledgements
Funding for this work was provided by NSF (EAR-0635685), RPSEA, and the USGS. Anna Martini
and Lynn Walter contributed significantly to research on the New Albany Shale. Joe Wade
helped with field sample collection and Tim Corley helped with laboratory analyses.
57
-------
References
Barrows M.H. and Cluff R.M. (1984) New Albany Shale Group (Devonian-Mississippian) source
rocks and hydrocarbon generation in the Illinois Basin. In AAPG Memoir: Petroleum
geochemistry and basin evaluation (eds. G. Demaison and R.J. Murris). AAPG, pp. 111-138.
Cluff R.M. and Byrnes A.P. (1990) Lopatin analysis of maturation and petroleum generation in
the Illinois Basin. In Interior Cratonic Basins (eds. M.W. Leighton, D.R. Kolata, D.F. Oltz and
J.J.Eidel).AAPG, pp. 425-454.
Coleman D.D., Liu C.-L. and Riley K.M. (1988) Microbial methane in the shallow Paleozoic
sediments and glacial deposits of Illinois, U.S.A. Chemical Geology 71 (1-3), 23-40.
Drobniak A., Mastalerz M., Rupp J. and Eaton N. (2004) Evaluation of coalbed gas potential of
the Seelyville Coal Member, Indiana, USA. International Journal of Coal Geology 57, 265-
282.
Glessner, J.J.G. and Roy, W.R. (2009) Paleosols in Central Illinois as Potential Sources of
Ammonium in Groundwater. Ground Water Monitoring and Remediation 29 (4), 56-64.
Hassenmueller N.R. (1993) New Albany Shale (Devonian and Mississippian) of the Illinois Basin.
In Petroleum Geology of the Devonian and Mississippian Black Shale of Eastern North
America (ed. R.C. Kepferle), USGS, pp. C1-C19.
Hatch J.R., King J.D. and Risatti J.B. (1991) Geochemistry of Illinois Basin oils and hydrocarbon
source rocks. In Interior Cratonic Basins (eds. M.W. Leighton, D.R. Kolata, D.F. Oltz and J.J.
Eidel). AAPG, pp. 403-423.
Hower J.C., Mastalerz M., Drobniak A., Quick J.C., Eble C.F. and Zimmerer M.J. (2005) Mercury
content of the Springfield coal, Indiana and Kentucky. International Journal of Coal Geology
63, 205-227.
Jones D.M., Head I.M., Gray N.D., Adams J.J., Rowan A.K., Aitken CM., Bennett B., Huang H.,
Brown A., Bowler B.F.J., Oldenburg T., Erdmann M. and Larter S.R. (2008) Crude-oil
biodegradation via methanogenesis in subsurface petroleum reservoirs. Nature 451,176-
180.
Mclntosh J.C. and Martini A.M. (2008) Hydrogeochemical indicators for microbial methane in
fractured organic-rich shales: Case studies of the Antrim, New Albany, and Ohio Shales. In
Gas Shale in the Rocky Mountains and Beyond (eds. D.G. Hill, P.G. Lillis and J.B. Curtis).
Rocky Mountain Association of Geologists 2008 Guidebook, Denver, pp. 162-174.
Mclntosh J.C., Walter L.M. and Martini A.M. (2002) Pleistocene recharge to mid-continent
basins: Effects on salinity structure and microbial gas generation. Geochimica et
Cosmochimica Acta 66 (10), 1681-1700.
Mclntosh J.C. and Walter L.M. (2006) Paleowaters in Silurian-Devonian carbonate aquifers:
Geochemical evolution of groundwater in the Great Lakes region since the Late Pleistocene.
Geochimica et Cosmochimica Acta 70, 2454-2479.
Schlegel, M.E., Mclntosh, J.C., Bates, B., Kirk, M., Martini, A.M. (in press, 2011) Comparison of
fluid geochemistry and microbiology of multiple organic-rich reservoirs in a sedimentary
basin: evidence for controls on methanogenesis and microbial transport. Geochimica et
Cosmochimica Acta.
58
-------
Strgpoc D., Mastalerz Mv Eble C. and Schimmelmann A. (2007) Characterization of the origin of
coalbed gases in southeastern Illinois Basin by compound-specific carbon and hydrogen
stable isotope ratios. Organic Geochemistry 38, 267-287.
Swann D.H. (1968) A summary geologic history of the Illinois Basin. In Geology and Petroleum
Production of the Illinois Basin. Illinois Geologic Society, Evansville. pp. 3-21.
Whiticar M.J., Faber E. and Schoell M. (1986) Biogenic methane formation in marine and
freshwater environments: C02 reduction vs. acetate fermentation - Isotope evidence.
Geochimica et Cosmochimica Acta 50, 693-709.
59
-------
Distinguishing the Source of Natural Gas Accumulations with a
Combined Gas and Co-produced Formation Water Geochemical
Approach
Stephen Osborn1 and Jennifer Mclntosh2
1Duke University, Center on Global Change
2University of Arizona, Hydrology and Water Resources
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
Identifying the source of dissolved gas accumulations in shallow groundwater systems may
provide some insight into potential impacts from underlying geologic formations targeted for
natural gas production and associated with drilling and hydro-fracturing operations. This
presentation describes a combined gas and formation water geochemical approach for
distinguishing relatively shallow microbial methane and gas sourced from depth by the thermal-
catalytic breakdown of organic matter (thermogenic gas). Recent results from the northern
Appalachian Basin, detailed in Osborn and Mclntosh (2010), will be presented and highlight this
approach using gas composition and carbon isotope (613C) values of methane and higher chain
hydrocarbons (ethane, propane, and butane) combined with elemental (major) and isotopic
analyses (613C-DIC) of formation waters.
Distinguishing the Source of Natural Gas
Thermogenic gas has a greater proportion of higher chain hydrocarbons (ethane, propane,
butane, and pentane) relative to methane than microbial gas, which dominantly contains CH4
and C02 (Schoell, 1980). Thus, gas composition may be used as a first order approximation
distinguishing between microbial and thermogenic gas. Carbon isotope values of methane
(613C-CH4) are widely used to further constrain the origin of natural gas. Schoell (1980)
suggested that 613C-CH4 values less than -64%o and ethane concentration less than 0.5 mole%
(methane to higher chain hydrocarbon ratio greater than 1000) may represent a purely
microbial gas end-member, where as more positive 613C-CH4 values (up to approximately -
50%o) may represent an admixture of thermogenic and microbial gas (Bernard, 1978; Schoell,
1980; Whiticar et al., 1986). Methane to higher chain hydrocarbon ratios less than
approximately 100 and 613C-CH4more positive than -50%o have been interpreted as indicative
of thermogenic gas (Bernard, 1978; Schoell, 1980; Whiticar et al., 1986). Studies on
thermogenic gas have focused on higher chain hydrocarbons to understand mixing
relationships among different sources of gas.
60
-------
Chung et al. (1988) proposed a kinetic fractionation model (natural gas plot) that assumed a
linear correlation between carbon isotope values of methane, ethane, propane, butane, and
pentane (Ci-C5); and the reciprocal carbon number as a qualitative indicator of purely
thermogenic gas. Deviations from linearity in the heavier hydrocarbons (C2-C5) may be
interpreted as a mixture of different thermogenic gas end-members and/or reflecting the
heterogeneity within a single organic matter source. Depleted carbon isotope values of
methane relative to a linear thermogenic end-member may result from mixed thermogenic and
microbial gas. The sole use of gas composition and 613C values to determine the origin of
natural gas is complicated by microbial production of higher chain hydrocarbons (i.e., ethane
and propane) and secondary modifications during microbial oxidation and diffusive
fractionation and mixing between multiple gas sources (Barker and Fritz, 1981; Prinzhofer and
Pernaton, 1997; Boetius et al., 2000; Taylor et al., 2000; Martini et al., 2003; Hinrichs et al.,
2006). Thus, additional indicators from formation water geochemistry are necessary to further
constrain the origin of natural gas.
Methanogens extract hydrogen from shale organic matter and in-situ formation waters to
produce methane. The covariance of hydrogen isotope values of water and methane have been
used as a strong indicator methanogenesis, as well as identifying the predominant metabolic
pathways for methane generation, as C02 reducing methanogens impart a different hydrogen
isotope fractionation factor (aCH4-H2o) than acetate fermenting methanogens (Schoell, 1980;
Whiticar et al., 1986; Martini et al., 1996). Production of C02 causes formation waters
associated with methanogenesis to be enriched in alkalinity (>10 meq/kg), and have very
positive 613C values of dissolved inorganic carbon (DIG) (>10%o).
The geochemical conditions (salinity and redox) favorable for microbial metabolism are also
important for identifying methanogenesis. Microbial methane generation is inhibited at high
salinities and sulfate concentrations. A salinity level of approximately 2,000 mmole/L chloride
has been proposed as a boundary for the onset of toxicity for methanogens, however, a range
of tolerances is observed up to 4,000 mmole/L (Zinder, 1993; Ollivier et al., 1994; Waldron et
al., 2007). Sulfate reducing bacteria (SRB) begin to out-compete methanogens in freshwater
sediments for hydrogen and acetate at sulfate concentrations greater than 1 mmole/L (Lovley
and Klug, 1982).
Results from the Appalachian Basin
Together the formation water and gas results suggest that the vast majority of methane in the
Devonian organic-rich shales and sandstones across the northern Appalachian Basin margin is
thermogenic in origin. Small accumulations of microbial methane are present at shallow depths
along Lake Erie and in western NY. This is based on the observation that the majority of natural
gases in Devonian organic-rich shales and sandstones at depth (>168 m) in the northern
Appalachian Basin have a low methane to higher chain hydrocarbon ratio (3-35 mole%) with
high 613C values of CH4 (-53.35 to -40.24%o).
Dissolved gases in shallow groundwaters in Devonian organic-rich shales along Lake Erie contain
detectable methane (0.01 to 50.55 mole%) with low 613C-CH4values (-74.68 to -57.86%o) and
61
-------
no higher chain hydrocarbons, characteristics typical of microbial gas. Nevertheless, these
groundwaters have only moderate alkalinity concentrations (1.14 to 8.72 meq/kg) and
relatively low 613C values of dissolved inorganic carbon (DIG) (-24.8 to -0.6%o), suggesting that
microbial methanogenesis is limited.
The geochemistry of co-produced brines shows evidence for microbial activity. High 613C values
of DIG (>+10%o), slightly elevated alkalinity concentrations (up to 12.01 meq/kg) and low S04
values (<1 mmole/L) in select Devonian organic-rich shale and sandstone formation water
samples suggest the presence of methanogenesis, while low 613C-DIC values (<-22%o) and
relatively high sulfate concentrations (up to 12.31 mmole/L) in many brine samples point to
sulfate reduction, which likely limits microbial methane generation in the Appalachian Basin.
Conclusions
Results from this case study demonstrate the importance of a combined gas and formation
water approach for constraining the origin of natural gas in sedimentary basins. This approach
may provide useful in understanding the origin of natural gas accumulations in shallow
groundwater systems that may be associated with drilling and hydro-fracturing operations.
Acknowledgements
Research support was provided by the New York State Energy Research and Development
Authority, the American Chemical Society-Petroleum Research Fund (#45284-G2), and the
United States Geologic Survey. We thank Robert Jackson, Avner Vengosh, and Nathaniel
Warner from the Center on Global Change and the Earth and Ocean Science Department at
Duke University for support during the preparation of this extended abstract.
62
-------
References
Barker, J.F., Fritz, P., 1981. Carbon isotope fractionation during microbial methane oxidation.
Nature 293, 289-291.
Bernard, B.B., 1978. Light hydrocarbons in marine sediments. Ph.D. Dissertation, Texas A&M
University, College Station, Texas.
Boetius, A., Ravenschlag, K., Schubert, C.J., Rickert, D., Widdel, F., Gleseke, A., Amann, R.,
Jergensen, B.B., Witte, U., Pfannkuche, 0., 2000. A marine microbial consortium
apparently mediating anaerobic oxidation of methane. Nature 407, 623-626.
Chung, H.M., Gormly, J.R., Squires, R.M., 1988. Origin of gaseous hydrocarbons in subsurface
environments: Theoretical considerations of carbon isotope distribution. Chemical
Geology 71, 97-103.
Hinrichs, K.U., Hayes, J.M., Bach, W., Spivack, A.J., Hmelo, L.R., Holm, N.G., Johnson, C.G., S.P.
Sylva, 2006. Biological formation of ethane and propane in the deep marine subsurface.
PNAS 103, no. 40, 14684-14689.
Lovley, D.R., Klug, M.J., 1982. Sulfate Reducers Can Out-compete Methanogens at Freshwater
Sulfate Concentrations. Applied and Environmental Microbiology 45, no. 1, 187-192.
Martini, A.M., Budai, J.M., Walter, L.M., Schoell, M., 1996. Microbial generation of economic
accumulations of methane within a shallow organic-rich shale. Nature 383,155-158.
Martini, A.M., Walter, L.M., Budai, J.M., Ku, T.C.W., Mclntosh, J.C., Schoell, M., 2003. Microbial
production and modification of gases in sedimentary basins: Ageochemical case study
from a Devonian Shale gas play, Michigan Basin. AAPG Bulletin 87, no. 8, 1355-1375.
Ollivier, B., Caumette, P., Garcia, J.-L, Mah, R.A., 1994. Anaerobic bacteria from hypersaline
environments. Microbiology Reviews 58, 27-38.
Osborn, S.G. and Mclntosh, J.C., 2010, Chemical and isotopic tracers of the contribution of
microbial gas in Devonian organic-rich shales and reservoir sandstones, northern
Appalachian Basin, Applied Geochemistry, 25, pp. 456-471.
Prinzhofer, A, Pernaton, E., 1997. Isotopically light methane in natural gas: bacterial imprint or
diffusive fractionation?. Chemical Geology 142, 193-200.
Schoell, M., 1980. The hydrogen and carbon isotopic composition of methane from natural
gases of various origins. Geochimica et Cosmochimica Acta 44, 649-661.
Taylor, S.W., Sherwood Lollar, B., Wassenaar, L.I., 2000. Bacteriogenic Ethane in Near-Surface
Aquifers: Implications for Leaking Hydrocarbon Well Bores. Environmental Science and
Technology 34, no. 22, 4727-4732.
Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and
freshwater environments: C02 reduction vs. acetate fermentation - Isotope evidence.
Geochimica et Cosmochimica Acta 50, 693-709.
Whiticar, M.J., Faber, E., Schoell, M., 1986. Biogenic methane formation in marine and
freshwater environments: C02 reduction vs. acetate fermentation - Isotope evidence.
Geochimica et Cosmochimica Acta 50, 693-709.
Zinder, S.H., 1993. Physiological ecology of methanogens. In: Ferry J.G. (Eds.), Methanogenesis
Ecology, Physiology, Biochemistry, and Genetics. Chapman and Hall, 128-206.
63
-------
The Origin of Some Natural Gases in Permian through Devonian
Age Systems in the Appalachian Basin & the Relationship to
Incidents of Stray Gas Migration
Fred Baldassare
Echelon Applied Geoscience Consulting
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
The molecular composition and isotope geochemistry for natural gases of Permian through
Devonian age system rocks across the Central & Northern Appalachian Basin reveal gases are
microbial, thermogenic, and mixed microbial/thermogenic in shallow systems, and thermogenic
gases of different thermal maturities in commercially productive formations.
Since 2008, the pace of gas well drilling to the Marcellus shale has markedly increased in the
Central & Northern Appalachian Basin. This pace is expected to continue as demand for natural
gas grows. As drilling activity to the Marcellus shale has increased so have the number of
reports of stray natural gas migration to shallow aquifer systems.
As allegations of hydraulic fracturing contaminating groundwater supplies received the
attention of various media outlets, the link to stray natural gas contamination of aquifer
systems created by hydraulic fracturing was simultaneously made, inexplicably. Pennsylvania
leads all states in the number of Marcellus shale gas wells drilled and in production.
Investigations of stray natural gas migration incidents in Pennsylvania reveal that not one
incident of stray natural gas migration to a shallow aquifer system was caused by hydraulic
fracturing of the Marcellus shale.
Background
Methane is the principal hydrocarbon detected in all Permian and Devonian age system rocks
across the Appalachian Basin and in all stray gas migration incidents. It is the simplest and most
abundant gas hydrocarbon, and is generated by microbial and thermogenic processes. Methane
has a specific gravity of .555, and a solubility of 28-32 mg/l in water at standard atmospheric
pressure (methane solubility increases with increases in hydrostatic pressure).
Though incidents of stray methane gas and stray methane gas migration have received
significant attention as gas well drilling to the Marcellus shale has increased, the occurrence of
methane gas, and stray gas migration incidents to shallow aquifer systems have occurred long
before the start of the Marcellus shale drilling industry. This is particularly evident in
Pennsylvania where there are often multiple potential sources of methane and subsurface
methane migration. Potential anthropogenic sources of stray methane migration include
64
-------
abandoned and operating coal mines, abandoned and operating gas wells, natural gas storage
facilities, coalbed methane wells, buried organic matter, and landfills.
Methane is also naturally occurring in some shallow, near surface environments. Microbial
methane is generated by bacterial decomposition of organic matter through C02 reduction and
acetate fermentation pathways. Thermogenic methane also occurs naturally in some aquifer
systems in the Central and Northern Appalachian basin. (Laughrey, Baldassare, 1998;
Baldassare, Laughrey unpublished data, 1999-2011; Revesez, and others, 2010)
Genetic Characterization of Natural Gas
Gas-origin/source correlations are determined from geochemical parameters including
molecular composition, (relative concentrations of CH4 to C6 hydrocarbons and non-
hydrocarbon gases including He, H, C02, and N2), and stable carbon and hydrogen isotope
ratios. Analyses of the molecular composition, stable carbon and hydrogen isotopes of Cl - C5
hydrocarbons, and the radioactive isotope of Cl provide the highest level of geochemical
evidence to interpret the origin of methane and other hydrocarbons in the subsurface.
Various researchers have determined by examination of stable carbon and hydrogen isotopes
of methane that there are common isotopic compositions for thermogenic gas associated with
coal and natural gas, drift gas, and other near-surface microbial gases (Coleman and others,
1977; Schoell, 1980; Whiticar,1986; Baldassare and Laughrey, 1997; Jenden and others, 1993;
Kaplan and others, 1997; Rowe and Muehlenbachs, 1999, Revesz, and others, 2010; and Osborn
and Mclntosh, 2010).
Stable isotope compositions are expressed as
<5 =(Rsample-Rstd)/Rstd:
SE = SfE) = <5(iE/jE) = /V(iE)B//V(JE)B-/V(iE)std//V(jE)std
/V('E)std//V(JE)std
where 5('E) refers to the delta value of isotope number / and j (heavy and light isotopes,
respectively) of element E of sample B relative to the same element E in an international
measurement standard (std); Rsample=/V('E)B//V(yE)B and Rstd=/V('E) std//V(yE)std are the ratios
of the number of isotopes in unknown sample (B) and the international measurement standard.
A positive 5('E) value indicates that the sample is more enriched in the heavy isotope than the
international measurement standard. A negative 5('E) value indicates that the sample is
depleted in the heavy isotope relative to the international measurement standard (Revesz, and
others, 2010). Stable carbon isotope ratios are relative to Vienna Pee Dee Belemnite (VPDB)
standard, and hydrogen isotope ratios are relative to Vienna Standard Mean Ocean Water
(VSMOW) standard. (Gat and Gonfiantini, 1981)
In general, microbial methane formed by the acetate fermentation pathway (marsh gas &
landfill gas) reveals isotope compositions of 613C1: -40 to -62 o/oo and 6D: -270 to -350 o/oo.
65
-------
Microbial methane formed via the C02 reduction (drift gas) pathway reveal isotope
compositions of 613C1: -62 to -90 o/oo and 6D: -180 to -240 o/oo. Thermogenic methane
reveal isotope compositions of 613C1: -26 to -50 o/oo and 6D: -110 to -250 o/oo as depicted in
Figure 9.
Additionally, 613C values of dissolved inorganic C (DIG) compared to the DIG in background
groundwater can reveal additional evidence of methane origin in groundwater. Microbial
processes such as CH4 production or CH4 consumption by oxidation can impart unique isotopic
signatures to gases and components of the associated waters. Using a combined gas and water
approach is significant to constrain the origin of natural gas in water wells. Microbial CH4
production by acetate fermentation results in C and H stable isotope signatures in associated
waters that are different from microbial CH4 produced by C02 reduction (Whiticar and Faber,
1986; Revesz et. al., 2010; Osborn, Mclntosh, 2010).
o
If
§
-100
-150
-200
-250
-300
-350
Stable carbon and hydrogen isotopic compositional ranges of methanes
from different sources.
Mixed gas
X
/ Sub-surface microbial gas
i (deep-sea sediments
and drift gas)
Thermogenic gas
(natural gas and
coalbed gas)
..'' Near-surface
microbial gas
! (marsh gas and
landfill gas).../
-90 -80 -70 -60 -50 -40
513C of CH4 (%o) (relative to VPDB)
-30
After Coleman and others (1993) based on the data set of Schoell (1980)
Figure 9. Isotope geochemistry for Marcellus Shale gas have expanded our view to reveal more thermally mature
thermogenic methane than fit within the standard isotope crossplot, and that also reveal isotope reversals (Cl, C2)
throughout the stratigraphic section. Similarly, stable carbon and hydrogen isotope compositions reveal a broader range for
early thermogenic methane.
66
-------
ISOTOPE CROSSPLOT
-100
£: -200
0)
as
"S
S
s
a
to
-300
Upper Devonian
Marcellus (SW PA)"
Themogenic Gas
("natural gas" and
coalbed methane)
Microbial Gas - ^\,
teduction /j
(driftgas) f-X *
early thermoaeiiic
Microbial Gas -
Fermentation
(landfill gases)
Marcellus (NC PA]
-90
-80
-70
-60
-50
-40
-30
-20
Representati've Thermogenic gases
Central/Northern Appalachian Basin
8Cof Methane (%c
Figure 10 provides a small datasetof a much broader database documenting isotope compositions of thermogenic gases in areas
across the Central and Northern Appalachian basin. In addition, recent geochemistry for Marcellus shale gas (unpublished data,
Baldassare, 2011) and deeper natural gas resources in areas of the Appalachian basin reveal a complicated thermogenic history
that cannot be explained by current hydrocarbon generation models (Burruss, Laughrey, 2010).
Marcellus gas well
Bradford County, PA
-50.0 -45.0
-25.0
-20.0
Upper Devonian
(Catskill fm?)
Marcellus shale
(note isotope reversal)
10000
12000
Figure 11 provides a small datasetof a much broader database documenting isotope compositions of thermogenic gases in areas
across the Central and Northern Appalachian basin. In addition, recent geochemistry for Marcellus shale gas (unpublished data,
Baldassare, 2011) and deeper natural gas resources in areas of the Appalachian basin reveal a complicated thermogenic history
that cannot be explained by current hydrocarbon generation models (Burruss, Laughrey, 2010).
67
-------
Conclusions
Defining the source of natural gas migration requires the investigation and synthesis of
different data types, notably gas geochemistry and mechanism of migration. Geochemical
evidence reveals that microbial and early thermogenic methane occurs naturally (pre-existing
anthropogenic activity) in some shallow aquifer systems in areas of the Appalachian Basin .
Water wells drilled to these formations create low pressure zones and potentially a migration
point for naturally occurring methane or stray methane from anthropogenic sources. Building
structures and other low pressure zones also represent potential migration pathways.
Given the occurrence of multiple potential anthropogenic and naturally occurring stray gas
sources, it is essential that a thorough characterization and definition of background
groundwater quality is implemented to define pre-existing conditions prior to drilling activity.
Stray gas migration incidents should be thoroughly investigated and supported by multiple lines
of evidence, principally, geochemistry & analyses documenting a mechanism of migration.
68
-------
References
Baldassare, F.J., Laughrey, CD., 1997, Identifying the source of stray methane by using
geochemical and isotopic fingerprinting: Environ. Geosc. 4(2), 85-94.
Burruss, R.C., Laughrey, C.D., 2010, Carbon and Hydrogen Isotopic Reversals in Deep Basin Gas:
Evidence for Limits to the Stability of hydrocarbons, Organic Geochem., 2010; doi:
10.1016/j.orggeochem.
Coleman, D.D., Meents, W.F., Liu, C.L., Keogh, R.A., 1977. Isotopic identification of leakage gas
from underground storage reservoirsA progress report. Illinois State Geol. Surv.,
Illinois Petroleum No. 111.
Gat, J.R., Gonfiantini, R., 1981. Stable isotope hydrology, deuterium and oxygen-18 in the water
cycle. International Atomic Energy Agency, Technical Report Series No. 210.
Jenden, P.D., Drazan, D.J., Kaplan, I.R., 1993. Mixing of thermogenic natural gases in northern
Appalachian Basin. Am. Assoc. Petrol. Geol. Bull. 77(6), 980-998.
Laughrey, C.D., Baldassare, F.J., 1998. Geochemistry and origin of some natural gases in the
Plateau Province, C. Appalachian Basin, Pennsylvania and Ohio. Am. Assoc. Petrol. Geol.
Bull. 82(2), 317-335.
Osborn, S.G., Mclntosh, J.C, 2010. Chemical and isotopic tracers of the contribution of microbial
gas in Devonian organic-rich shales and reservoir sandstones, northern Appalachian
Basin. Appl. Geochem. 25, 456-471.
Revesz, K.M., Breen, K.M., Baldassare, A.J., and Burruss, R.C.,2010. Carbon and hydrogen
isotopic evidence for the origin of combustible gases in water-supply wells in north-
central Pennsylvania. Appl. Geochem. 25, 1845-1851
Rowe, D., Muehlenbachs, K., 1999. Isotopic fingerprints of shallow gases in the Western
Canadian sedimentary basin: tools for remediation of leaking heavy oil wells. Org.
Geochem. 30, 861-871.
Schoell, M., 1980. The hydrogen and carbon isotopic composition of methane from natural
gases of various origins. Geochem. Cosmochem. Acta 44(5), 649-661.
Whiticar, M.J., Faber, E., 1986. Methane oxidation in sediment and water column environments
-Isotope evidence. Org. Geochem. 10, 759-768.
Whiticar, M.J., 1999. Carbon and hydrogen isotope systematics of bacterial formation and
oxidation of methane. Chem. Geol. 161, 291-314.
69
-------
Tracking Fracture Fluid Movement with Chemical and Gamma-
emitting Tracers with Verification by Microseismic Recording
George E. King
Apache Corporation
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Chemical tracers, gamma-emitting tracers, microseismic (acoustic monitoring of small breakage
sounds made by the rock during fracturing) as well as other simple measurements during and
after fracturing can be very useful in describing hydraulic fracture development. The output
from such monitoring can supply information on fracture complexity, frac conductivity, height
growth, frac barrier effectiveness, well-to-well and frac-to-frac interference, water entry points
and general fracturing execution.
This work focuses on both the "certainty" of the data (and sources) and the linkages between
measurements with the intent to improve understanding of both the reliability of the data and
the unknowns that can be addressed, either in part or in full, by the collected data.
Frac optimization begins with understanding the elements that control the placement and flow
behaviors in a well. High permeability, conventional reservoir rocks basically have one-factor-
dominated systems of porosity, permeability and fluid saturations. Such formations are
relatively easy to describe in numerical units that yield to reservoir models and predictive
behavior. As formation grain size, permeability and porosity reach the values in the rocks
classified as potential-productive "shales", the permeability drops into the 100's of nano-
Darcies. Effective, interconnected porosity may exist in fractures, interstitial portions of the
fine-grained rock matrix and in openings within the kerogen, created by shrinkage and
alternation of the organic content during maturation and subsequent hydrocarbon expulsion.
As shale forms gas storage and flow structures, the rock fabric becomes radically different from
convention reservoirs with natural fractures and weak zones within the fabric offering potential
for complex fracturing and very unusual fluid behavior.
Field data sets are rarely perfect or complete, so having a number of data sources and an
understanding of the type of information that can be drawn from often disparate data sources
plus the level of accuracy possible from these measurements is very useful. The sources of
information in well work may include:
3-D Seismic is well known for identification of geologic hazards, but is also important in
combination with overlays of microseismic, micro-losses of mud and frac breakdown
pressures to identify natural fracture locations.
Geological structure mapping is valid for locating structures, reefs and areas of uplift
that may be associated with increased incidence of natural fractures.
70
-------
Shale fabric considerations are part of the wider reaching candidate and sweet-spot
identification as well as evaluation of interconnection of shale variables such as
mineralogy, multiple porosity types and natural fracture occurrences.
Frac Design and Pumping Records - pumping behavior (pressure, rate, loading)
frequently raise questions about what is happening in frac initiation, frac extension and
overall frac growth. Some frac analysis, either during or post frac, have been accepted in
shale fracturing while others are of questionable use in the ultra-low permeability shales
where most fluid loss is through the fractures in pressure dependent leakoff.
Microseismic with pressure, rate and prop loading matched to the event time is a useful
tool for tracking shear fracturing events that are common in many shale fracs where
natural fractures open and form the flow paths and the extensive fracture-to-formation
contact areas necessary for shale development. Microseismic is commonly used to
describe the stimulated reservoir volume (SRV) in shales. Low-level gamma energy
tracer-marking of proppant has developed many uses and supplies conformation of frac
initiation points, near wellbore isolation between fracs in multi-frac wells, verifies frac
diversion, proppant interference from frac-to-frac and even well-to-well on moderately
closely spaced wells.
Chemical tracers in the backflow from both stimulated wells and offset wells have been
utilized for tracking water return from individual stages, polymer clean-up, well-to-well
frac interference and possibly helping confirm complex fracture development.
Recovered brine salinity, volume and ion analysis describe the return of frac fluid, the
amount of mixing between frac water and shale connate fluids at time of sampling. Care
must be exercised with this technology since single data points are often skewed by
irregular flow patterns of the well (e.g., slugging). Using the data as a trend over time is
the only accurate use.
Production logging can be the final word on production from each frac stage; however,
flow in horizontal wells is complex and examples of active production and simultaneous
counter-current flow (into the well) are well described in both physical simulators and
actual wells. Correct application of this technology is very valuable to identification and
quantification of fluid entry and exit points.
Production plots contribute significantly to evaluation of overall frac performance in a
well-known area and as a piece of data in evaluating new areas.
Tracing Examples
One of the first necessities is to lay out the wall path with as much support information as
possible. Figure 12 illustrates such a layout with well path, structural impacts, faults below and
possibly through the play, the potential well-to-well linkage from expected frac direction, and
comparison of frac behavior of offset wells.
71
-------
-2.000 -1.MO 1.0
Frac spead and well
layout details,
Including primary frac
direction to the NE and *'
secondary frac
direction to the SE. *m Possible Faults
Frac complexity
development average
using Cipolta's FCI is
approximately 2.0
(mlcroseimic event
width about twice the
half length generation.
Figure 12. A sequential or zipper traced well pair in Tier 2 of the Barnett was drilled in a NNW direction in an area with
possible faulting in the Ellenberger (immediately below the Barnett) with general primary frac direction of NE/SW. Well T1H
has 8 stages; well T2h has 10 stages. Each stage was approximately 200,000 Ib 100 mesh sand (1000 Ib/ft); 7400 bbls water (
40 bbls/ft of lateral); 6 perf clusters per frac stage; 1 bpm/perf; design frac rate was 50 bpm.
In this first example case, the available assessment information includes chemical fluid tracer
(CFTs), proppant tracers ("100" mesh sand), microseismic monitoring and frac pump charts. The
initial production or IP (24 hr) was 2.3 mmscf/d in the T2H and 1.0 mmscf/d in the T1H.
The first post-drilling information is usually the frac treatment pressure response, Figure 13.
The break in pressure in stage 1 after about an hour on the T2H well was unusual, but the
microseismic showed that the events were still solidly in the target zone, although frac
direction had changed.
JJOOO 1,W» -1,CM
2MO t.VM I.40*
3JKH!
Figure 13 shows an unexplained pumping pressure break at about 60 minutes.
Fracture initiation pressure (breakdown) is low in this area and is marked by frequent frac
direction switching along the natural fractures as often captured by the microseismic. The
technique of ramping up slowly was common to allow the frac to open natural fractures and
initiate active complex fracturing. This avoided both planar fracturing and breaking out of zone
(important to avoid the salt water containing Ellenberger formation below the Barnett). Stage 1
(the toe stage) on the T1H well exhibited nearly the same behavior, Figure 14.
72
-------
'3.
Figure 14. Near identical primary and secondary frac behavior on the T1H stage 1. The breakout pressure shift was hidden in
the ramp up but occurs at about 45 minutes.
Flow back monitoring of returning frac and formation fluids was accomplished with chemical
tracer monitoring and a logging run to measure the gamma levels along the wellbore. Results
are in Figure 15.
Other Information Sources
Surface Treating Pressure (STP) is one source of information that is always available.
Understanding the pressure relationship to down-hole events is never an exact method, but
there are many cases where this surface information does have a strong linkage to
demonstrated events. The closest agreement is often generated in a field-specific study and
demonstrated knowledge of the geologic area. One example is the behavior of the net pressure
development (rate of increase or decrease) during the fracs. This data has been more useful in
specific shales such as the Barnett than any other conventional or unconventional formation
(King, 2008; King, 2011).
The earliest data from the frac is the breakdown and the well's reaction to the increases in rate
and proppant. Increasing the rate after the initial breakdown in increments of 5 to 10 bpm was
recognized as a way to keep the frac in zone, first in the Devonian shale and later in the Barnett
Shale (Yost, 1988; King, 2008; King, 2010). The explanation then and now is that the
incremental increase gives the natural fractures time to break down and start the initial
complexity development. After breakdown, every incremental increase in rate brings a sharp
rise in pressure followed by a sharp drop grading into an attenuated decline as the fracture
opens and reaches a steady leakoff rate. Notice that the declines in pressure become shallower
and less in total in the later stages of rate increase. This is the effect of hydraulic diversion as
friction through the perforations begins to control the flow rate into the subject perforations.
73
-------
Figure 15 for T1H (left) and T2H (right) are the mass balances of fluid tracer flow coupled with information on frac
breakdown from proppant tracer. The information from fluid tracing shows CFT concentration measurements in the
backflow against volume produced (left side of each figure) are useful for estimating communication of the frac stage with
the wellbore and how that communication changes over time or volume produced. The left side of each figure is the amount
of initial fluid tracer recovered from each traced stage and is a reflection of the amount of traced load fluid actually
recovered. The shaded area in the center is the contribution of each stage to the flow as calculated from the mass balance.
74
-------
Selected References
Britt, L.K., Schoeffler, J.: "The Geomechanics of a Shale Play: What Makes a Shale Prospective!",
Paper SPE 125525, presented at 2009 SPE Eastern Regional Meeting, Charleston, WVA,
USA, 23-25 September.
Cipolla, C.L., Lolon, E.P., Dzubin, B.: "Evaluating Stimulation Effectiveness in Unconventional Gas
Reservoirs," Paper SPE 124843, presented at 2009 SPE Annual Technical Meeting, New
Orleans, LA, USA, 4-7 October.
King, G.E.: "Thirty Years of Shale Gas Fracturing: What Have We Learned?," SPE 133456,
Presented at the 2010 SPE Annual Technical Conference and Exhibition, Florence, Italy,
Spetember 20-23, 2010.
King, G.E., Haile, L. Shuss, J, Dobkins, T.A.: "Increasing Fracture Path Complexity and Controlling
Downward Fracture Growth in the Barnett Shale," Paper 119896, presented at the 2008
SPE Shale Gas Production Conference, Fort Worth, TX, 16-18 November.
King, G.E., Leonard, Dick: "Deciphering Chemical Tracer Results in Multi-Fractured Well
Backflow in Shales: A Framework for Optimizing Fracture Design and Application," SPE
140105, SPE Hydraulic Fracturing Meeting, The Woodlands, TX, USA, 24-26 Jan 2011.
Warpinski, N.R.: "Microseismic Monitoring: Inside and Out," SPE Distinguished Author Series,
J.P.T., November 2009, PP-80-85.
Warpinski, N.R., Mayhofer, M.J., Vincent, M.C., Cipolla, C.L., Lolon, E.P.: "Stimulating,
Unconventional Reservoirs: Maximizing Network Growth While Optimizing Fracture
Conductivity," Paper SPE 114173, presented at the 2008 Unconventional Reservoirs
Conference, Keystone, Colorado, 10-12 February.
Woodroof, R.A., Asadi, M., Warren, M. N.,: "Monitoring Fracturing Fluid Flowbackand
Optimizing Fluid Cleanup Using Frac Tracers," Paper SPE 82221, presented at 2003 SPE
European Formation Damage Conference, The Hague, The Netherlands, 13-14 May.
Woodroof, R.A., Asadi, M., Leonard, R.S., Rainbolt, M.: "Monitoring Fracturing Fluid Flowback
and Optimizing Fracturing Fluids Cleanup in the Bossier Sand Using Chemical Frac
Tracers," Paper SPE 84486, presented at 2003 SPE Annual Technical Conference and
Exhibition, Denver, CO, USA, 5-8 October 2003.
75
-------
Designing a Water Quality Program for Shale Exploration
Uni Blake
Majitox for Gastem USA
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
Protecting water quality is the key objective of developing a comprehensive water quality
monitoring program. The design of an effective program requires characterizing and
understanding all the chemical variables associated with the ambient water resources and the
natural gas well site. This includes characterizing the: (1) quality of the surrounding water
resources before any natural gas exploration activity starts (establishing a baseline), (2)
additives used in hydraulic fracturing, (3) flowback fluids, and (4) ambient water after hydraulic
fracturing activities. The combination of the above information is crucial when it comes to
delineating and differentiating potential contaminations that may occur as a result of hydraulic
fracturing.
Project Goals
There is concern about the potential contamination of surface and ground water primarily by
the additives used in the hydraulic fracturing process and secondarily by the flowback from the
well after the frac job is completed. The flowback contains some additives, salts, metals and
organic compounds. Understanding the exact nature of these chemical streams is important
especially when it comes to selecting indicator parameters that can be used to trace sources or
track a contamination plume.
Project Design
Gastem USA designed a water quality monitoring program for use at the Ross Well site in
Maryland, NY.
1. A sampling and analysis plan which incorporated quality assurance and quality control
protocols was written to ensure that the data collected was accurate, precise,
representative and complete.
2. A two mile radius centered at the Ross Well site was established as the area of interest.
This region was determined to have the greatest potential of being impacted by
activities at the well site because of proximity.
3. Analytical data from flowback fluid studies was gathered to aid in the development of
the monitoring program (Hayes, 2009, Clark, 2009). This data was instrumental in
selecting potential sentinel indicators and chemical tracers for potential contamination
caused by hydraulic fracturing.
76
-------
6.
Water samples were collected at the surface and ground water (private wells) locations
and tested for general water quality indicators and for other parameters that are
associated with natural gas exploration and development activities. See Table 3.
The Well underwent hydraulic fracturing in the fall of 2008. Day one flowback water was
sampled and analyzed. This water contained mostly additives and their chemically
transformed constituents.
Water samples from the area of concern were collected for a twelve months following
the hydraulic fracturing of the Shale.
Table 3. Parameters Tested in Surface Water and Residential Wells
Dissolved methane, propane and ethane
Glycols
Conductivity
Alkalinity
Total Dissolved Solids
Calcium
Sodium
Iron
Arsenic
Chromium
Lithium
Volatile Organic Compounds
pH
Chloride
Sulfate
Total Suspended Solids
Magnesium
Potassium
Manganese
Barium
Lead
Strontium
Challenges
There were challenges faced when developing the Shale water monitoring.
Variability in baseline water quality: Residential water wells vary in depth and aquifer
location. This variability has an impact on the concentrations of natural constituents
found in the groundwater. It is known that drinking water wells located close to natural
gas deposits may contain elevated levels of BTEX, methane, and strontium (Lesage et al.,
1997). Gastem found that some wells had detectable levels of methane, and some had
constituents that exceeded local water quality standards. This variability has an impact
on creating a representative baseline water quality database. Solution: A team of
stakeholders in Otsego County are working on developing a program that will
characterize groundwater quality.
Additives. Additive selection depends on the nature of each frac job, type of well, depth
of the well, and the company performing the hydraulic fracturing. Solution: Comparing
fracturing fluids collected during the stimulation of various target formations to
determine if the variability has an appreciable impact.
Chemical fate and transport. Fate and transport mechanisms of the potential
contaminants have the ability to change the distribution and concentrations of the
contaminants in the soil and groundwater. Solution: Literature search on reported fate
77
-------
and transport models of the potential indicator parameters to determine their suitability
as indicators.
Selecting Chemical Tracers
Specific indicator parameters can help overcome some of the challenges faced when
developing a monitoring program. There are different types of chemical tracers and fingerprints
that have been recommended to monitor potential chemical migration in hydraulic fracturing.
Isotopes
Frac Fluid Constituents
Added Chemical Tracers
Frac Fluid Constituents as Unique Tracers
In Gastem's water quality monitoring program, frac fluid constituents were used as indicators
of potential migration. The process of identifying indicator parameters in potential
groundwater contamination is a complex problem because there are both natural and
anthropogenic sources of chemically similar contaminants. However, because of flowback aging
studies, it is possible to determine which constituents are anthropogenic and which are natural.
Table 4, summarizes the constituents that were selected based in their available concentrations
in the flowback fluids.
Table 4. Selected Flowback Constituents
Detected Constituents
(Highest concentration to Lowest)
Inorganic anions
Strontium
Lithium
Sodium
Calcium
Potassium
Organics
Toluene
Ethylene Glycol
Inorganic cations
Chlorides
Indicator
Potential
N
N
N
N
Y
N
N
Y
Source
Natural
Natural
Natural
Natural
Natural/Added
Natural
Added
Natural/Added
Notes
Biodegrades rapidly
78
-------
Results
Table 5. Comparison of Select Parameters from Utica Shale Stimulation Results
ANALYTES
Non X-VOC
SVOC
X-VOCs
Potassium
Chlorides
RESIDENTIAL
WATER WELLS
(before)
Non detect
Non detect
Non detect
0.27-0.97mg/L
0-69 mg/l
RESIDENTIAL
WATER WELLS
(after)
Non detect
Non detect
Non detect
0.23-0.98mg/L
1.1-84 mg/l
ADDITIVES IN
FLUID
Ethylene glycol*
35 mg/L
Acetone*
13 mg/L
Chloromethane*
8.0 mg/L
61.6 mg/L
5070 mg/l
*highest concentration parameter in the chemical group
Indicator parameters selected
Chlorides: Chloride's fate and transport has been extensively studied; as a conservative solute,
chloride is commonly used in groundwater transport and plume monitoring studies (barber,
1992). It is modeled using advection and dispersion. However, chlorides are not necessarily
unique to hydraulic fracturing. Chlorides are found in the formation brine solutions, and are
also used in the region for de-icing road surfaces in concentrations high enough to impact
groundwater (Godwin, 2003).
Potassium: Potassium has also served as an indicator in groundwater studies. It is highly soluble
and not likely to precipitate. It is found in appreciable concentrations. It is used in well drilling
activities as a clay stabilizer. If high levels of chloride are detected in the field in a water
resource, and if elevated potassium can be confirmed, then the probability increases that there
has been migration of frac fluid.
Table 6. Selected Indicator Parameters
Ease of detection in field
Ease of detection in laboratory
Non-reactive with water
Low Sorption
Availability in Frac Fluid
Background interference
K+
X
X
X
X
K-bentonite from the
Utica Shale
Cl
X
X
X
X
X
Road salt
Conclusion
The possibility exists for potential migration indicators to be drawn from the field of additives
used in hydraulic fracturing. The right choices can be a demonstrative and powerful tool.
However, deciphering which constituent(s) can serve as a sentinel tracers and which other
79
-------
constituents can serve as confirmation tracers requires meticulous data review. The process
includes understanding the chemical additives used, and the chemical processes that occur as a
result of mixing at the surface and downhole.
Study Limitations: (1) Impact of produced water is not included within the scope of the study,
since is generally understood that produced water does not contain significant concentrations
additives (Hayes, 2009), (2) water quality contribution of the introduced supply water, and (3)
site specific fate and transport information of organic constituents were not reviewed
intensively.
Future: A comprehensive study to determine a potential organic confirmation tracer.
References
Hayes, T. (2009b, December 31). Sampling and analysis of water streams associated with the
development of Marcellus Shale gas, final report. Canonsburg, PA: Marcellus Shale
Coalition, Gas Technology Institute.
Lesage, S., Xu, H., and Novakowski, K.S. 1997. Distinguishing natural hydrocarbons from
anthropogenic contamination in groundwater. J. Groundwater, 35(1), 149-160.
Clark, C.E. and J.A. Veil. 2009. Produced Water Volumes and Management Practices in the U.S.
USDOE/Argonne Laboratories Report No. ANL/EVS/R-09/1.
K. S. Godwin, S. D. Hafner, M. F. Buff., Long-term trends in sodium and chloride in the Mohawk
River, New York: the effect of fifty years of road-salt application Environmental
Pollution, Volume 124, Issue 2, July 2003, Pages 273-281
D.F. Dwyer and J.M. Tiedje, Degradation of ethylene glycol and polyethylene glycols by
methanogenic consortia, Appl. Environ. Microbiol. 46 (1983), Pages 189-190
80
-------
Elucidating Water Contamination by Fracturing Fluids and
Formation Waters from Gas Wells: Integrating Isotopic and
Geochemical Tracers
Avner Vengosh, Nathaniel Warner, Stephen Osborn, Rob Jackson
Nicholas School of the Environment, Duke University
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
This presentation highlights the use of several isotope-fingerprinting methods coupled with a
geochemical evaluation of the possible impacts of fracturing fluids and formation waters on the
quality of water resources in gas-drilled affected areas. The presentation is based on
preliminary data generated as part of Duke University research on the impact of gas drilling and
fracturing fluids on shallow groundwater resources in Pennsylvania and New York. The study
presented here is based on (1) systematic sampling of shallow groundwater from private wells
from the Catskill aquifer in northeastern Pennsylvania in an attempt to establish baseline water
quality data in areas that are expected to be affected by gas drilling and hydro- fracturing; (2)
systematic analysis of the chemical and isotopic compositions of brines from the Marcellus
Shale, one of the major target formations for unconventional gas exploration in the
Appalachian basin; and (3) integration of geochemical (major and trace elements), water
isotopes (5180, 52H), and dissolved salt isotopes (nB/10B, 87Sr/86Sr, 228Ra/226Ra) of co-produced
waters from a gas well in Pennsylvania. Boron and strontium isotopes were measured by
thermal ionization mass spectrometry (Triton mass spectrometer) at Duke University using
advanced methods with analytical reproducibility of 0.6%o and 0.01%o, respectively
(http://www.nicholas.duke.edu/tims/). Radium isotopes were measured at the Laboratory for
Environmental Analysis of RadioNuclides at Duke University
(http://www.nicholas.duke.edu/learn/). Trace metals were measured by VG PlasmaQuad-3
inductively coupled plasma mass-spectrometer (ICP-MS) at Duke University.
The Upper Devonian Catskill aquifer in northeastern Pennsylvania is the focus of the baseline
survey and is composed of various amounts of gray to red shale, siltstone, sandstone, and
conglomerate (Low and Galeone, 2007). The bedrock aquifer is underlying glacial deposits,
particularly in valleys, and overlying the Lock Haven Formation. Groundwater flow in the
Catskill aquifer is mainly through fractures. The investigated private wells had a reported depth
range 110 to 430 (median=270) feet.
The chemical data of the Catskill groundwater was evaluated based on preliminary sampling of
the Duke team combined with archive water-quality database from Pennsylvania Department
of Conservation and Environmental Resources (Taylor, 1984). The data show that the Catskill
groundwater has low salinity (TDS<400 mg/L) with a typical Ca-HC03 composition. All levels of
trace inorganic contaminants were below the EPA Maximum Contaminant level (MCL) drinking
81
-------
water guidelines (e.g, arsenic < 1.4 u,g/L, barium< 230 u,g/L). Preliminary 226Ra data also show
activities of 0.1 to 0.5 pCi/L, significantly lower than the MCL guideline of 5 pCi/L for combined
226Ra and 228Ra activities. The stable oxygen and hydrogen isotope ratios of the Catskill
groundwater were found to overlap the local meteoric isotopic composition, with 52H - 5180
relationships identical to the Local Meteoric Water Line (Kendall and Coplan, 2001). The
strontium isotope ratios show 87Sr/86Sr range of 0.71201 to 0.71553. Boron isotopes (expressed
as 5nB values, normalized to SRM951) of the Catskill groundwater have a range of 13.1 to
28.1%o.
In contrast, integration of existing data on the chemical and water isotope compositions of the
Marcellus Shale (Osborn and Mclntosh, 2010) and new measurements of strontium and boron
isotopes of the Marcellus brines and flowback water from a gas well in PA (Meadow #4,
provided by the USGS) show that formation/ fracturing fluids waters from gas wells in PA have
a significantly different composition relative to the local shallow groundwater in eastern PA. In
addition to the large difference in the salinity of the water (TDS of 200-300 to mg/L in shallow
groundwater relative to 250,000 mg/L in the Marcellus brines), the shallow groundwater has a
Ca-HC03 composition relative to a Ca-chloride composition the produced water from the gas
well. Given the multiple salinity sources (e.g., sewage and domestic waste water, road salt
deicing, coal mining drainage, leachates from coal combustion products), the salinity factor
cannot be a sole indicator for the contamination of groundwater by formation/ fracturing
waters. Likewise, in spite of the fact that the stable water isotopes of the Marcellus shale are
more enriched with 180 and 2H relative to the local groundwater with a significant lower 52H-
5180 slope, this parameter cannot be used alone, as other salinity sources, particularly those
that involved surface evaporation, would have similar isotopic composition.
A detailed analysis of the geochemical and isotopic variations of the flowback waters samples
from Meadows #4 well in PA represents a mixture of fracturing fluids and the original Marcellus
Shale brines shows that the concentrations of Ca, Ba, Sr, B, and Li, among others, increase in
flowback fluids during the first twenty days sampling following fracturing. Our data also show
that 87Sr/86Sr ratios increase from 0.71012 in the initial flow to 0.71122 on 20th flow day, as the
isotopic shift occurred already on the first day of flow. Likewise, the 5nB show a slight increase
from 28.3%o in the initial flow to 30.1%o in the twentieth flow day. The Sr and B isotope ratios
of the latest days are similar to those of the Marcellus Shale brines and thus indicating a
blending of two end-members (1) fracturing fluids with lower salinity and concentrations of
trace elements, lower 87Sr/86Sr ratios (<0.71012) and 5nB (<28%o) values; and (2) the Marcellus
Shale end-member with higher 87Sr/86Sr ratios (0.7115) and 5nB values (32-33%o). This flow
experiment also demonstrates that under operational conditions of continuous pumping, the
composition of the Marcellus brines is expected to predominant the produced waters from gas
wells.
Mixing modeling shows that the differences in both strontium concentrations and isotopic
ratios between the Marcellus Shale brines and/or the mixed Marcellus-fracturing fluids with the
local groundwater in eastern PA could be very useful in delineating even small contribution of
82
-------
formation water. Sensitivity tests show even a very small contribution of formation water
(1/100) could be identified in a system that is controlled by mixing of the Marcellus Shale brines
(87Sr/86Sr =0.7115) and/or a mixture of the Marcellus Shale and fracturing fluids (87Sr/86Sr
<0.71012). Consequently, strontium isotopes could be a very sensitive tracer to track the
possible contamination of shallow groundwater and surface water by formation/ fracturing
waters from gas wells. While the boron concentrations of the flowback waters (17,000 to
63,000 u,g/L) are higher by three orders of magnitude relative to the local groundwater (10 to
400 u,g/L), the 5nB values are only slightly higher (32-33%o in the Marcellus brines relative to
13-28%o in the Catskill groundwater). Nonetheless, this isotopic composition is different from
other potential salinity sources (e.g., coal ash and acid mine drainage with elevated B
concentrations and much lower 5nB values of -15 to 0%o) and thus should be used, in
conjunction with the other tracers as a critical tool to assess the possible interaction of
formation/ fracturing waters with shallow groundwater or surface waters.
Overall, our study indicates that a detailed survey of the baseline chemical and isotopic
compositions of shallow groundwater is crucial in evaluating possible contamination of
flowback fracturing fluids and co-produced formation waters from gas wells. Our preliminary
investigation of the Catskill formation groundwater from northeastern PA clearly shows
significant chemical and isotopic differences between the shallow groundwater and the
Marcellus brines and/or fracturing fluids mixed with the Marcellus brines. The ability to
delineate trace levels of formation / fracturing waters in very early stages of water
contamination depends on integration of multiple geochemical and isotopic tracers that could
provide reliable and sensitive detections of contamination by effluents associated with hydro-
fracturing and gas drilling.
References
Low, D.J., and Galeone, D.G., (2007) Reconnaissance of arsenic concentrations in ground water
from bedrock and unconsolidated aquifers in eight northern-tier counties of
Pennsylvania: U.S. Geological Survey Open-File Report 2006-1376, 35 p.
Osborn S.G., Mclntosh, J.C. (2010) Chemical and isotopic tracers of the contribution of microbial
gas in Devonian organic-rich shales and reservoir sandstones, northern Appalachian
Basin. Appl Geochem 25: 456-471.
Taylor I.E., (1984) Groundwater Resources of the Upper Susquehanna River Basin, Pennsylvania
(Pennsylvania Department of Environmental Resources-Office of Parks and Forestry -
Bureau of Topographic and Geologic Survey Water Resources Report 58).
Kendall, C. and Coplan, T. B., (2001) Distribution of oxygen-18 and deuterium in river waters
across the United States. Hydrological Processes 15:1363-1393.
83
-------
Summary and Abstracts from Theme 3: Field and Analytical
Challenges
84
-------
Summary of Presentations from Theme 3: Field and Analytical Challenges
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
Technical Presentations
The first technical presentation in this theme addressed sampling issues such as
representativeness, handling, and preservation.
Mark Engle, USGS, discussed the challenges of obtaining representative flowback fluid samples.
Because fluids will undergo physical, chemical, and/or biological changes as they are moved
from a geologic reservoir to the surface, sampling and preservation techniques affect the
results. Dr. Engle also described modeling results indicating that composition of fluid varies
non-linearly with flowback progress, which further emphasizes the need for time-series
sampling.
The second set of technical presentations addressed practical quantitation and method
reporting limits.
David Thai, Environmental Standards, Inc., discussed quality assurance (QA), quality control
(QC), and method performance considerations for HF-related sampling. Mr. Thai discussed key
quality parameters, including precision, accuracy, representativeness, completeness,
comparability, sensitivity, and selectivity. He described a series of QA/QC best practices that
can be applied to HF-related analyses and the interferences that could lead to QA/QC
challenges.
Keith McLeroy, Texas A&M University, described a method for total organic carbon (TOC)
determination utilizing a supercritical water oxidation (SCWO) procedure. The SCWO method
allows for the effective analysis of TOC in samples where concentrations of inorganic species
would usually inhibit oxidation.
Kesavalu M. Bagawandoss, Accutest Labs, discussed the challenges associated with practical
quantitation limits, method detection limits, interferences, and dilution with respect to HF
samples. Dr. Bagawandoss made a series of recommendations for dealing with highly viscous
samples and emphasized the importance of establishing expectations and predefining analytical
parameters.
The final set of technical presentations specifically addressed interference and dilution
challenges associated with radionuclides.
85
-------
Terry Romanko, TestAmerica Labs, described methods for radionuclide analysis (specifically
radium-226/228) and provided suggestions for addressing interferences in different situations.
He emphasized the importance of properly representing sample results.
Zoltan Szabo, USGS, discussed the need for high-precision radium analysis and described
analytical methods and approaches that could be used to overcome the challenges of analysis
in high brine matrices. He stressed the importance of yield monitoring in these analyses.
Summary of Discussions Following Theme 3: Field and Analytical
Challenges Presentations
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by meeting participants have not been verified or endorsed by EPA. Any mention of
trade names or commercial products does not constitute endorsement or recommendation for
use.
The availability and capabilities ofSCWO equipment. The equipment is available commercially,
and Mr. McLeroy has developed specific methodologies and protocols for this application. The
SCWO technique is the same as standard method 5310-D, though there is a difference in
procedure. Mr. McLeroy stated that the SCWO method allows a user to focus on the organic
carbons, which provide a food source for bacteria as well as act as potential tracers. Traditional
TOC analyzers can only handle dilute samples, according to Mr. McLeroy.
Sample preservation and holding times. One participant recommended sub-sampling at the
wellhead based on analyte; this way, samples can be treated individually based on their specific
needs. For example, the participant explained that some samples can be immediately frozen
while others can be filtered and preserved on site. A participant stated that holding times also
vary depending on the type of sample. Speed is important for volatiles and reactive species, and
these samples should be processed within 48 hours, according to this participant. However, the
participant noted that less time-sensitive samples can be held for several days, or up to several
weeks. Many analytes have prescribed holding times, and there is also guidance available from
different sources. Participants emphasized that planning for holding times is an important part
of the sampling and quality assurance planning process. Participants have found written plans
and instructions for field personnel to be essential. In addition, participants suggested that
operators and researchers should take into account the challenges of taking time-sensitive
samples while in remote field locations. Participants also noted the importance of recording
pressure checks and temperature for pressurized samples, as well as proper sample
preparation and extraction.
Sampling methods and challenges. Participants recommended using bombs (pressurized piston
cylinders) for sampling at the well. However, not all environmental labs analyze these types of
samplesparticipants recommended that a lab with a hydrocarbon division may be better
86
-------
equipped to address them. Decontamination of bomb samples can be an issue, according to
some participants, depending on the level of IDS. Use of Gas Processors Association (GPA)
method 2174-93 was recommended by one participant in order to preserve the samples under
well conditions. Participants noted that collection of samples utilizing this method at the
wellhead would also provide valuable information on the gases present in the sample stream.
Analysis methods and challenges. One participant noted that only a handful of labs have the
capability to do isotope labeling analysis. Participants also expressed concern about the lack of
reference standards for HF-related analyses. In addition, participants indicated that labs and
states do not often have standard protocols for dealing with flowback samples. One participant
mentioned additional methods for radionuclide analysis that are not yet EPA-certified, as well
as the availability of specialized ion exchange resins for radium. The geothermal industry may
be a source of relevant information on this topic, according to a participant.
Cost-control strategies. Participants stated that good process knowledge and screening
processes can reduce cost of analysis. Participants also suggested that reducing the sample load
and pooling method development efforts within the industry may be possible. A participant
noted that some of the specialized protocols described in the presentations, such as TOC
analysis, only need to be carried out on the "rich samples" that are high in salt and
interferences, and analysis may not be necessary for every sample that is analyzed. One
participant suggested using university labs instead of commercial labs, though other
participants noted that results from academic labs may not be defendable in court. Many
participants stated that academic labs are essential for method development, evaluation of
procedures and methodologies, and analyses of specialized tracers.
Sources of chemical species and the disposal of solid waste. One participant suggested that
barium and radium could be introduced into the HF waste stream through the drilling mud, and
also raised the possibility of temperature-dependent precipitates could be removing barium
and radium from samples that are taken after precipitation occurs. A participant noted that
uranium is generally not a concern in the Marcellus Shale. Participants added that radioactivity
does build up in treatment sludge, as dispersed radioactivity is concentrated into a smaller,
solid form. A participant confirmed that well cuttings are sent to landfills. Radiation from
reused materials may be a concern, according to some participants.
Fluid disposal and injection wells. A participant stated that a particular operator only uses tanks,
not pits, for storing produced water. Another participant noted that a different company uses
double-lined pits with leak detection capabilities to hold produced water, and the pits are
designed, constructed, and overseen by a third-party engineering firm. The participant
continued to explain that the pits are drained and inspected on an annual basis. There are very
few Underground Injection Control (UIC) wells that accept brine from oil and gas production
(Class II injection wells) in the Marcellus Shale and only eight in Pennsylvania.
Produced water in the Marcellus Shale and other formations. In most shale plays, participants
noted that the majority of produced water is produced in the first months after HF. In the
87
-------
Marcellus Shale, participants stated that approximately 15-20% of the injected volume is
returned fairly quickly and then decreases to 1-3 barrels per million cubic feet of gas. This
varies slightly geographically, according to participants: the Marcellus in West Virginia is
somewhat wetter compared to northern Pennsylvania, though the Marcellus tends to be the
driest U.S. shale play. The Haynesville returns 15-30% of injected water quickly, the Fayetteville
returns about 30% quickly, and the Barnett, which is very wet, returns about 50-100%,
according to participants. One participant noted that he has never seen a Marcellus Shale well
that has returned the volume of water that had been injected. Other participants noted that
Marcellus Shale wells do return some water, and the composition appears to represent
formation water. Some participants explained that the salinity in produced water samples
appears to be from paleo-evaporated seawater brines; this chemical signature is distinct from
the fracture fluids and from salt dissolution.
88
-------
Abstracts for Theme 3: Field and Analytical Challenges
Abstracts were submitted to U.S. EPA by the presenters for use in this proceedings document.
Not all presenters submitted abstracts of their presentations.
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
89
-------
Quality Assurance, Quality Control and Method Performance
Considerations for Chemical Testing of Environmental Samples
Impacted by Hydraulic Fracturing Fluids
David Thai
Environmental Standards Inc.
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
A discussion of the traditional quality assurance and quality control (QA/QC) as practiced in
environmental studies directed by the USEPA is presented, focused on the most common
analyses provided for samples of hydraulic fracturing (HF) fluids, flowback waters and related
samples. The types of analyses being applied most commonly are presented, based on
interviews and correspondence with five commercial laboratory companies. The specific types
of QA/QC studies and samples applied in a robust quality assurance program are presented for
each type of analysis. A survey of interferences is presented, based on hundreds of samples of
HF fluids and flowback waters, along with a cautionary summary, alerting the data user to the
need for review of HF data by senior analytical chemists.
The USEPA requires that a systematic planning process be used to plan for data quality in
environmental data operations that receive funding from the agency. To support this
requirement, EPA developed a process called the Data Quality Objectives (DQO) Process. The
DQO Process is a reasonable starting point to build a model for data quality planning in general.
It identifies the data quality indicators (DQIs) that need to be measured to help assure that data
of known quality will be obtained. The DQO process and DQIs are described in the publication
Guidance for the Data Quality Objectives Process (QA/G-4).1 Other agencies have also
developed and have collaborated on guidance that define essential data quality parameters for
Department of Defense and Federal Facilities in general.2'3 A review of these documents reveals
very close agreement between them on the list of DQI's.
The USEPA also provides instructions and specific requirements for the information needed in a
QA Project Plan (QAPP) for environmental data operations funded by EPA. This document is
titled EPA Requirements for Quality Assurance Project Plans (QA/R-5).4 Conducting the DQO
Process and developing a QAPP that covers all the requirements in the R-5 document is a very
good foundation for quality planning. However, experience has shown that the traditional
approaches can fall short in providing data for difficult matrices and cases in which the need for
high confidence at low detection limits affects decision making. It is important to understand
how the properties and composition of hydraulic fracturing fluids and flowback waters can
impact data quality.
90
-------
The G-4 lists principal DQIs as precision, bias, representativeness, completeness, comparability,
and sensitivity. (PARCCS). In many derivative publications, and guidance documents these are
referred to as the "PARCC" parameters, leaving sensitivity out of the mnemonic, and
unfortunately, out of the data quality planning process. Additionally, the sensitivity of a method
as practiced in laboratories is usually demonstrated on a reference matrix, with no chemical
interferences present. In reality, interferences often result in dilutions, or reduced sample
volumes being analyzed in the laboratory. In many (but by no means all) cases, the EPA
methods themselves direct the use of dilutions and reduced sample volumes to overcome the
impacts of matrix-borne interferences. Here again, and understanding of the types of
interferences likely to be encountered.
A survey of commercial laboratories (5 companies) was conducted to determine the commonly
applied testing technologies for samples impacted by or suspected of containing hydraulic
fracturing fluids or flow back waters. The following types of testing method were common to
almost all responses. The testing technologies applied were:
Classical Chemistry, often referred to as wet chemistry. This general category of testing usually
involves wet chemical techniques and physical characterization, and may rely on more than one
type of determinative technique (e.g., titrimetry, spectrometry, gravimetry, ion selective
electrodes, ion chromatography, voltatmetry). Examples typical of classical chemistry
techniques used fracturing fluid and flow back water studies include: Alkalinity, Ammonia, BTU
Content, Carbonate, Chloride, Methylene Blue Active Substances, Oil and Grease, Potassium,
Sodium, Specific Conductance, Sulfate, Sulfide, Total Dissolved Solids, Total Suspended Solids.
Most interferences on anionic evaluations come from competing anions.
Gas Chromatography of Volatile Organic Compounds and Gasoline Range Organics, (GC-VOC
&GRO). This type of testing involves the collection of samples which will be purged with
nitrogen in the laboratory to bring volatile components into a gas chromatograph (GC). The GC
which separates the components by physicochemical properties (polarity, molecular weight,
electrostatic interactions), and ports them into a detector such as a flame ionization detector or
mass spectrometer. Examples of HF analytes measured by this type of test include: 2-
butoxyethanol, gasoline range organics, methane, ethane. Most interferences are from non
target volatile organic compounds. Cross contamination can be a significant issue due the
mobility of the analytes.
Gas Chromatography of Semivolatile Organic Compounds, Diesel and Oil Range Organics,GC-
SVOC, DRO, ORO). This type of testing involves the collection of samples which will be extracted
with solvents in the laboratory to bring semivolatile components into a gas chromatograph
(GC). The GC which separates the components by physicochemical properties (polarity,
molecular weight, electrostatic interactions), and ports them into a detector such as a flame
ionization detector or mass spectrometer. Examples of HF analytes measured by this type of
test include: Diesel Range Organics, Oil Range Organics, Ethylene Glycol, Formaldehyde and
many other analytes. Most interferences are from non-target organic compounds including
fats, oils, organic acids and sulfur in the extracts.
91
-------
Inductively Coupled Plasma -Atomic Emission Spectrometry. (ICP-AES) The sample is typically
digested with acids under heated conditions and the liquid digestate is aspirated into an
inductively coupled plasma. Light focused on the excited metals further excites the electon
states, and they emit spectral light that is passed through a gradient filter and detected by
wavelength and intensity. Examples of analytes that are detected by this technique include.
Calcium, Barium, Iron, Potassium and Sodium and many others.
Inductively Coupled Plasma - Mass Spectrometry (ICP-MS) The sample is typically digested
with acids under heated conditions and the liquid digestate is aspirated into an inductively
coupled plasma. The analytes are then passed into a quadupole mass analyzer, where they are
separated by their mass-to-charge ratio. Some instruments pass the analytes through a collision
cell prior to the quadrupole to reduce polyatomic interferences. The analytes then strike an
electron multiplier and are detected by the current generated. Examples of analytes of interest
in HF studies that are detected by this technique include: aluminum, arsenic, barium, cadmium,
chromium, cobalt, copper, iron, lead, magnesium, manganese, nickel, potassium selenium,
silver, sodium, zinc.
Naturally Occuring Radioactive Materials. (NORM). Several radiochemical methods can be
applied to this general field of testing. In each case the fundamental analytical technique
involves carefully arranging the sample into a reproducible geometry and measuring the
radioactive emissions using gross alpha, gross beta or gamma Spectrometry when information
on specific radionuclides is needed. Examples of nuclides of interest would include radium 226,
radium 228.
Specialized Gas Chromatography or Liquid Chromatography- Mass Spectrometry. (GC/LC-MS
SVOC). Liquid Chromatography is well suited to highly polar, or ionic compounds, such as some
phenols, glycols and organic acids. Specialized GC-MS can be used for alkylphenols, and some
glycols and alcohols. For analytes that are easily fragmented, or in need of highly specific
analyses, specialized mass Spectrometry, such as ms/ms, msn, negative ion chemical ionization
and high resolution mass Spectrometry may be of use. The suite of QA/QC measures for these
tests should be the same as those indicated for routine GC/MS, as long as the methods of
sample preparation are analogous.
The commonly practiced types of QA Study or QC sample are defined and briefly described
below.
Initial demonstration of proficiency or capability. (IDOC) This QA measure is typically
comprised of a series of tests on positive controls (spiked reference samples) analyzed as a part
of staff training or method capability testing. Precision and percent recovery are evaluated. This
may be set up as four consecutive, or four simultaneously prepared samples.
Laboratory control sample (LCS). The LCS is a sample comprised of a reference matrix spiked
with analytes of concern and taken through all steps of preparation and analysis within the
92
-------
laboratory. Recovery is compared to control limits established in the method, QAPP, or
laboratory (based on historical data). It is prepared and analyzed with each batch of (20 or
fewer) samples, and may be prepared in duplicate (LCSD) if insufficient sample material is
available to perform the matrix spike/matrix spike duplicate or sample duplicate.
Matrix Spike (MS). An aliquot of the field sample is analyzed as received, and another aliquot is
analyzed with a known quantity of analyte added prior to sample preparation and analysis. This
measures recovery in the presence of matrix interferences.
Matrix Spike Duplicate (MSD). A separate aliquot of the field sample used for the MS is spiked
and prepared in the same manner as the MS. The results are compared and are measures of
both precision and recovery in the presence of the sample matrix.
For metals analyses, If the MS/MSD recoveries are unacceptable, the same sample from which
the
MS/MSD aliquots were prepared should also be spiked with a post digestion spike.
Post digestion spike. For metals analyses, this step involves adding the metals directly to the
prepared sample to determine whether the recovery problem is related to the original matrix,
or the analytical technique.
Unspiked Duplicate, also termed Laboratory Duplicate. This is useful when the parameter
being studies is always present, or when the parameter is difficult to add. It is simply a separate
aliquot of sample analyzed alongside the original, and compared to the original sample as a
measure of precision.
Dilution test. For metals analyses. If the analyte concentration is sufficiently high (minimally, a
factor of 10 above the lower limit of quantitation after dilution), an analysis of a 1:5 dilution
should agree within ±
10% of the original determination. If not, then a chemical or physical interference effect should
be suspected.
Method Blank (Sample Preparation Blank). A reference matrix is taken through all the steps of
sample preparation and analysis, alongside the field samples, to determine whether
contamination may be present in the laboratory environment or materials. This is a measure
that helps evaluate the accuracy of the data, particularly at low levels.
Instrument Blank. A pure aliquot of the sample carrier (e.g.,gas, aqueous acid, solvents
depending on the type of method) is introduced to the instrument in the same manner as a
sample. This checks for sample-to-sample carryover potential in the analysis.
Trip Blank (Volatile Organic Compounds). A sample of analyte-free water is placed in a sample
shipment cooler and travels with the sampling materials to the sampling location, and then
travels with the samples back to the laboratory. This sample is analyzed to identify the potential
for contamination of samples or sample bottles in the field or under shipping conditions.
93
-------
The types of QA/QC that may reasonably be expected and required in a robust QA/QC program
are presented in Table 7.
Quality Check
Initial
Demonstration of
Capability
Method Detection
Limit Studies
Standard Reference
Materials (NIST
traceable if
available.)
Blind PE Samples
Trip Blanks (VOCs)
Equipment Blanks
Instrument Blanks
Field Duplicates
Method Blanks
Laboratory Control
Samples
Matrix Spikes/
Matrix Spike
Duplicates
Laboratory
Duplicates
Dilution Check
Post Dilution Spike
Classical
Chemistry
Yes
Most
methods
Yes
Yes
Rare
Some
Yes
Yes
Yes
Some
Yes
GC-VOC
(&GRO)
&
GC/MS-
VOC
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
GC-SVOC
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
ICP-AES
/ICP-MS
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
NORM
Yes
Note 1
Yes
Yes
Yes
Rare
Yes
Yes
Yes
Yes
GC/LC-MS
SVOC
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Notes
1. For NORM Minimum detectable activities are calculated using instrument backgrounds
and sample characteristics.
94
-------
Matrix Interferences Typical of HF Fluids, Flowback Waters and Related
Samples
A series of chemical testing results from hydraulic fracturing fluid and flowback waters (within
the first 5 days post-fracturing) from various Marcellus sites was examined for potential matrix-
borne interferences. The results were provided by a commercial laboratory providing analytical
services to many sites across the Marcellus formation. (TestAmerica Nashville). Ranges of
observed levels of interferences were also taken from literature sources.5'6 Known
interferences for the test methods are described below, along with an evaluation of the
potential impact from matrix-borne interferences in this data set. Many of the known
interferences are described verbatim as they appear in the USEPA Methods being applied.
Total Dissolved Solids (Gravimetric Method 160.1). Highly mineralized waters containing
significant concentrations of calcium, magnesium, chloride and/or sulfate may be hygroscopic
and will require prolonged drying, desiccation and rapid weighing. Samples containing high
concentrations of bicarbonate will require careful and possibly prolonged drying at 180°C to
insure that all the bicarbonate is converted to carbonate. Too much residue in the evaporating
dish will crust over and entrap water that will not be driven off during drying. Total residue
should be limited to about 200 mg. (On a 100 ml sample this would be 2000 mg/L, 3.4% of the
sample
Alkalinity (Titrimetric Method 310.1): Oil and grease may coat the electrode causing sluggish
response for both alkalinity and pH. Salts of weak organic and inorganic acids may cause
interference in the electrometric pH measurements. For samples having high concentrations of
mineral acids, such as mine wastes and associated receiving waters, method modifications are
needed (i.e., the laboratory is to titrate to an electrometric endpoint of pH 3.9,using ASTM
Standard, D- 067, Method D or superseding methods).
Anions (Chloride, Fluoride, Bromide, Nitrite, Nitrate, Phosphate, Sulfate): Any species with a
retention time similar to that of the desired anion will interfere. Large quantities of ions eluting
close to the anion of interest will also result in an interference. For example, high levels of
organic acids that may interfere with inorganic anion analysis may be present. Two common
species, formate and acetate, elute between fluoride and chloride. (Several organic acid
homopolymers, polymers and telomers are used in fracturing fluids.
Total Kjeldahl Nitrogen (Methods 350.1,2,3,4 -Automated Colorimetry, Titrimetry,
Potentiometry, Ion Selective Electrode): High nitrate concentrations (10X or more than the
TKN level) result in low TKN values. If interference is suspected, samples should be diluted and
reanalyzed. The reaction between nitrate and ammonia can be prevented by the use of an
anion exchange resin (chloride form) to remove the nitrate prior to the TKN analysis.
Barium (ICP-AED or MS, e.g. 200.7, 6010, 6020): The total recoverable sample digestion
procedure will solubilize and hold in solution only minimal concentrations of barium in the
presence of free sulfate. For the analysis of barium in samples having varying and unknown
95
-------
concentrations of sulfate, analysis should be completed as soon as possible after sample
preparation.
Metals by ICP-AED (e.g., 200.7, 6010): Flow variation due to physical interferences with the
sample nebulization and transport processes can results from changes in viscosity and surface
tension. This in turn can result in significant inaccuracies, especially in samples containing high
dissolved solids or high acid concentrations. The use of a high-solids nebulizer or sample
dilution is often necessary. Another problem that can occur with high dissolved solids is salt
buildup at the tip of the nebulizer, which affects aerosol flow rate and causes instrumental drift.
Here again, use of a high-solids nebulizer, wetting the argon prior to nebulization, using a tip
washer, or diluting the sample may help alleviate the difficulty. Also, it has been reported that
better control of the argon flow rates, especially for the nebulizer, improves instrument
stability and precision; this is accomplished with the use of mass flow controllers.
Chemical interferences include molecular-compound formation, ionization effects, and solute-
vaporization effects. The QA/QC program should provide for matrix matching, and standard-
addition procedures. Chemical interferences are highly dependent on matrix type and the
specific analyte element. The Spectral Interference Check addresses most interferences at 50
mg/L. ( 300 mg/L Fe; 200 mg/L Al: 50 mg/L Ba; 50 mg/L Be; 50 mg/L Cd; 50 mg/L Ce; 50 mg/L
Co; 50 mg/L Cr; (i) 50 mg/L Cu; (j) 50 mg/L Mn; (k) 50 mg/L Mo; (I) 50 mg/L Ni; (m) 50 mg/L Sn;
(n) 50 mg/L Si02; 50 mg/L Ti; 50 mg/L Tl and 50 mg/L
Of all these limits, barium was exceeded for one sample.
For ICP-MS (6020A)
The interference check standards for this method check for interferences at the following
levels: Ca (300 mg/L); Fe (250 mg/L); Mg (100.0 mg/L); Na (250 mg/L) 250.0; K (100 mg/L); Cl
(2000 mg/L).
Of these, 10 percent of the samples exceeded the Na check standard level.
Four percent (4%) exceeded the K check standard level.
One percent (1.3%) of the 381 samples exceeded the Mg check standard level. Higher
values reported in the literature exceed 25 times the interference check standard level.
One of 74 samples exceeded the Fe check standard level.
Two percent (1.8%) of the 542 samples exceeded the Cl check standard level.
Three percent ( 3.0%) of the 439 samples exceeded the Ca check standard level. Higher
literature values exceed the check standard concentration 100-fold.
Both metals methods note that dissolved solids should not exceed 0.2 % (True for 4% of 462
samples.) Literature values range up to 2.5%.
Finally, memory interferences result when analytes in a previous sample contribute to the
signals measured in a new sample. Memory effects can result from sample deposition on the
uptake tubing to the nebulizer, and from the buildup of sample material in the plasma torch
and spray chamber.
96
-------
Summary
Many standard analytical methods apply to the analysis of HF fluids and flowback water
samples. However, they will perform poorly in some cases involving high levels of interferents.
The interferences are well characterized, but a careful evaluation of the entire data set is
needed to understand the potential biases.
For Metals, ICP-MS method alone is not a good fit for this matrix. ICP-Atomic Emission
Spectrometry is needed to help overcome the many interferences well above the interference
check standard.
Total Dissolved Solids data should be scrutinized for potential impacts on the metals data, and
classical chemistry parameters.
Field duplicates should be used and evaluated carefully. If the anion data set does not make
sense, consider using the method of standard additions.
Watch barium levels closely. If they are not consistent, or if they are high, engage a senior
analytical chemist to help evaluate the data set.
It is important to evaluate the sulfate levels.
References
USEPA. Guidance on Systematic Planning Using the Data Quality Objectives Process. EPA QA/G-
4. EPA/240/B-06/001. Office of Environmental Information Washington, DC. February,
2006.
Intergovernmental Data Quality Task Force Uniform Federal Policy for Quality Assurance Project
Plans, Evaluating, Assessing, and Documenting Environmental Data Collection and Use
Programs Part 1: UFP-QAPP Manual
http://www.epa.gov/fedfac/documents/qualityassurance.htm.
DoD Quality Systems Manual for Environmental Laboratories Version 4.2. Based on NELAC
Voted Revision 5 June 2003. Issued 10/25/2010. http://www.navvlabs.navy.mil/
USEPA. EPA Requirements for Quality Assurance Project Plans. EPA QA/R-5. EPA/240/B-
01/003. Office of Environmental Information Washington, DC, March 2001.
Kirby, Carl S., Pritz, Molly E., Lunde, A. Scott, and Tate, Garyn P., Inorganic Geochemistry of
Marcellus Shale Natural Gas Hydrofracturing Waters. University, Department of
Geology, Bucknell University, Lewisburg PA.
Blausch, M.E., Meyers, R.R., Moore, T.R., Lipinski, B.A.,Marcellus Shale Post-Frac Flowback
Waters - Where is All the Salt Coming From? Proceedings of the Society of Petroleum
Engineers Eastern Regional Meeting, Sept 2009.
97
-------
Determination of Total Organic Carbons in Difficult Sample
Matrices Utilizing the Supercritical Water-Oxidation TOC
Procedure
L. Keith McLeroy
Texas Engineering Extension Service, Texas A&M University
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Abstract
The requirement for monitoring Total Organic Carbon (TOC) in oil/gas produced waters is
inevitable. Most often, organics in these types of waters are volatiles, petroleum distillates, and
natural organic acids or purposely added organic chemical compounds. These are generally
difficult to analyze due to the presence of several interfering factors such as high turbidity,
increased alkalinity, drilling mud additives and very high concentrations of various
salts/minerals. Additionally, inorganic compounds in the waters (i.e.: brines, alkalinity) can give
false high readings in traditional total organic analyses if not accounted for. The methodology
of Supercritical Water Oxidation (SCWO) for monitoring TOC in oil/gas produced waters is being
utilized and evaluated. The SCWO method has shown to be a robust procedure for analyzing
TOC in otherwise difficult water matrices.
Introduction
An inherent issue with TOC analyses of oil/gas production water is the problem with difficult
sample matrices. Produced waters contain a variety of organic and inorganic compounds which
may be capable of exacerbating the chemistry mechanisms of analytical testing. Additionally,
environmental monitoring of oil and gas produced wastewaters is dependent upon protocols
that are robust enough to be of reliability. The work presented here is the evaluation of the
modified wet-oxidation method described as Supercritical Water Oxidation (SCWO) for the
analysis of TOC in very difficult oil/gas produced water sample matrices.
TOC Methodologies
TOC analysis is carried out through a choice of methodologies. Given that all TOC analyzers only
actually measure total carbon, TOC analysis always requires some accounting for the 1C that is
always present. One analysis technique involves a two-stage process commonly referred to as
TC-IC. It measures the amount of 1C evolved from an acidified aliquot of a sample and also the
amount of TC present in the sample. TOC is calculated by subtraction of the 1C value from the
TC the sample. Another variant employs acidification of the sample to evolve carbon dioxide
and measuring it as 1C, then oxidizing and measuring the remaining non-purgeable organic
carbon (NPOC). This is called TIC-NPOC analysis.
98
-------
A more common method directly measures TOC in the sample by again acidifying the sample to
a pH value of two or less to release the 1C gas to the atmosphere. Any remaining NPOC-C02
dissolved in the liquid aliquot is then oxidized releasing the gases. These gases are then
transmitted to the detector for quantification.
TOC analysis does have intrinsic problems with difficult sample matrices. Turbidity, salinity, oils
and other high mineral content samples can be problematic to adjust for and calibrate for in
most traditional TOC methodologies. Concerns are environmentally controlled waters and
oil/gas production wastewaters may not be monitored forTOCs' due to those issues.
If laboratories do have the capability to analyze difficult samples, they generally require more
frequent maintenance, and weekly or daily recalibrations. To eliminate those issues, the SCWO
was developed by GE Analytical Instruments (Boulder, CO.) to provide complete oxidation of
organic carbons in complex sample matrices.
Supercritical Water Oxidation
By convention, there are three phases of matter; solid, liquid, and gas. Science has since shown
that there are fifteen (15) states of matter, including super critical. When describing
supercritical water, it is fundamental to understand the different phases of water. When liquid
water is sitting in an open container at room temperature and ambient atmospheric pressure
and then cooled to below 0 °C, the water transitions from liquid phase to solid phase. If the
temperature of the liquid water in the same open container is raised above 100 °C, the water
boils and transitions from liquid phase to gas phase. This is normal behavior when atmospheric
pressure is kept constant at ambient conditions. At the same weight, or mass, solid water and
liquid water occupy about the same volume at atmospheric pressure. The equivalent weight of
gaseous water occupies about 1,600 times the volume at atmospheric pressure. However, if a
gas-tight lid was attached on a container and the temperature rose to 130 °C, the liquid water
would transition into the gas phase. Given that the volume needed for expansion is limited, the
pressure in the headspace of the sealed container would begin to increase as more gas is
formed. As shown in the phase diagram of water (Figure 1), an increase in head pressure raises
the temperature at which all the liquid water would transition into the gas phase. As a result,
the liquid water can be heated to a higher temperature without boiling. In this example, the
resulting pressure inside the container increases to nearly twice atmospheric pressure.
At a specific point, pressure can no longer be increased to maintain the liquid phase. Beyond
374 °C and 218 atm (3200 psi), the gas and liquid phases merge to form another phase of
matter. This phase is called Supercritical Water (SCW). When in a supercritical state, water
exhibits the characteristics and benefits of both a liquid and a gas. The SCW has a density closer
to that of a liquid, but can still diffuse similar to a gas. Organic material and gases become
highly soluble in SCW and, conversely, inorganic salts become insoluble. These conditions are
ideal for SCWO reaction. The GE InnovOx (InO) TOC Analyzer employed in the study uses the
wet chemical oxidation technique. This process seeds the solution with an oxygen donating
reagent. The InO employs a 30% weight/volume solution of sodium persulphate as the oxidizer.
It then heats the sample and oxidizer in a sealed reactor past the critical point and SCWO is
99
-------
achieved. The increased pressure within the reaction cell noticeably enhanced the efficiency of
the oxidation process, thus offering better recovery of C02 from difficult sample matrices. The
C02 is then analyzed through a highly stable, non-dispersive infrared (NDIR) detector.
Noted in the current evaluation was the difference from the traditional Standard Methods for
The Examination of Water and Wastewater (SM) 5310-D. Wet-Oxidation Combustion
procedure, was that the SCWO process completely removed all by-products from the sample
flow path between sample runs. This allowed for no carry over contamination from sample to
sample. Analysis data from organic-free water blanks between sample runs, demonstrated this.
Additionally noted was that as discussed previously, when water reaches a supercritical state,
organic material and gases become highly soluble in SCW, while inorganic salts become
insoluble. This concept is of importance, since salts will typically scavenge the oxidizer, resulting
in an incomplete organic C to C02 conversion. This was of significance when analyzing the oil
and gas produced wastewaters which were excessive in chlorides.
Evaluation of SCWO
The evaluation of the SCWO method employed the GE Analytical InO TOC Analyzer.
Water samples described in Table 8, were obtained from various sources that represented a
spectrum of industries that may make use of the TOC results for organic loading monitoring
into the environment, experimental TOC removal technology in the oil/gas industry, frac-water
studies and raw water characteristics.
For this assessment, oil/gas produced and source waters that were to be treated by various
membrane or chemical treatment systems were collected. The Global Petroleum Research
Institute (GPRI) and the Membrane Separations Laboratory (MSL) at Texas A&M University
(TAMU), College Station, Texas provided the samples and treatment technology. Samples were
analyzed at The Texas Engineering Extension Service (TEEX) Water and Environmental
Laboratory, College Station, Texas.
100
-------
Table 8. Samples evaluated for SCWO-TOC
Sample Type
Brazos River
Fractured Pond Waste
Brine Pond
Glycol Condensate
Mixed Oil/Gas Wastewater
Industry
Source Water
Gas Tracing Drilling
Crude Oil Production
Natural Gas Drying
Environmental Monitoring
Table 9. Results of SCWO-TOC analyses
Raw Sample
Brazos
Frac-Pond
Gly. Cond.
Mixed WW
Brine Pond
TOC PPM
180.3
83.2
44.3
91.0
15.3
Dup. PPM
181.4
82.5
39.5
92.7
17.3
Blank PPM
0.32
1.90
0.14
0.05
0.11
%RSD
0.30
.42
5.72
0.92
6.13
Cal. Range
1000.0 PPM
1000.0 PPM
1000.0 PPM
10000.0 PPM
1000.0 PPM
Post Treated
Brazos
Frac-Pond
Gly. Cond.
Mixed WW
Brine Pond
TOC PPM
15.7
22.3
40.0
15.7
16.0
Dup. PPM
21.0
19.2
33.1
19.2
19.1
Blank PPM
0.09
0.12
0.11
0.01
0.94
%RSD
14.4
7.4
9.58
10.0
8.83
Cal. Range
1000.0 PPM
1000.0 PPM
1000.0 PPM
10000.0 PPM
1000.0 PPM
Raw Sample
Brazos
Frac-Pond
Gly. Cond
Mixed WW
Brine Pond
Concentrate
Frac-Pond
Brine Pond
Chloride PPM
33.4
25,492
10,041
125
31,202
TOC PPM
59.3
2.44
Alkalinity PPM
162.9
427.0
34.7
79.1
519.0
Dup. PPM
58.2
2.42
Sulfate
PPM
3.1
67.8
273.8
157.4
6.81
Blank PPM
0.32
1.01
pH
S.U.
7.44
8.01
9.05
8.02
7.73
%RSD
0.94
0.41
Turbidity
NTU
775
360
12
152
88
Cal. Range
1000.0 PPM
1000.0 PPM
Observations
The InO and SCWO had performed well above expectations for this Phase I evaluation. The
Brazos River sample containing the heavy silt was duplicated within a RSD of 0.30%. This was
unexpected due to the high levels of solids. The brine pond water required no dilution, filtration
or adjustments to the persulfate feed on the InO and primary standards ran after the analysis
showed very good accuracy and no carry-over of contamination from the samples previously
analyzed. Repeat samples of the frac-water had a precision of 7.4 %RSD, despite having a heavy
101
-------
black colored turbidity and strong sulfide odor. The glycol condensate water was clear but high
in chloride. Organic-free water blanks ran between each sample repeat batch showed that the
InO rinse cycles performed efficiently and no carry-over had occurred. The membrane
concentrate wastes contained a heavy brine level and effectively, the SCWO method was able
to compensate for those levels and still maintain RSDs' of 0.94 and 0.41 respectively.
The InO can determine Total Inorganic Carbon (TIC), NPOC, and TOC by difference (TC-IC). This
allows for any combination of these modes in a single sample run or multiple duplicates on the
same sample.
Conclusions
The Phase I evaluation of the SCWO-TOC Method had proved to be a robust and accurate
method on otherwise difficult sample matrices. The SCWO methodology had demonstrated
that at specific calibration levels, readings were analyzed without needed dilutions. Despite
heavy solids loading and color in some samples, the InO was able to analyze the samples and
provide proper line flushing and rinse cycles to prevent carry over. The analyst friendly software
guided the technician through the step-by-step sample set-up and provided printer friendly
reports. Calibration curves were not difficult to develop and the certified primary standards
were delivered in a ready to analyze kit and there was no need to make-up laboratory
standards. The portability of the analyzer was also a convenience as some samples were
analyzed in different laboratory locations and outdoors during field trials. The InO required no
carry gas cylinders making it an ideal analyzer for field analysis.
The SCWO methodology demonstrated to be a very robust method and had the ability to
handle very difficult matrices and still provide accurate and precise results.
102
-------
Practical Quantitation, Method Detection Limits, Interferences
and Dilution Challenges
Kesavalu M. Bagawandoss, Ph.D., J.D.
Accutest Laboratories
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
Introduction
The matrices of fracturing fluids vary significantly in consistency and viscosity prior to
fracturing. Hence, the analytical challenges increase with the viscosity of the samples provided
for analyses. The samples on their own merits are generally representative. The flow back
water samples are representative and uniform in nature unless the amount of sediment
collected is a significant quantity.
Samples should be collected before fracturing and after fracturing to understand the
contaminants present before fracturing and after fracturing. This would yield valuable
information in terms of the transformations occurring during the fracturing process and
comingling of components from the host materials or the extracted materials. Another
suggested step would be collection of a sample of the host material if available or possible, to
assess the material balance or source of components detected. One recommendation is that
gas samples be collected at the point of exposure as the flow back water is discharged to
determine if gaseous components are being emitted and account for any emissions. A
suggested method for determining gaseous components in backflow water samples flowing
from the wells would be the use of GPA 2174-93. (GPA=Gas Processors Association)(Constant
pressure sampling method) This method allows for the samples to be captured at the same
conditions as the stream of water.
Sample Preservation
Preservation of the samples of fracturing fluids for the various analytical processes does not
pose any issues. Preservation of the Backflow waters generally does not pose issues. Volatile
components in both the fracturing fluids and the backflow waters need not be preserved with
acid and analyzed within 7 days of collection of samples thus satisfying the regulatory
guidelines on holding times prescribed in the various methodologies.
Practical Quantitation Limits (PQLs) and Method Detection Limits (MDLs)
Methods employed for the analyses of the fracturing fluids include US EPA 500 and 600 series
methods and Standard Methods for Water and Wastewater in combination with SW846
methods. The typical methods utilized are listed in Table 10. The components of the analytical
suite are presented in Table 11 through Table 14. The practical quantitation limits are also listed
103
-------
in the aforementioned tables. However, depending on the viscosity of the frac fluids and the
interferences present in the various samples, the PQLs may be affected. The MDLs utilized are
those determined for liquid samples. The MDLs are statistically derived, as outlined in the
method requirements for each of the analytes.
Interferences and Dilution Challenges
The PQLs are affected when dilutions are performed due to the matrix interferences. Dilutions
are generally performed to alleviate the viscosity issues or the interferences. Based on analyses
performed, flow back waters do not pose the same challenges as the fracturing fluid matrices.
Matrix interferences, specially, with respect to samples which are highly viscous are handled by
dilutions. Surrogates and internal standards and its behavior in the matrix vary based on the
samples. In some cases matrix spike compounds behave differently. For example, in a VGA
analysis the gases may recover fine yet the rest of the compounds may not recover at the same
level as the gaseous components. In a semi volatile analysis the acid surrogates may not
recover where as in the same sample the base surrogate recoveries may be adequately
recovered. Based on our experience, approximately 10 % of the fracturing fluid matrices pose
analytical challenges for the parameters analyzed.
Table 10. Typical Methods
Acrylamide-SW846-8316
Metals-Method 200.8/245.1
Cyanide/ Weak and dissociable, Total - SM 4500 CN E & I
Fecal and Total Coliform - SM 9222D & B
Total Phenolics-EPA Method 420.1
Herbicides-SW846-8151
Ion Chromatography - EPA Method 300
Residual Chlorine-SM 4500 CLG
VGA's-EPA Method 624
SVGA's-EPA Method 625
Pesticides/ PCB's - EPA Method 608
2,3,7,8 - TCDD - EPA Method 1613B
Halo Acetic Acids -EPA Method 552.3
1, 2 - Dibromoethane - EPA Method 504.1
Table 11. 500 Series Methods
Method
EPA 552.3
EPA 552.3
EPA 552.3
EPA 552.3
EPA 552.3
EPA 552.3
EPA 504.1
Analyte
Dibromoacetic Acid
Dichloroacetic Acid
Monobromoacetic Acid
Monochloroacetic Acid
Trichloroacetic Acid
Haloacetic Acid
1,2-Dibromoethane
PQL, ug/L
0.8
0.8
1.6
0.8
0.8
0.8
0.025
104
-------
Table 12. 600 Series Methods
Method
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 608
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
Analyte
4,4'-DDD
4,4'-DDE
4,4'-DDT
Aldrin
alpha-BHC
Aroclor 1016
Aroclor 1221
Aroclor 1232
Aroclor 1242
Aroclor 1248
Aroclor 1254
Aroclor 1260
beta-BHC
Chlordane
delta-BHC
Dieldrin
Endosulfan 1
Endosulfan II
Endosulfan sulfate
Endrin
Endrin aldehyde
gamma-BHC
Heptachlor
Heptachlor epoxide
Methoxychlor
Toxaphene
1,1,1-Trichloroethane
1,1,2,2-Tetrachloroethane
1,1,2-Trichloroethane
1,1-Dichloroethane
1,1-Dichloroethene
1,2,4-Trichlorobenzene
l,2-Dibromo-3-chloropropane
1,2-Dichlorobenzene
1,2-Dichloroethane
1,2-Dichloropropane
1,3-Dichlorobenzene
1,4-Dichlorobenzene
2-Chloroethyl vinyl ether
Acrolein
Acrylonitrile
PQL,ug/L
0.05
0.05
0.05
0.05
0.05
1
1
1
1
1
1
1
0.05
0.5
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
1
1
1
1
1
1
1
5
1
1
1
1
1
5
50
5
105
-------
Method
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 624
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
Analyte
Benzene
Bromodichloromethane
Bromoform
Bromomethane
Carbon tetrachloride
Chloro benzene
Chloroethane
Chloroform
Chloromethane
Dibromochloromethane
Dibromomethane
Epichlorohydrin
Ethyl benzene
Methylene chloride
Styrene
Tetrachloroethene
Toluene
Trichloroethene
Vinyl chloride
cis-l,2-Dichloroethene
cis-l,3-Dichloropropene
m,p-Xylene
o-Xylene
trans-l,2-Dichloroethene
trans-l,3-Dichloropropene
1,3-Dichloropropylene
Trihalomethane, Total
Xylenes/Total
Epichlorohydrin
1,2,4-Trichlorobenzene
1,2-Diphenylhydrazine
2,4,6-Trichlorophenol
2,4-Dichlorophenol
2,4-Dimethyl phenol
2,4-Dinitrophenol
2,4-Dinitrotoluene
2,6-Dinitrotoluene
2-Chloronaphthalene
2-Chlorophenol
2-Nitrophenol
3,3'-Dichlorobenzidine
4, 6-Dinitro-2-methyl phenol
PQL,ug/L
1
1
1
1
1
1
1
1
1
1
1
10
1
1
1
1
1
1
1
1
1
2
1
1
1
1
1
1
25
5
5
5
5
5
25
5
5
5
5
5
10
25
106
-------
Method
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
EPA 625
4-Bromophenyl phenyl ether
4-Chlorophenyl phenyl ether
4-Nitrophenol
Acenaphthene
Acenaphthylene
Anthracene
Atrazine
Benz(a)anthracene
Benzidine
Benzo(a)pyrene
Benzo(b)fluoranthene
Benzo(g,h,i)perylene
Benzo(k)fluoranthene
Bis(2-chloroethoxy) methane
Bis(2-chloroethyl)ether
Bis(2-chloroisopropyl)ether
Bis(2-ethylhexyl)phthalate
Butyl benzyl phthalate
Chrysene
Di-n-butyl phthalate
Di-n-octyl phthalate
Dibenz(a,h)anthracene
Diethyl phthalate
Dimethyl phthalate
Fluoranthene
Fluorene
Hexachlorobenzene
Hexachlorobutadiene
Hexachlorocyclopentadiene
Hexachloroethane
lndeno(l,2,3-cd)pyrene
Isophorone
N-Nitrosodi-n-propylamine
N-Nitrosodimethylamine
N-Nitrosodiphenylamine
Naphthalene
Nitrobenzene
Pentachlorobenzene
Pentachlorophenol
Phenanthrene
Phenol
Pyrene
PQL,ug/L
5
5
25
5
5
5
5
5
20
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
5
10
5
5
5
5
5
5
5
5
5
25
5
5
5
107
-------
Method
EPA 625
EPA 625
Analyte
2-Methylphenol
3 & 4-Methylphenol
PQL,ug/L
5
5
Table 13. SW846 Series
Method
EPA 8316
EPA 8270
EPA8151A
EPA8151A
EPA8151A
EPA8151A
Analyte
Acrylamide
Atrazine
2,4,5-TP (Silvex)
2,4-D
Dinoseb
Dalapon
PQL, ug/L
80
50
1
0.94
0.47
1.1
Table 14. Inorganic Parameters
Method
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 200.8
EPA 245.1
SM4500 CN E
SM45000 CN 1
Method
SM9222 D
SM9222 B
Method
EPA 300.0
EPA 300.0
EPA 300.0
EPA 420.1
SM 4500 Cl G
Analyte
Antimony
Arsenic
Barium
Beryllium
Cadmium
Chromium
Copper
Lead
Nickel
Selenium
Silver
Thallium
Zinc
Mercury
Cyanide
Cyanide, Weak and dissociable
Analyte
Coliform,Fecal
Coliform, Total
Analyte
Nitrogen,Nitrate (As N)
Nitrogen,Nitrite (As N)
Fluoride
Phenolics, Total Recoverable
Residual Chlorine,Total
PQL,mg/L
0.005
0.005
0.005
0.004
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
0.005
PQL,colonies/100 ml
1
1
PQL,mg/L
0.5
0.5
0.5
0.005
0.1
108
-------
Challenges for Precise Radium Analysis in Brine
Zoltan Szabo
U.S. Geological Survey
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA.
For detailed evaluation of changes in radionuclide concentrations and isotopic ratios for
studying waste-products generation during the fracturing process, analyses require high
precision and accuracy to study progressive changes. For high precision determinations of the
concentrations of Ra-226 with low minimum critical detection levels (MDLs), the most
commonly used technique is the USEPA-approved (Method 903.1) radon (Rn) de-emanation
technique, with scintillation counting of the Rn-222 progeny after cold-trapping on charcoal.
Alpha spectrometry has become increasingly commonly used to determine concentrations of
Ra-226 and Ra-224. A 100-minute count for all (liter) sample is typically long enough to
achieve detection of 1 pCi/L or less for aliquots with simple matrices. Spectral analysis can also
explicitly include short-lived progeny for verification and improved quantification (example,
polonium-216 progeny for Ra-224). The gamma-spectrometric analytical technique (Standard
Method 7500-E of the American Public Health Association, 2005) has become more commonly
used, though it has a precision and MDL that is higher by factors of 2 to 5 times that of the
alpha-spectrometry technique using reasonable operating parameters. The strong benefit of
the technique is that it can determine all four of the naturally occurring Ra isotopes, but the
background gamma count cannot be maintained at levels as low as those achievable for
measurements of alpha particles, thus is also more imprecise under comparable optimal
operating conditions. The counting time required for alpha-spectrometric or Rn de-emanation
technique is shorter (60 or 100 minutes as opposed to 1000 minutes), the sample volume
required for achieving low MDL is smaller (1 L as opposed to 4 to 20 L), and the instrumentation
is less bulky and expensive than that required for gamma-spectrometric determination.
Delayed coincidence counting has also allowed for an increase in precision and level of
detection for multiple isotopes at once.
Multiple-counting inductively coupled plasma mass spectrometry (MC-ICP-MS) and thermal
ionization mass spectrometry (TIMS) for U and Th isotopes and for Ra-226 have increased the
level of available precision and detection substantially. The limitation to these atom-counting
techniques remains the ability to effectively extract and purify the isotopes from brine. One
option in using TIMS is once the Ra-226 is precisely quantified, other techniques such as gamma
spectroscopy that define the ratios of Ra isotopes relative to each other can be used on
remaining sample aliquots to provide a reasonable estimation of the other Ra isotope levels.
Extraction, purification, and pre-concentration along with successful yield monitoring are
critical steps in achieving low detection levels and high precision. Chemical separation by
forming a Ba-Ra-sulfate precipitate is most commonly used to extract Ra isotopes from sample
109
-------
aliquots for Ra-226 and Ra-224 analyses by the Rn de-emanation technique, by alpha
spectrometry, and by gamma spectroscopy. Cation-exchange chromatography (Bio-Rad AG
50W-X8 resin) is typically used to first separate the Ra and Ba, which are then eluted with 8-
molar HN03, and are co-precipitated with barite using a seeding suspension. The seeding
suspension is needed to ensure the formation of uniform fine-grained crystals of the barite that
is required for efficient counting by alpha spectrometry, which is most sensitive of the
techniques to variable geometry of the Ra-bearing precipitate crystals. Experimentation is
ongoing with forming the precipitate on Sr-specific resin filter plates. Using a bed of previously
precipitated microcrystalline barite as the seeding agent has also been tried to improve the
precipitate formation. A widely used approach is to improve barium-specific precipitation
performance by using ethylenediaminetetraacetic acid (EDTA) as a complexing agent to limit
co-precipitation of impurities. The Ra isotopes can also be extracted using Mn-coated fibers
(prepared in various forms: in tubes, filters, or disks). Oxic and alkaline waters are optimal for
use with the Mn-coated fibers, however, in acidic and especially in reducing waters such as
might be encountered in deep brines extraction efficiency is limited by the properties of
manganese oxide. Numerous element-specific resins have been developed for Ra, Pb, U, Th
that may allow for efficient extraction of the target radionuclide from briney solution. The issue
is cost, as the more specialized extraction resins are expensive.
A radioactive tracer (Ba-133) is added to the samples during precipitation of the Ra-bearing
barium sulfate to determine yield. The Ba-133 is analyzed by gamma spectroscopy after sample
purification. The Ba-133 tracer on occasion exhibits variable and low recoveries in Ca- and Sr-
rich waters, and in these cases, may not match Ra recovery quantitatively; thereby, cation-
exchange removal or EDTA-complexing of the competing divalent cations is helpful. For atom
counting techniques, a mixture containing Ra-228 and Th-229 can be added to monitor
recovery of Ra-226 and Th-228. The use of these tracers more precisely quantify Ra yield, which
is appropriate for the more precise measurements possible with atom counting as opposed to
activity counting techniques. Organic compounds provide interference for TIMS analysis, Ba for
MC-ICP-MS analysis, and Th-228 interferes with yield tracing with Ra-228; of these, limiting Ba
poses the greatest challenge. The MDLs and precision for all analysis types are influenced by
dilution, interference effects, temperature or pressure that need to be understood and
minimized.
110
-------
Revisiting the Major Discussion Points of the Technical
Presentation Sessions
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by participants have not been verified or endorsed by EPA. Any mention of trade
names or commercial products does not constitute endorsement or recommendation for use.
The workshop lead and theme leads addressed the workshop participants and EPA at the
conclusion of presentations for each theme and at the end of the workshop. Leads summarized
the major discussion points and commented on research needs and data gaps.
Wilma Subra of Subra Company, the workshop lead, thanked the participants for their
questions and summarized the main points of the presentations and discussions. She said
several presenters emphasized the need for baseline/background sampling, whether required
by state regulations or by industry initiative. According to participants, salinity or chloride alone
is not sufficient as a tracer; many other important parameters should be considered. Ms. Subra
noted an important knowledge gap related to tracers and changing chemistry of recycled
produced water and flowback water. Reuse/recycling is clearly an important topic, but the
effects on tracers and chemicals deserve more study. Other key points included the difference
between sanitizing and sterilizing (with regard to biocide) and air quality. Air quality is not
addressed in the current study plan, but it is an important issue. The need for method and
procedure development was reiterated by Ms. Subra. While this is somewhat outside of the
scope of the EPA study, it forms the basis of investigating potential instances of ground water,
shallow subsurface, and surface contamination. She noted that there are many issues that need
to be addressed by industry, EPA, and academia. Ms. Subra described a few key issues,
including disclosure, wastewater recycling, and tracer selection.
Cal Cooper of Apache Corporation, the Fracture Fluid Chemistry theme co-lead, emphasized
that fracture fluid composition is quite variable, depending on many geologic and operational
factors. However, fracture fluids for individual wells are not composed of large numbers of
chemicals. Industry has changed fluid composition over time to address technical, economic,
and environmental factors. The gas industry produces from formations with a very wide range
of pressure, temperature, and salinity characteristics, but industry works to solve any
challenges posed by these geologic differences. Dr. Cooper noted the movement toward using
more environmentally-friendly chemicals in fracture fluids. In addition, Dr. Cooper said that
industry understands that surface contamination is a key concern that needs to be addressed;
at the same time, the public confuses surface and subsurface issues, which causes perception
problems. Dr. Cooper also mentioned that more information is needed on fate and transport of
deep subsurface flow and on how pressure dissipates at these HF sites.
Tracy Bank of University at Buffalo, the Fracture Fluid Chemistry theme co-lead, speaking from
the academic perspective, said she feels encouraged by the presentation of data from the
industry participants. She asked industry to share data and samples with academia. Dr. Bank
111
-------
also invited workshop participants to submit results to an upcoming Geological Society of
America session related to HF that she is co-chairing with Dr. Kirby.
Jennifer Mclntosh of University of Arizona, the Fingerprinting theme lead, described the need
for collecting baseline water chemistry data before and during HF operations, as well as for
continued time-series sampling after HF operations are completed. Data are needed regarding
water quality from ground water (including shallow aquifers) and native formation waters from
the target shales, as well as data on the chemistry of the injected fluids. She noted that an
understanding of inherent variability is a key component of baseline studies. An understanding
of local (shallow) methane sources is also important. Dr. Mclntosh also described the power of
combining multiple tracers, including water chemistry, stable and radiogenic isotopes, and gas
components. The combination of tracers is especially important as some chemical constituents
(e.g., pH, metals) are not conservative and can change as conditions (physical, chemical,
biological) change within the formation, within the well bore, inside storage tanks, etc. Dr.
Mclntosh also emphasized the importance of understanding the basic source of solutes in shale
formation waters to distinguish them from injected HF fluids.
Kesavalu Bagawandoss of Accutest Labs, the Field and Analytical Challenges theme lead,
thanked the speakers for an interesting discussion on analytical techniques and sampling. He
described the major points of the discussion, including the need for the development of
additional analytical methods and the need for defining expectations for analytical techniques.
He also noted the need for good data on the chemical components that are used in HF, as well
as the natural chemical characteristics of the subsurface. He emphasized the importance of
sample collection; an analytical method is only as good as the sample collected. Because of the
pressures and temperatures involved, sampling for HF purposes is a sophisticated process and
requires trained personnel following established protocols. The radionuclides analysis methods
must be updated to account for the interferences due to the presence of barium, strontium,
and calcium in the matrix, as outlined in the presentations.
Summary of Discussions Following Workshop and Theme Lead Summaries
The statements made during the workshop do not represent the views or opinions of EPA. The
claims made by meeting participants have not been verified or endorsed by EPA. Any mention of
trade names or commercial products does not constitute endorsement or recommendation for
use.
The importance of surface water. A participant stated that in EPA Region 3 (Pennsylvania,
Maryland, West Virginia, Virginia, Delaware, and Washington, DC), 80% of the population is
served by surface water. One participant noted that surface water issues will be addressed at
the fourth workshop on Water Resource Management.
The transport, storage, and use ofHCI. A participant explained that HCI for HF is brought to the
site in its own tank and is not stored on location. The HCI reacts rapidly downhole, usually with
112
-------
common salts, and is completely spent, claimed the participant, and it is not returned to the
surface. The participant indicated that the concentration used is usually 15% HCI by weight, and
the maximum concentration that can be held by the tanks is 28% by weight. One industry
participant noted that his company recently began testing an acid reduction program and
subsequently eliminated acid use from 80% of their HF jobs in southwestern Pennsylvania. He
noted that using acid is a site-specific decision.
The possibility of a test site. A participant suggested that a test site be established where
industry, academia, and EPA could all have access for sampling. One participant pointed out
that there is no such thing as a "typical" HF site, due to geographic variations in the shales.
Several participants raised potential safety concerns. One participant recommended that EPA
participate in a short course on the basics of HF.
113
-------
Glossary of Terms
The sources of the definitions found in this glossary are noted at the end of each definition.
Sources include the following:
Abbreviated Source Full Source Name
SPE Society of Petroleum Engineers Exploration & Production Glossary
(http://www.spe.org/glossary/wiki/doku.php/)
Schlumberger Schlumberger Oilfield Glossary
(http://www.glossary.oilfield.slb.com/default.cfm)
ABBREVIATIONS
BTEX benzene, toluene, ethyl benzene and xylene
CMC critical micelle concentration
DBNPA 2,2-dibromo,3 nitriloproprionamide
DIC dissolved inorganic carbon
DQI data quality indicator
DRH diesel range hydrocarbons
EDTA ethylenediaminetetraacetic acid
EXAFS extended x-ray absorption fine structure
FIFRA Federal Insecticide, Fungicide and Rodenticide Act
FR friction reducer
GMWL global meteoric water line
GPA Gas Processors Association
HF hydraulic fracturing
HPG hydroxypropyl guar; a chemically modified guar (SPE)
1C inorganic carbon
ICPMS inductively coupled plasma mass spectrometry
INAA instrumental neutron activation analysis
InO GE InnovOxTOC Analyzer
LPG liquified petroleum gas
MC-ICP-MS multiple-counting inductively coupled plasma mass spectrometry
MCL maximum contaminant level
mD millidarcy
MDL method detection limit
MIT mechanical integrity test
MSI Marcellus Shale Initiative
NDIR non-dispersive infrared
NORM naturally occurring radioactive material
NPOC non-purgeable organic carbon
NSLS National Synchrotron Radiation Lightsource
PQL practical quantitation limit
QA quality assurance
QAPjP quality assurance project plan
QC quality control
SCW supercritical water
SCWO supercritical water oxidation
SEM/EDX scanning electron microscopy energy dispersive x-ray
SRB sulfate reducing bacteria
114
-------
SRV stimulated rock volume
STP surface treating pressure
SVOCs semivolatile organic compounds
IDS total dissolved solids
THPS tetrakis hydroxymethyl phosphonium sulfate
TIC total inorganic carbon
TIMS thermal ionization mass spectrometry
TMAC tetramethylammonium chloride
TOC total organic carbon
TOP SIMS time-of-flight secondary ion mass spectrometry
UIC Underground Injection Control
ULWP ultra-lightweight proppants
USGS United States Geological Survey
VES viscoelastic surface-based system
VOCs volatile organic compounds
VSMOW Vienna Standard Mean Ocean Water
XANES x-ray absorption near edge spectroscopy. Also called uXANES.
U.-SXRF micro-x-ray fluorescence
GLOSSARY
baseline water quality water quality data collected prior to the beginning of an activity
flowback The process of allowing fluids to flow from the well following a treatment, either in preparation for a
subsequent phase of treatment or in preparation for cleanup and returning the well to production
(Schlumberger)
imbibition absorption and adsorption of fluids into the pores of the rock (SPE)
interfacial tension a measurement of the difficulty of moving a fluid past another fluid (SPE)
irreducible water saturation the fraction of the pore space occupied by water when the hydrocarbon content is
at maximum. This level of water can only be reduced by flow of very dry gas that evaporates the water. (SPE)
kerogen An initial stage of oil that never developed completely into crude. Typical of oil shales. (SPE)
leak off The magnitude of pressure exerted on a formation that causes fluid to be forced into the formation. The
fluid may be flowing into the pore spaces of the rock or into cracks opened and propagated into the formation
by the fluid pressure. (Schlumberger)
retrograde condensate condensate that precipitates in the pore space of the rock in a previously dry gas (single
phase) reservoir as the pressure drops below the dew point. The building condensate severely reduces the
permeability to gas (SPE)
shut-in pressure The surface force per unit area exerted at the top of a wellbore when it is closed at either the
Christmas tree or the BOP stack. The pressure may be from the formation or an external and intentional source.
The SIP may be zero, indicating that any open formations are effectively balanced by the hydrostatic column of
fluid in the well. If the pressure is zero, the well is considered to be dead, and can normally be opened safely to
the atmosphere. (Schlumberger)
slick water a water base fluid with only a very small amount of a polymer added to give friction reduction benefit
(SPE)
tight formation non specific term meaning lower permeability (SPE)
type-ll kerogen Exinite; has an intermediate hydrogen to carbon and oxygen to carbon ratio. Oil and gas prone
with yields of 40 to 60%. From plant sources. (SPE) See also kerogen.
type-Ill kerogen Vitrinite; has a low hydrogen to carbon and high oxygen to carbon ratio.Low quality gas prone
with low yields. Source is wood and high order plant debris. Coal precursor. (SPE) See also kerogen.
workover repairing a well. Usually implies opening the well and running in with a tubing string. May or may not
involve killing the well and may or may not involve a conventional rig (SPE)
115
-------
Printed with vegetable-based ink on paper that
contains a minimum of 50% post-consumer fiber
content and processed chlorine free.
United States
Environmental Protection
Agency
PRESORTED STANDARD
POSTAGES FEES PAID
EPA
PERMIT NO. G-35
Office of Research and Development (8101R)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |