<pubnumber>816R04003</pubnumber>
<title>Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic Fracturing of Coalbed Methane Reservoirs</title>
<pages>463</pages>
<pubyear>2004</pubyear>
<provider>NEPIS</provider>
<access>online</access>
<origin>PDF</origin>
<author></author>
<publisher></publisher>
<subject></subject>
<abstract></abstract>
<operator>mja</operator>
<scandate>03/02/11</scandate>
<type>single page tiff</type>
<keyword></keyword>
Evaluation of Impacts to
Underground Sources of
Drinking Water by Hydraulic
Fracturing of Coalbed
Methane Reservoirs
Final
image:
Office of Water
Office of Ground Water and Drinking Water (4606M)
EPA816-R-04-003
www.epa.gov/safewater
June 2004
.•TV
£:£
Printed on Recycled Paper
image:
EPA816-R-04-003
Evaluation of Impacts to Underground Sources of Drinking
Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
FINAL
June 2004
United States Environmental Protection Agency
Office of Water
Office of Ground Water and Drinking Water
Drinking Water Protection Division
Prevention Branch
1200 Pennsylvania Avenue, NW (4606M)
Washington, DC 20460
in
image:
TABLE OF CONTENTS
EXECUTIVE SUMMARY ES-1
ES-l How Does CBM Play a Role in the Nation's Energy Demands? ES-2
ES-2 What Is Hydraulic Fracturing? ES-4
ES-3 Why Did EPA Evaluate Hydraulic Fracturing? ES-7
ES-4 What Was EPA's Project Approach? ES-8
ES-5 How Do Fractures Grow? ES-10
ES-6 What Is in Hydraulic Fracturing Fluids? ES-11
ES-7 Are Coalbeds Located within USDWs? ES-13
ES-8 Did EPA Find Any Cases of Contaminated Drinking Wells Caused by
Hydraulic Fracturing in CBM Wells? ES-13
ES-9 What Are EPA's Conclusions? ES-16
CHAPTER 1. INTRODUCTION 1-1
. 1 EPA's Rationale for Conducting This Study
.2 Overview of Hydraulic Fracturing
.3 EPA's Authority to Protect Underground Sources of Drinking Water
.4 Potential Effects of Hydraulic Fracturing of Coalbed Methane Wells
onUSDWs
.5 Study Approach
.6 Stakeholder Involvement
.7 Information Contained within This Report
-2
-3
-4
-6
-7
-9
CHAPTER 2. STUDY METHODOLOGY 2-1
2.1 Overview of the Study Methods 2-1
2.2 Information Sources 2-3
2.2.1 Literature Reviews 2-4
2.2.2 Department of Energy 2-5
2.2.3 Interviews 2-5
2.2.4 Field Visits 2-6
2.2.5 Federal Register Notice to Identify Reported Incidents 2-7
2.3 Review Process 2-7
CHAPTER 3. CHARACTERISTICS OF COALBED METHANE PRODUCTION
AND ASSOCIATED HYDRAULIC FRACTURING PRACTICES 3-1
3.1 Introduction 3-1
3.2 Hydraulic Fracturing 3-4
3.2.1 The Hydraulic Fracturing Process 3-4
3.2.2 Factors Affecting Fracture Behavior 3-5
3.3 Fracturing Fluids 3-10
3.3.1 Quantifying Fluid Recovery 3-11
3.3.2 Mechanisms Affecting Fluid Recovery 3-12
3.4 Measuring and Predicting the Extent of Fluid Movement 3-15
3.4.1 Direct Measurements 3-16
3.4.2 Indirect Measurements 3-18
3.4.3 Model Estimates 3-19
3.4.4 Limitations of Fracture Diagnostic Techniques 3-20
3.5 Summary 3-22
IV
image:
CHAPTER 4. HYDRAULIC FRACTURING FLUIDS 4-1
4.1 Introduction 4-1
4.2 Types of Fracturing Fluids and Additives 4-2
4.2.1 Gelled Fluids 4-3
4.2.2 Foamed Gels 4-5
4.2.3 Water and Potassium Chloride Water Treatments 4-6
4.2.4 Acids 4-6
4.2.5 Fluid Additives 4-7
4.2.6 Proppants 4-8
4.3 The Fate and Transport of Stimulation Fluids Injected into Coal and Surrounding
Rock During Hydraulic Fracturing of Coalbed Methane Reservoirs (with a Special Focus
on Diesel Fuel) 4-11
4.3.1 Point-of-Injection Calculation 4-13
4.3.2 Fracturing Fluid Recovery 4-15
4.3.3 The Influence of the Capture Zone 4-16
4.3.4 Fate and Transport Considerations 4-16
4.4 Summary 4-29
CHAPTER 5. SUMMARY OF COALBED METHANE BASIN DESCRIPTIONS 5-1
5.1 The San Juan Basin 5-1
5.2 The Black Warrior Basin 5-2
5.3 The Piceance Basin 5-3
5.4 The Uinta Basin 5-4
5.5 The Powder River Basin 5-5
5.6 The Central Appalachian Basin 5-6
5.7 The Northern Appalachian Basin 5-7
5.8 The Western Interior Coal Region 5-8
5.9 The Raton Basin 5-10
5.10 The Sand Wash Basin 5-11
5.11 The Washington Coal Regions (Pacific and Central) 5-12
5.12 Summary 5-14
CHAPTER 6. WATER QUALITY INCIDENTS 6-1
6.1 The San Juan Basin (Colorado and New Mexico) 6-2
6.1.1 Summary of Reported Incidents 6-2
6.1.2 State Agency Follow-Up in the San Juan Basin 6-4
6.1.3 Major Studies That Have Been Conducted in the San Juan Basin 6-5
6.2 The Powder River Basin (Wyoming and Montana) 6-9
6.3 The Black Warrior Basin (Alabama) 6-10
6.3.1 Summary of Reported Incidents 6-10
6.3.2 State Agency Follow-Up (Alabama Oil and Gas Board) 6-11
6.4 The Central Appalachian Basin (Virginia and West Virginia) 6-13
6.4.1 Summary of Virginia Incidents 6-14
6.4.2 State Agency Follow-Up (Virginia DMME) 6-15
6.5 Summary 6-16
CHAPTER?. CONCLUSIONS AND RECOMMENDATIONS 7-1
7.1 Reported Water Quality Incidents 7-1
7.2 Fluid Injection Directly into USDWs or into Coal Seams Already in Hydraulic
Communication with USDWs 7-2
7.3 Breach of Confining Layer 7-3
7.4 Conclusions 7-5
image:
REFERENCES MR-1
APPENDIX A: Department of Energy - Hydraulic Fracturing White Paper APP. A-l
APPENDIX B: Quality Assurance Plan APP. B-l
1.0 Project Management APP. B-l
1.1 Project and Task Organization APP. B-l
1.2 Problem Definition and Background APP. B-3
1.3 Project and Task Description APP. B-4
1.4 Quality Objectives and Criteria APP. B-5
1.5 Special Training and Certification APP. B-6
1.6 Documents and Records APP. B-7
2.0 Data Generation and Acquisition APP. B-7
2.1 Non-Direct Measurements APP. B-7
2.2 Data Management APP. B-9
3.0 Assessment and Oversight APP. B-9
4.0 Data Validation and Usability APP. B-10
4.1 Data Review, Verification, and Validation APP. B-10
4.2 Reconciliation with User Requirements APP. B-l 1
4.2.1 Drawing Conclusions APP. B-12
4.2.2 Communication of Findings APP. B-12
ATTACHMENT 1. THE SAN JUAN BASIN Al-1
1.1 Basin Geology Al-1
1.2 Basin Hydrology and USDW Identification Al-3
1.3 Coalbed Methane Production Activity A1 -5
1.4 Summary Al-8
ATTACHMENT 2. THE BLACK WARRIOR BASIN A2-1
2.1 Basin Geology A2-1
2.2 Basin Hydrology and USDW Identification A2-2
2.3 Coalbed Methane Production Activity A2-2
2.4 Summary A2-6
ATTACHMENTS. THE PICEANCE BASIN A3-1
3.1 Basin Geology A3-1
3.2 Basin Hydrology and USDW Identification A3 -3
3.3 Coalbed Methane Production Activity A3-5
3.4 Summary A3-6
ATTACHMENT 4. THE UINTA COAL BASIN A4-1
4.1 Basin Geology A4-1
4.2 Basin Hydrology and USDW Identification A4-2
4.3 Coalbed Methane Production Activity A4-4
4.4 Summary A4-5
ATTACHMENTS. THE POWDER RIVER COAL BASIN A5-1
5.1 Basin Geology A5-1
5.2 Basin Hydrology and USDW Identification A5-4
5.3 Coalbed Methane Production Activity A5-5
5.4 Summary A5-9
VI
image:
ATTACHMENT 6. THE CENTRAL APPALACHIAN COAL BASIN A6-1
6.1 Basin Geology A6-1
6.2 Basin Hydrology and USDW Identification A6-3
6.3 Coalbed Methane Production Activity A6-5
6.4 Summary A6-7
ATTACHMENT 7. THE NORTHERN APPALACHIAN COAL BASIN A7-1
7.1 Basin Geology A7-1
7.2 Basin Hydrology and USDW Identification A7-2
7.3 Coalbed Methane Production Activity A7-5
7.4 Summary A7-6
ATTACHMENT 8. THE WESTERN INTERIOR COAL REGION A8-1
8.1 Basin Coals A8-1
8.1.1 Arkoma Basin Coals A8-2
8.1.2 Cherokee Basin Coals A8-2
8.1.3 Forest City Basin Coals A8-2
8.2 Basin Hydrology and USDW Identification A8-3
8.2.1 Arkoma Basin Hydrology and USDW Identification A8-3
8.2.2 Cherokee Basin Hydrology and USDW Identification A8-4
8.2.3 Forest City Basin Hydrology and USDW Identification A8-6
8.3 Coalbed Methane Production Activity A8-9
8.3.1 Arkoma Basin Production Activity A8-9
8.3.2 Cherokee Basin Production Activity A8-10
8.3.3 Forest City Basin Production Activity A8-11
8.4 Summary A8-11
ATTACHMENT 9. THE RATON BASIN A9-1
9.1 Basin Geology A9-1
9.2 Basin Hydrology and USDW Identification A9-2
9.3 Coalbed Methane Production Activity A9-3
9.4 Summary A9-5
ATTACHMENT 10. THE SAND WASH COAL BASIN A10-1
10.1 Basin Geology A10-1
10.2 Basin Hydrology and USDW Identification A10-3
10.3 Coalbed Methane Production Activity A10-4
10.4 Summary A10-5
ATTACHMENT 11. THE WASHINGTON COAL REGION (PACIFIC AND
CENTRAL) All-1
11.1 Basin Geology All-1
11.1.1 Pacific Coal Region Geology All-2
11.1.2 Central Coal Region Geology All-4
11.2 Basin Hydrology and USDW Identification Al 1-5
11.3 Coalbed Methane Production Activity A11 -5
11.3.1 Pacific Coal Region Production Activity All-6
11.3.2 Central Coal Region Production Activity All-6
11.4 Summary All-7
vn
image:
LIST OF TABLES
EXECUTIVE SUMMARY
ES-l Coal Basins Production Statistics and Activity Information in the U.S. ES-3
ES-2 Evidence in Support of Coal-USDW Co-Location in U.S. Coal Basins ES-14
CHAPTER 3
3-1 Limitations of Fracture Diagnostic Techniques 3-21
CHAPTER 4
4-1 Characteristics of Undiluted Chemicals Found in Hydraulic Fracturing Fluids 4-9
(Based on MSDSs)
4-2 Estimated Concentrations of Diesel Contaminants in Fracturing Fluids at the
Point-of-Injection and Factors Affecting Their Concentrations and Movement
in Groundwater 4-18
CHAPTER 5
5-1 Evidence in Support of Coal-USDW Co-Location in U.S. Coal Basins 5-15
APPENDIX A: Department of Energy - Hydraulic Fracturing White Paper
Table 1 Sources of Data APP. A-4
Table 2 Fracturing Fluids and Conditions for Their Use APP. A-6
Table3 Typical Ranges of Young's Modulus for Various Lithologies APP. A-8
Table 4 Acceptable Levels for Mixed Water APP. A-12
Table 5 Summary of Chemical Additives APP. A-13
Table 6 Limitations of Fracture Diagnostic Techniques APP. A-21
APPENDIX B: Quality Assurance Plan
B-l Peer Review Panel APP. B-2
ATTACHMENT 1
A1 -1 Chemical Components of Typical Fracture/Stimulation Fluids Used for
San Juan Coalbed Methane Wells A1 -9
ATTACHMENT 2
A2-1 Chemical Components Previously Used in Typical Fracturing/Stimulation
Fluids for Alabama Coalbed Methane Wells A2-8
ATTACHMENT 5
A5-1 Average Water Quality Results from Produced Waters A5-8
ATTACHMENT 6
A6-1 Relative Locations of USDWs and Methane-Bearing Coalbeds A6-5
ATTACHMENT 7
A7-1 Relative Locations of USDWs and Methane-Bearing Coalbeds A7-4
ATTACHMENT 8
A8-1 Relative Locations of USDWs and Potential Methane-Bearing Coalbeds,
Arkoma Basin A8-4
A8-2 Relative Locations of USDWs and Potential Methane-Bearing Coalbeds,
Cherokee Basin A8-6
A8-3 Relative Locations of USDWs and Potential Methane-Bearing Coalbeds,
Forest City Basin A8-8
viii
image:
LIST OF FIGURES
EXECUTIVE SUMMARY
ES-l Major U.S. Coal Basins ES-3
ES-2 A Graphical Representation of the Hydraulic Fracturing Process
in Coalbed Methane Wells ES-5
ES-3 Hypothetical Mechanisms-Direct Fluid Injection into a USDW ES-9
ES-4 Hypothetical Mechanisms - Fracture Creates Connection to USDW ES-10
ES-5 and ES-6 Photos from a Hydraulic Fracturing Visit ES-12
CHAPTER 1
1-1 Locus Map of Major U.S. Coal Basins 1-2
1-2 Hypothetical Mechanisms - Direct Fluid Injection into a USDW 1-10
1-3 Hypothetical Mechanisms - Fracture Creates Connection to USDW 1-11
CHAPTER 3
3-1 Major United States Coal Basins 3-24
3 -2 Geography of an Ancient Peat-Forming System 3-25
3-3 Schematic Representation of "Face Cleat" and "Butt Cleat" 3-26
3-4 A Graphical Representation of the Hydraulic Fracturing Process in
Coalbed Methane Wells 3-27
3-5 Water and Gas Production Over Time 3-29
3-6 Side-View of a Vertical Hydraulic Fracture Typical of Coalbeds 3-30
3-7 Plan View of Vertical, Two-Winged Coalbed Methane Fracture Showing the
Reservoir Region Invaded by Fracturing Fluid Leakoff 3-31
3-8 Plan View (Looking Down the Wellbore) of a Vertical Hydraulic Fracture 3-32
CHAPTER 4
4-1 through 4-11 Photos from a Hydraulic Fracturing Field Visit 4-20
CHAPTER 5
5-1 Locus Map of Major United States Coal Basins 5-17
APPENDIX A: Department of Energy - Hydraulic Fracturing White Paper
Fig. 1 Typical Input Data for a PSD Model APP. A-4
Fig. 2 Fracture Treatment Optimization Project APP. A-5
Fig. 3 Local In-Situ Stress and Depth APP. A-7
Fig. 4 Cased Hole Test Configuration APP. A-9
Fig. 5 Typical Stress Test Pump-In/Shut-in APP. A-9
Fig. 6 Closure Pressure Analysis APP. A-9
Fig. 7 PKN Geometry APP. A-10
Fig. 8 KGD Geometry APP. A-10
Fig. 9 Width from a PSD Model APP. A-l 1
Fig. 10 Width and Height from PSD Model APP. A-l 1
Fig. 11 Definition of Fracture Conductivity APP. A-14
Fig. 12 Effective Stress on Proppant APP. A-14
Fig. 13 Effect of Stress on Conductivity APP. A-15
Fig. 14 Selecting a Fracture Fluid APP. A-16
Fig. 15 Fracturing Using Coil Tubing APP. A-17
Fig. 16 Proppant Selection Based on Closure Pressure APP. A-17
Fig. 17 Economic Analysis APP. A-18
Fig. 18 Principle of Microseismic Fracture Mapping APP. A-20
ix
image:
ATTACHMENT 1
A1 -1 Regional Tectonic Setting of the San Juan Basin A1 -10
Al-2 Generalized Hydrogeologic Cross-Section of the San Juan Basin Al-11
A1 -3 Isopach Map of the Fruitland Formation Including Pictured Cliffs Tongues A1 -12
Al-4 Cross Section S-10 from Figure Al-3, a Stratigraphic Strike Section Al-13
Al-5 Cross Section D-20 from Figure Al-3, a Stratigraphic Dip Section from the
Fruitland to the Oj o Alamo A1 -14
Al-6 Cross Section E-W from Figure Al-3, a Stratigraphic Strike Section from
the Southeastern San Juan Basin, Showing the Erosional Fruitland-Ojo
Alamo Contact Al-15
Al-7 Areas of the San Juan Basin That Exhibit Similar Characteristics for Production,
Coal Properties, and Hydrologic Pressure Al-16
A1 -8 Map of the Potentiometric Surface of the Fruitland Aquifer A1 -17
Al-9 General Ground Water Flow in the Fruitland/Pictured Cliffs Aquifer System,
San Juan Basin Al-18
Al-10 Generalized Flow Paths of the Fruitland/Pictured Cliffs Aquifer System,
San Juan Basin Al-19
A1 -11 Equipotentials and Flow Paths from Ground Water Flow Modeling
of the San Juan Basin Al-20
Al-12 Chloride Concentration Map (mg/L) of Waters of the Fruitland Aquifer,
San Juan Basin A1 -21
Al-13 Histograms of Water Analyses (mg/L) from the Fruitland/Pictured Cliffs
Aquifer System in the North Central and South-Margin Areas of the San Juan Basin Al-22
Al-14 Direction of Ground Water Flow and Dissolved Solids Concentration in
Tertiary Rocks Al-23
Al-15 Outline of the Fairway Zone of Area 1 of the San Juan Basin Al-24
Al-16 Conceptual Schematic (Plan View) of Tensile Fracture and Shear Failure
in Coal Formed by Openhole Cavitation Cycling Al-25
Al-17 Table of Fracture Stimulation Treatments in the Fruitland Formation of the
San Juan Basin Al-26
Al-18 Density of Wells in the Northern Portion of Area 1 in the San Juan Basin,
as of 12/31/1990 Al-27
Al-19 Fruitland Net Coal Map Al-28
Al-20 Plan View of a Vertical, Two-Winged Coalbed Methane Fracture Showing
the Reservoir Region Invaded by Fracturing Fluid Leakoff Al-29
ATTACHMENT 2
A2-1 Coalbed Methane Fields of Alabama A2-9
A2-2 Coal Cycles of the Pottsville Formation in the Black Warrior Basin A2-10
A2-3 Cross-Section of the Pottsville Formation in the Deerlick Creek Field A2-11
A2-4 Hydrogeologic Cross-Section of the Pottsville Formation in the Brookwood Field A2-12
A2-5 TDS Concentration of Pottsville Aquifer, Black Warrior Basin, Alabama A2-13
A2-6 TDS Concentration of the Mary Lee Coal Seam of the Pottsville Aquifer in the
Eastern Part of the Black Warrior Basin, Alabama A2-14
A2-7 Generalized Increase in TDS Concentration with Depth in the Pottsville Aquifer,
Black Warrior Basin, Alabama A2-15
A2-8 "Stiff Diagram of Water Salinity A2-16
A2-9 Relationship of Water Salinity to Structural Features in the Oak Grove Field,
Black Warrior Basin, Alabama A2-17
A2-10 Productive Coal Seams and the Typical Number of Stimulations Per Well as
of 1993, Black Warrior Basin, Alabama A2-18
A2-11 Sketch of a Vertical, Unconfined Fracture Typical of the Black Warrior Basin A2-19
image:
ATTACHMENT 3
A3-1 Tectonic Map of the Piceance Basin A3-8
A3-2 Stratigraphic Section of the Piceance Basin A3-9
A3-3 Generalized Depth to Base of Coal - Cameo Group A3-10
A3-4 Locations of Gas Fields A3-11
A3-5 Exploration Target Areas, Piceance Basin A3-12
A3 -6 Diagrammatic East - West Sections of Hydrologic System A3 -13
A3-7 Dominant Chemical Constituents in the Two Major Bedrock Aquifers A3-14
ATTACHMENT 4
A4-1 Index Map of Coal Fields in Uinta Basin, Utah A4-6
A4-2 Stratigraphic Column for the Cretaceous of the Castle Valley A4-7
A4-3 Cross Section of Cretaceous Rocks A4-8
ATTACHMENT 5
A5-1 Location of the Powder River Basin of Wyoming and Montana A5-10
A5-2 Map of the Development of Coalbed Methane in the Powder River Basin A5-11
A5-3 Conceptual Cross Section of the Powder River Basin (West - East) A5-12
A5-4 Stratigraphic Diagram of Geology in the Powder River Basin A5-13
A5-5 Detailed Cross Section of the Wasatch and Fort Union Formations
in the Powder River Basin A5-14
A5-6 Conceptual Cross Section (West - East) of the Fort Union Formation
in the Eastern Powder River Basin Near Gillette, Wyoming A5-15
A5-7 Conceptual Cross Section Near the Center of the Powder River Basin A5-16
A5-8 Conceptual Cross Section in the Western Powder River Basin
Near Lake De Smet A5 -17
A5-9 Graph of Coalbed Methane Production and Wells in Service in the
Powder River Basin, 1989 to 1999 A5-18
ATTACHMENT 6
A6-1 Area of Highest Methane Concentration A6-8
A6-2 Structural Features A6-9
A6-3 Representative Stratigraphic Column of Pennsylvanian Age Formations A6-10
A6-4 Isopach Map: Thickness of Cover Over the Pocahontas No. 3 Coalbed A6-11
A6-5 Isopach Map: Thickness of Cover Over the Pocahontas No. 4 Coalbed A6-12
A6-6 Isopach Map: Thickness of Cover Over the Fire Creek - Lower Horsepen Coalbed A6-13
A6-7 Isopach Map: Thickness of Cover Over Beckley - War Creek Coalbed A6-14
A6-8 Isopach Map: Thickness of Cover Over the Sewell/Lower Seaboard Coalbed A6-15
A6-9 Isopach Map: Thickness of Cover Over the Leager/Jawbone Coalbed A6-16
ATTACHMENT 7
A7-1 Index Map Showing County Names A7-7
A7-2 Structure Map A7-8
A7-3 Generalized Stratigraphic Column of the Northern Appalachian Coal Basin A7-9
A7-4 Isopach Map: Depth of Cover to the Brookville/Clarion Group Coals A7-10
A7-5 Isopach Map: Depth of Cover to the Kittanning Group Coals A7-11
A7-6 Isopach Map: Depth of Cover to the Freeport Group Coals A7-12
A7-7 Isopach Map: Depth of Cover to the Pittsburgh Group Coals A7-13
A7-8 Isopach Map: Depth of Cover to the Sewickley Group Coals A7-14
A7-9 Isopach Map: Depth of Cover to the Waynesburg Group Coals A7-15
XI
image:
ATTACHMENT 8
A8-1 Western Interior Coal Basin - Index Map of the Arkoma, Cherokee and
Forest City Basins A8-13
A8-2 Western Interior Coal Basin - Generalized Stratigraphic Column of the
Pennsylvanian System in the Arkoma Basin A8-14
A8-3 Western Interior Coal Basin - Stratigraphic Column of the Cherokee Group
in the Cherokee Basin A8-15
A8-4 Western Interior Coal Basin - Forest City Basin Study Area A8-16
A8-5 Counties, Aquifers, and Physiographic Provinces of Arkansas A8-17
A8-6 Counties, Aquifers, and Physiographic Provinces of Oklahoma A8-18
A8-7 Counties, Aquifers, and Physiographic Provinces of Kansas A8-19
A8-8 Counties, Aquifers, and Physiographic Provinces of Missouri A8-20
A8-9 Western Interior - Detail of Forest City Basin with Detail of Cherokee Basin
in Missouri A8-21
A8-10 Water Quality (TDS) of Lower Paleozoic Aquifers in Kansas A8-22
A8-11 Counties, Aquifers, and Physiographic Provinces of Iowa A8-23
A8-12 Western Interior Coal Basin - Quality of Ground Water in the Paleozoic Aquifers
of Missouri A8-24
A8-13 Counties, Aquifers, and Physiographic Provinces of Nebraska A8-25
ATTACHMENT 9
A9-1 Structure Contour Map on Top Trinidad Sandstone A9-6
A9-2 Structural Cross Section A9-7
A9-3 Generalized Stratigraphy of Cenozoic and Mesozoic Units A9-8
A9-4 Vermejo Formation - Total Coal Isopach A9-9
A9-5 Overburden to Coal Interval A9-10
A9-6 Location of Stratigraphic Cross Sections A9-11
A9-7 Cross Section A - A1 A9-12
A9-8 Cross Section C - C1 A9-13
A9-9 Potentiometric Surface Map for Raton Basin A9-14
A9-10 Relationship Between Gas Content and Depth Below Potentiometric
Surface for Two Groups of Coal Rank A9-15
A9-11 Historical Gas and Water Production for Typical Well Showing How
Water Withdrawal Decreases and Methane Production Increases A9-16
A9-12 Historical Gas and Water Production for Ozzello 42-1 Well Showing
Water Withdrawal Increasing with Gas Production A9-17
ATTACHMENT 10
A10-1 Location of the Sand Wash Basin of Colorado and Wyoming A10-7
A10-2 Diagram of Geologic Formations within the Sand Wash Basin
and Neighboring Basins A10-8
A10-3a Map of Coal and Geologic Features Within the Sand Wash Basin A10-9
A10-3b Conceptual Cross Section C - C1 A10-10
A10-4 Location of the Sand Wash Basin in Relation to the Western Interior Seaway
of Upper Cretaceous Times A10-11
A10-5 Ground Water Quality Trends in the Sand Wash Basin A10-12
A10-6 Comparison of Features Relevant to Coalbed Methane
Production - San Juan Basin and Sand Wash Basin A10-13
ATTACHMENT 11
Al 1-1 The Pacific Coal Region Showing Targeted Subbasins Used for
Coalbed Methane Estimates Al 1-8
Al 1-2 The Central Coal Region Al 1-9
Al 1-3 Major Coal-Bearing Areas in Western Washington Al 1-10
A11 -4 Stratigraphy for Three Coal Districts of the Pacific Coal Region A11 -11
All-5 Structural Map of the Central Columbia Basin and Yakima Fold Belt All-12
xii
image:
EPA816-R-04-003
Executive Summary
A USDW is defined as an aquifer or a portion of an
aquifer that:
A. 1. Supplies any public water system; or
2. Contains sufficient quantity ofgroundwater to
supply a public water system; and
i. currently supplies drinking water for human
consumption; or
ii. contains fewer than 10,000 milligrams per
liter (mg/L) total dissolved solids (TDS); and
B. Is not an exempted aquifer.
NOTE: Although aquifers with greater than 500 mg/L
TDS are rarely used for drinking water supplies
without treatment, the Agency believes that protecting
waters with less than 10,000 mg/L TDS will ensure an
adequate supply for present and future generations.
The U.S. Environmental Protection
Agency (EPA, or the Agency)
conducted a study that assesses the
potential for contamination of
underground sources of drinking
water (USDWs) from the injection
of hydraulic fracturing fluids into
coalbed methane (CBM) wells. To
increase the effectiveness and
efficiency of the study, EPA has
taken a phased approach. Apart
from using real world observations
and gathering empirical data, EPA
also evaluated the theoretical
potential for hydraulic fracturing to
affect USDWs. Based on the
information collected and reviewed, EPA has concluded that the injection of hydraulic
fracturing fluids into CBM wells poses little or no threat to USDWs and does not justify
additional study at this time. EPA's decision is consistent with the process outlined in
the April, 2001 Final Study Design, which is described in Chapter 2 of this report.
The first phase of the study, documented in this report, is a fact-finding effort based
primarily on existing literature to identify and assess the potential threat to USDWs
posed by the injection of hydraulic fracturing fluids into CBM wells. EPA evaluated that
potential based on two possible mechanisms. The first mechanism was the direct
injection of fracturing fluids into a USDW in which the coal is located, or injection of
fracturing fluids into a coal seam that is already in hydraulic communication with a
USDW (e.g., through a natural fracture system). The second mechanism was the creation
of a hydraulic connection between the coalbed formation and an adjacent USDW.
EPA also reviewed incidents of drinking water well contamination believed to be
associated with hydraulic fracturing and found no confirmed cases that are linked to
fracturing fluid injection into CBM wells or subsequent underground movement of
fracturing fluids. Although thousands of CBM wells are fractured annually, EPA did not
find confirmed evidence that drinking water wells have been contaminated by hydraulic
fracturing fluid injection into CBM wells.
EPA has determined that in some cases, constituents of potential concern (section ES-6)
are injected directly into USDWs during the course of normal fracturing operations. The
use of diesel fuel in fracturing fluids introduces benzene, toluene, ethylbenzene, and
xylenes (BTEX) into USDWs. BTEX compounds are regulated under the Safe Drinking
Water Act (SOWA).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-1
image:
EPA816-R-04-003
Given the concerns associated with the use of diesel fuel and the introduction of BTEX
constituents into USDWs, EPA recently entered into a Memorandum of Agreement
(MOA) with three major service companies to voluntarily eliminate diesel fuel from
hydraulic fracturing fluids that are injected directly into USDWs for CBM production
(USEPA, 2003). Industry representatives estimate that these three companies perform
approximately 95 percent of the hydraulic fracturing projects in the United States. These
companies signed the MO A on December 15, 2003 and have indicated to EPA that they
no longer use diesel fuel as a hydraulic fracturing fluid additive when injecting into
USDWs.
ES-1 How Does CBM Play a Role in the Nation's Energy Demands?
CBM production began as a safety measure in underground coalmines to reduce the
explosion hazard posed by methane gas (Elder and Deul, 1974). In 1980, the U.S.
Congress enacted a tax credit for non-conventional fuels production, including CBM
production, as part of the Crude Oil Windfall Profit Act. In 1984, there were very few
CBM wells in the U.S.; by 1990, there were almost 8,000 CBM wells (Pashin and
Hinkle, 1997). In 1996, CBM production in 12 states totaled about 1,252 billion cubic
feet, accounting for approximately 7 percent of U.S. gas production (U.S. Department of
Energy, 1999). At the end of 2000, CBM production from 13 states totaled 1.353 trillion
cubic feet, an increase of 156 percent from 1992. During 2000, a total of 13,973 CBM
wells were in production (GTI, 2001; EPA Regional Offices, 2001). According to the
U.S. Department of Energy, natural gas demand is expected to increase at least 45
percent in the next 20 years (U.S. Department of Energy, 1999). The rate of CBM
production is expected to increase in response to the growing demand.
In evaluating CBM production and hydraulic fracturing activities, EPA reviewed the
geology of 11 major coal basins throughout the United States (Figure ES-1). The basins
shown in red have the highest CBM production volumes. They are the Powder River
Basin in Wyoming and Montana, the San Juan Basin in Colorado and New Mexico, and
the Black Warrior Basin in Alabama. Hydraulic fracturing is or has been used to
stimulate CBM wells in all basins, but it has not frequently been used in the Powder
River, Sand Wash, or Pacific Coal Basins. Table ES-1 provides production statistics for
2000 and information on hydraulic fracturing activity for each of the 11 basins in 2000.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-2
image:
EPA816-R-04-003
Figure ES-1. Major United States Coal Basins
Table ES-1. Coal Basins Production Statistics and Activity Information in the U.S.
Basin
Powder River
Black Warrior
San Juan
Central Appalachian
Raton Basin
Uinta
Western Interior
Northern Appalachian
Piceance
Pacific Coal
Sand Wash
Number of CBM
Producing Wells
(Year 2000)*
4,200
3,086
3,051
1,924
614
494
420
134
50
0
0
Production of CBM
in Billions of Cubic
Feet (Year 2000)*
147
112
925
52.9
30.8
75.7
6.5
1.41
1.2
0
0
Does Hydraulic
Fracturing Occur?
Yes (but infrequently)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes (but infrequently)
Yes (but infrequently)
* Data provided by the Gas Technology Institute and EPA Regional Offices. Production figures include CBM
extracted using hydraulic fracturing and other processes.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-3
image:
EPA816-R-04-003
ES-2 What Is Hydraulic Fracturing?
CBM gas is not structurally trapped in the natural fractures in coalbeds. Rather, most of
the methane is adsorbed to the coal (Koenig, 1989; Winston, 1990; Close, 1993). To
extract the CBM, a production well is drilled through the rock layers to intersect the coal
seam that contains the CBM. Next, fractures are created or existing fractures are
enlarged in the coal seam through which the CBM can be drawn to the well and pumped
to the surface.
Figure ES-2 illustrates what occurs in the subsurface during a typical hydraulic fracturing
event. This diagram shows the initial fracture creation, fracture propagation, proppant
placement, and the subsequent fracturing fluid recovery/groundwater extraction stage of
the CBM production process. The actual extraction of CBM generally begins after a
period of fluid recovery/groundwater extraction. The hydraulically created fracture acts
as a conduit in the rock or coal formation, allowing the CBM to flow more freely from
the coal seams, through the fracture system, and to the production well where the gas is
pumped to the surface.
To create or enlarge fractures, a thick fluid, typically water-based, is pumped into the
coal seam at a gradually increasing rate and pressure. Eventually the coal seam is unable
to accommodate the fracturing fluid as quickly as it is injected. When this occurs, the
pressure is high enough that the coal fractures along existing weaknesses within the coal
(steps 1 and 2 of Figure ES-1). Along with the fracturing fluids, sand (or some other
propping agent or "proppant") is pumped into the fracture so that the fracture remains
"propped" open even after the high fracturing pressures have been released. The
resulting proppant-containing fracture serves as a conduit through which fracturing fluids
and groundwater can more easily be pumped from the coal seam (step 3 of Fig. ES-1).
To initiate CBM production, groundwater and some of the injected fracturing fluids are
pumped out (or "produced" in the industry terminology) from the fracture system in the
coal seam (step 4 of Figure ES-1). As pumping continues, the pressure eventually
decreases enough so that methane desorbs from the coal, flows toward, and is extracted
through the production well (step 5 of Figure ES-1). In contrast to conventional gas
production, the amount of water extracted declines proportionally with increasing CBM
production. In some basins, huge volumes of groundwater are extracted from the
production well to facilitate the production of CBM.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-4
image:
EPA816-R-04-003
Figure ES-2. A Graphical Representation of the Hydraulic Fracturing Process in
Coalbed Methane Wells
-3* w-air
- ^Klfj caimec a pwiBilr* t-pl'jup in:* cw*n*
Hi* tndun awaf Tam HM VMB
dtmnttri Ml t=iir
2. hluld iralny rrl,j™«! in ffi« dr»dtan ct it*
3,
jiltod fun} can>(n j a prrqipant (typcaly land}
n intTiijfj*d rta VM 'nnraffcai to prep
•ir.afe
or rncjoemrl Frappanl ramaiM. In ttw *rarhj™. along YVttti surrt
. Water Is utoo cmtrad^d b raduca tha tv^drnwalt
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-5
image:
EPA816-R-04-003
Figure ES-2. A Graphical Representation of the Hydraulic Fracturing Process in
Coalbed Methane Wells (Continued)
Extraction
Mriharw Production
5. The HmfcaeK pr««4 MilUBiw f»& tactHur*
at training fluds and tie withcfrnwnJ of
Brand WMM eui o' ihe tamaikih
ij.f- •,-.,• ,.-, f.ji-....-..
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-6
image:
EPA816-R-04-003
ES-3 Why Did EPA Evaluate Hydraulic Fracturing?
SDWA requires EPA and EPA-authorized states to have effective programs to prevent
underground injection of fluids from endangering USDWs (42 U.S.C. 300h et seq.).
Underground injection is the subsurface emplacement of fluids through a well bore (42
U.S.C. 300h(d)(l)). Underground injection endangers drinking water sources if it may
result in the presence of any contaminant in underground water which supplies or can
reasonably be expected to supply any public water system, and if the presence of such a
contaminant may result in such system's noncompliance with any national primary
drinking water regulation (i.e., maximum contaminant levels (MCLs)) or may otherwise
adversely affect the health of persons (42 U.S.C. 300h(d)(2)). SDWA's regulatory
authority covers underground injection practices, but the Act does not grant authority for
EPA to regulate oil and gas production.
In 1997, the Eleventh Circuit Court ruled, in LEAF v. EPA [LEAF v. EPA, 118F.3d 1467
(11th Circuit Court of Appeals, 1997)], that because hydraulic fracturing of coalbeds to
produce methane is a form of underground injection, Alabama's EPA-approved
Underground Injection Control (UIC) Program must effectively regulate this practice. In
the wake of the Eleventh Circuit's decision, EPA decided to assess the potential for
hydraulic fracturing of CBM wells to contaminate USDWs. EPA's decision to conduct
this study was also based on concerns voiced by individuals who may be affected by
CBM development, Congressional interest, and the need for additional information
before EPA could make any further regulatory or policy decisions regarding hydraulic
fracturing.
The Phase I study is tightly focused to address hydraulic fracturing of CBM wells and
does not include other hydraulic fracturing practices (e.g., those for petroleum-based oil
and gas production) because: (1) CBM wells tend to be shallower and closer to USDWs
than conventional oil and gas production wells; (2) EPA has not heard concerns from
citizens regarding any other type of hydraulic fracturing; and (3) the Eleventh Circuit
litigation concerned hydraulic fracturing in connection with CBM production. The study
also does not address potential impacts of non-injection related CBM production
activities, such as impacts from groundwater removal or production water discharge.
EPA did identify, as part of the fact-finding process, citizen concerns regarding
groundwater removal and production water.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-7
image:
EPA816-R-04-003
ES-4 What Was EPA's Project Approach?
Based on public input, EPA decided to carry out this study in discrete phases to better
define its scope and to determine if additional study is needed after assessing the results
of the preliminary phase(s). EPA designed the study to have three possible phases,
narrowing the focus from general to more specific as findings warrant. This report
describes the findings from Phase I of the study. The goal of EPA's hydraulic fracturing
Phase I study was to assess the potential for contamination of USDWs due to the
injection of hydraulic fracturing fluids into CBM wells and to determine based on these
findings, whether further study is warranted.
Phase lisa fact-finding effort based primarily on existing literature. EPA reviewed
water quality incidents potentially associated with CBM hydraulic fracturing, and
evaluated the theoretical potential for CBM hydraulic fracturing to affect USDWs. EPA
researched over 200 peer-reviewed publications, interviewed approximately 50
employees from industry and state or local government agencies, and communicated with
approximately 40 citizens and groups who are concerned that CBM production affected
their drinking water wells.
For the purposes of this study, EPA assessed USDW impacts by the presence or absence
of documented drinking water well contamination cases caused by CBM hydraulic
fracturing, clear and immediate contamination threats to drinking water wells from CBM
hydraulic fracturing, and the potential for CBM hydraulic fracturing to result in USDW
contamination based on two possible mechanisms as follows:
1. The direct injection of fracturing fluids into a USDW in which the coal is
located (Figure ES-3), or injection of fracturing fluids into a coal seam that is
already in hydraulic communication with a USDW (e.g., through a natural
fracture system).
2. The creation of a hydraulic connection between the coalbed formation and an
adjacent USDW (Figure ES-4).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-8
image:
EPA816-R-04-003
Figure ES-3. Hypothetical Mechanisms - Direct Fluid Injection into a USDW
(Where Coal Lies Within a USDW or USDWs)
I >: lr|«l*d into Ctnibed SEDITII
:»tHd««n l Cmnmamn-ft* /\
•
Frvrturc Oeated
•. •
.
•Sanw F*ilit Staindid IXrlng Productoi
Siandac Fluid Mtfucfi Pcil.Pr«Jj:tKifi
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-9
image:
EPA816-R-04-003
Figure ES-4. Hypothetical Mechanisms - Fracture Creates Connection to USDW
aup i
£fattj'B fluid n Infected imc Ccalbed Sown:
SmnS«i fluid Migration r Cw* Fcumiikpi ind USBV PoU-ProUurfwi
ES-5 How Do Fractures Grow?
In many CBM-producing regions, the target coalbeds occur within USDWs, and the
fracturing process injects "stimulation" fluids directly into the USDWs. In other
production regions, target coalbeds are adjacent to the USDWs (i.e., either higher or
lower in the geologic section). Because shorter fractures are less likely to extend into a
USDW or connect with natural fracture systems that may transport fluids to a USDW, the
extent to which fractures propagate vertically influences whether hydraulic fracturing
fluids could potentially affect USDWs.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-10
image:
EPA816-R-04-003
The extent of the fractures is difficult to predict because it is controlled by the
characteristics of the geologic formation (including the presence of natural fractures), the
fracturing fluid used, the pumping pressure, and the depth at which the fracturing is being
performed. Fracture behavior through coals, shales, and other geologic strata commonly
present in coal zones depends on site-specific factors such as the relative thickness and
in-situ stress differences between the target coal seam(s) and the surrounding geologic
strata, as well as the presence of pre-existing natural fractures. Often, a high stress
contrast between adjacent geologic strata results in a barrier to fracture propagation. An
example of this would be where there is a geologic contact between a coalbed and an
overlying, thick, higher-stress shale.
Another factor controlling fracture height can be the highly cleated nature of some
coalbeds. In some cases, highly cleated coal seams will prevent fractures from growing
vertically. When the fracturing fluid enters the coal seam, it is contained within the coal
seam's dense system of cleats and the growth of the hydraulic fracture will be limited to
the coal seam (see Appendix A).
Deep vertical fractures can propagate vertically to shallower depths and develop a
horizontal component (Nielsen and Hansen, 1987, as cited in Appendix A: DOE,
Hydraulic Fracturing). In the formation of these "T-fractures," the fracture tip may fill
with coal fines or intercept a zone of stress contrast, causing the fracture to turn and
develop horizontally, sometimes at the contact of the coalbed and an overlying formation.
(Jones et al., 1987; Morales et al., 1990). For cases where hydraulically induced
fractures penetrate into, or sometimes through, formations overlying coalbeds, they are
most often attributed to the existence of pre-existing natural fractures or thinly inter-
bedded layering.
ES-6 What Is in Hydraulic Fracturing Fluids?
Fracturing fluids consist primarily of water or inert foam of nitrogen or carbon dioxide.
Other constituents can be added to fluids to improve their performance in optimizing
fracture growth. Components of fracturing fluids are stored and mixed on-site. Figures
ES-5 and ES-6 show fluids stored in tanks at CBM well locations.
During a hydraulic fracturing job, water and any other additives are pumped from the
storage tanks to a manifold system placed on the production wells where they are mixed
and then injected under high pressure into the coal formation (Figure ES-6). The
hydraulic fracturing in CBM wells may require from 50,000 to 350,000 gallons of
fracturing fluids, and from 75,000 to 320,000 pounds of sand as proppant (Holditch et al.,
1988 and 1989; Jeu et al., 1988; Hinkel et al., 1991; Holditch, 1993; Palmer et al., 1991,
1993a, and 1993b). More typical injection volumes, based on average injection volume
data provided by Halliburton for six basins, indicate a maximum average injection
volume of 150,000 gallons of fracturing fluids per well, with a median average injection
volume of 57,500 gallons per well (Halliburton, Inc., 2003).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-11
image:
EPA816-R-04-003
Figure ES-5. Water used for the fracturing fluid is stored on-site in large, upright
storage tanks and in truck-mounted tanks.
EPA reviewed
material safety
data sheets to
determine the
types of additives
that may be
present in
fracturing fluids.
Water or nitrogen
foam frequently
constitutes the
solute in
fracturing fluids
used for CBM
stimulation. Other components of fracturing fluids contain benign ingredients, but in
some cases, there are additives with constituents of potential concern. Because much
more gel can be dissolved in diesel fuel as compared to water, the use of diesel fuel
increases the efficiency in transporting proppant in the fracturing fluids. Diesel fuel is
the additive of greatest concern because it introduces BTEX compounds, which are
regulated by SDWA.
A thorough discussion of fracturing fluid components and fluid movement is presented in
Chapter 4.
Figure ES-6. The fracturing fluids, additives, and proppant are pumped from the
storage tanks to a manifold system placed on the wellhead where they are mixed
just prior to injection.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
ES-12
image:
EPA816-R-04-003
ES-7 Are Coalbeds Located within USDWs?
EPA reviewed information on 11 major coal basins to determine if coalbeds are co-
located with USDWs and to understand the CBM activity in the area. If coalbeds are
located within USDWs, then any fracturing fluids injected into coalbeds have the
potential to contaminate the USDW. As described previously, a USDW is not
necessarily currently used for drinking water and may contain groundwater unsuitable for
drinking without treatment. EPA found that 10 of the 11 basins may lie, at least in part,
within USDWs. Table ES-2 identifies coalbed basin locations in relation to USDWs and
summarizes evidence used as the basis for the conclusions.
ES-8 Did EPA Find Any Cases of Contaminated Drinking Water Wells Caused by
Hydraulic Fracturing in CBM Wells?
EPA did not find confirmed evidence that drinking water wells have been contaminated
by hydraulic fracturing fluid injection into CBM wells. EPA reviewed studies and
follow-up investigations conducted by state agencies in response to citizen reports that
CBM production resulted in water quality and quantity incidents. In addition, EPA
received reports from concerned citizens in each area with significant CBM development.
These complaints pertained to the following basins:
• San Juan Basin (Colorado and New Mexico);
• Powder River Basin (Wyoming and Montana);
• Black Warrior Basin (Alabama); and
• Central Appalachian Basin (Virginia and West Virginia).
Examples of concerns and claims raised by citizens include:
• Drinking water with strong, unpleasant taste and odor.
• Impacts on fish, and surrounding vegetation and wildlife.
• Loss of water in wells and aquifers, and discharged water creating artificial
ponds and swamps not indigenous to region.
Water quantity complaints were the most predominant cause for complaint by private
well owners. After reviewing data and incident reports provided by states, EPA sees no
conclusive evidence that water quality degradation in USDWs is a direct result of
injection of hydraulic fracturing fluids into CBM wells and subsequent underground
movement of these fluids. Several other factors may contribute to groundwater
problems, such as various aspects of resource development, naturally occurring
conditions, population growth, and historical well-completion or abandonment practices.
Many of the incidents that were reported (such as water loss and impacts on nearby flora
and fauna from discharge of produced water) are beyond the authorities of EPA under
SDWA and the scope of Phase I of this study.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-13
image:
O o W
o ^ ^
gL O EL
cT 3. C
d> a &3
g||
s^ a
§ ^
2. &
en £2
o
o
^
Table E&2, Evidence in Support of CdaKTSDW Co-Location in lr.S, Coal
W
Basin
San Juan
Black Warner
Piceance
Powder Rrv*r
Centra]
Appalachian
found
wittiin
USOWtt
Yes
Yos
t.ir*kelv
Likuly
LiMy
Explanation mUor widtnet
A Imge area ril the l-fuiljand system produces water onntwiirig less- thai 1CMKM) mgl. lotel cfcss-rilved sriln1
(IDS), the water quality cnierron for a USDW Analyses taker* from 3 sefeded coal well area f 16 of 27 wells)
£ilcnv that pi (Mints wate dLiiiteiiirai tess Mian 1H,U(H1 niijA IDS {Kacw?* et al , 19*l4'i
Some portions of the Fottsville Fii«natiop oontatn waters Jhat meet the qualrtv crterra of less than 10,000 rngA.
lt)S kw a I ftilJVV A«xndirM] la- HIH AJafiwriB Oil anil Gas Hoaid, suinw wJ^Hrs- in lli« l-'ullsvilte FofiiiHliun du
iKrf moot the definition of a IJSOW and two TDS considerably higher than 10.000 mg.4. (Alabama Ol
and Board, 200?) In the early 1990s, several authors mpoifed fresh water production from ooalhed
al ialHS Lq] lo HU (paium |j« rriinilH (jit f^-isliiii el al, 1W1. f-lard e1 rf . 1397)
The CBW pf oiljcaig Canieo Coal and the lower system in Hie Green Rwar Foimation are more
than 6.000 feel apart. The ooal zorw, lies a) great cteplh, roughly 6,000 feet below the ground surface n a
orlinn i"l tts? IIHSIII (I ylei tA nl . 1P9€) A mrnp-nsrlfi walw qu^ily samffc taken from ^.637 tti B,<f!Hj
p within the Cam^o Curt Zone m Ihe yVilltanis Fo-rk Fomiatiun exhibited a TDS level of 15.500 mgA.
(Graham, 2t){it) The produced water from COM extraction in the Ptoeanoj Basin is of such kw qualify Itiat il
rnnsl hi* flisposfl-rj nJn Hvapc'taiinii rxmds, rH-iii|eiiHd siln Itit? lofinalinn tr(in wtiidi il name, or fH-«i|<s*d!et! at
even greater depths (Tessm. 2001}
I he walfir quatty in Hie r-erron and F-Ia«±hi»wk varies y eally willi kirsihcin, eaiii twvjitg I |]Ji Iwels i:«lnw and
above 10.000 rngi'l (Utah DepaTment of Natural Resources. 2002)
A lopurT pr»>par^d tiy ttnj United iSlafo G<?oli>3ii';al Sur^ny ^i.HsGSj s'lowi-d 'hal satti|s"i"?i ot wdtur •» pn-duojil
4-niri 4/ f,LIW w*f?s in Hie PfAvcx-r |4i%'W Rasin aSI had TDK M^-.-els -:if less, Ihan 10.mi'..' rn^-l <NJi>i ^1 al ?IMKI)
file wrf.Mi prmlui^il liv CHM •v-f-lls in Ihc MtXiVdfjf Kr.-^ (Ji.;al |-IM|I{ oiftuiKiiily iv^uls (t"inking 'rtah?i nf
*i y*J pruductMi A'altTs vui^h jy ;!K^ iia-.-i; NJI-H prupC'V.-d a^ n sep**afu ur Mi|^>li'iii>'»ilal suufo.1 fur
ilrinkni'1 vvaftti in SOITI- arc-as (DeFMurf ^ al ?nOt))
Ot'p'hs id D:.IH| i|nnips ar^ ^vinr.iilHnl with
fit al IHflB. vVilbuii j'ont. l\ibli!
ttirif orivar».' ®>4$. in -Arqiniri an..
i w.'-Jtnr in ,;J lensl Iwn :il t|i»» sfnlns «J('il
ti. I'SGbi 19TJ) Ar
*' soams (Wils-iin 2oi'i I,
-in I'IH nvH'^ill ha»n
i.'cdutal intunnati'jfi
VDMML ?nol!
image:
o g, w
§. O £i
5s 3. C
fu a £5
P. R- C.
^> CS" O
o & e^
s'l a
3. GO
2> 8
en
Basin
Northern
Appalachian
W«$ttm
Interior:
Arfroma
Forest City
Ralon
SaruJ Wash
AT*
found
within
USDW5?
Yfis
Yes (in
Arkansas)
{in
YRS
Yes
M^fcr «wid*nc«
The depth of each coal group within (tie basin is coincident the depths of iJSDWs (Kelafant et al, 1988,
Ptatt. 2001, Foster. 1960. Hopkins. 1»6, USGS, 1973. and Stein, 1070, IJ8G8, 19H. Diagon. 1&85;
Water tpjaliy clata hum eigli listen it; Northern Appalachian dial Basin propels show t'DS levels below
10,000 mg/L {Zebrowit et al ,,1991),
The depths of coaiheds within Arkansas are conradent wth deplhs to fresh water (Andrews ^ al, 1998.,
GjfdciM'a, 1%3. FntMlman, 19432, Quarterly Rwxnv. 1993-) OaserJ on prowitted by Itie
C«'|,Mjiation Cutrniissiofi (OfJC) stiowiig depths of tfw 10,00(1 iiMji'L. ] OS gruundv/uter qualify buuiKJijry lit
Okiafiatna. tiie locatitiii of well* and USDWs would mosl no! coinoidte in thai state This is based
on depths In noals typioallv pr^^^f 'hai* 1,000 feet (Anrfrev/a el si , I'WS) and depths So- !h* base n-l ihfi (ISDVV
typically less Iftan 900 feet (OCC Dop«h to Base of treatable Water Map Sones. 2001}
Ttw depths of coatb«Js in Kansas are coincident v/ilh depths to fresh waler fdnarterly Review. 1993.
OAHC, ?CKI1a)
Tte thinness of the aqutfor suggasts mat Uiefe is separation from Ihe deepe* withal Hie
basin (Bo54ioR( al , 1W11, DASC, ConAaand Heed. 19BQ1 t-lawRfrtay «l ai , 1»ei
iV;«t«r qualityrf-s.1 ilts lf-.''in QFM w^lls in fh»? Ratnn E1as4fi f|i-mf!nslratn TDS ovnlMfit "-f IHAS than t;i(Min n>p.1
N'.-any dll w«H>f sut-oyt^t <-ln?w d T PS ul Icsis ftlaci ? .'i<nl ino-t, arid niun- tlldii hair tiad T[iS uf ks> tllc*ri 1 .Ofm
f wo yas uuiiipaniws |.f uduojd watei ^rurri ix-rfs tll^il ytMjwi-d 1 U!S k^-yls b'.HoA1 V1 U'">U iiicj,'L Al Craig Dorttu ill
Mf/llftf C^'jiirtv (luokrHll rilfjfirpnrrjl«iiii dnllHff Ifi f,F?t/ v/flls 'CIH v.'ffSs yi-U1i:ff UirfjM 'dlumes ol fresh wfller
™lh rt)S *-1 '"I* "i nnj.'l I'fkJi.jHdiJ f '«d arwl (];-j« '>:iiiirinssi'<.iii. ?iKH) I ij^loi was '..ip'fialnw.] 11 wells al or jy
ChurfkHe r«fdi vVatec purripi-Kl In'.rn IIIH wHIs cfinlanurt 1 f:in; n\f]!\ tif I Lift ;>nfi ".vs.?. ifa^figrgecl to Ihe
yruLMtl uihk'i a Nt^MUil f-'ulliilKjn Lii*M^iaf•-}•.' i'limnatiO'i 8v^tw?i sNIJOi H) pi'nv'l O-Juailcfl/ ROVIQW. 1993)
firfrta fr. iin a f^CM s-lud^ iHiiun^tratiiS "Hf -TI Jij'^lmn »\ a - m-al si!am nrnl a i iSRW in P«ijr«» Gourt>- Walor
qudS-ly infuiriiatiun 1iuni tout yas li^l w^lb indif^tes ID8 lu- frfy t«jtxv^.'ii 1 '.vJi:- ail*J 1 <j(i':i iny/L. well tfle
1fi nn*i mg'l :t!l*:rn'ai (W\t>n WA'< Wells m (tie t'l.isriaN aimmncilv V11'!lf* 1-'1' lf= •^ll11-.' (iftBnns per minute-
I fis> t-'-isK 1*1 frm ivsvi-r pr'rfkf>,"d CIHIKTHIIV iriiuj^ 'f'"--*1 ^rjl11" 'll|i:i fH'l (Pmn, t^t>4i
w
oo
ON
image:
EPA 816-R-04-003 Executive Summary
ES-9 What Are EPA's Conclusions?
Based on the information collected and reviewed, EPA has determined that the injection
of hydraulic fracturing fluids into CBM wells poses little or no threat to USDWs.
Continued investigation under a Phase II study is not warranted at this time.
As proposed in the Final Study Design (April 2001), Phase I of the study was a limited-
scope assessment in which EPA would:
• Gather existing information to review hydraulic fracturing processes,
practices, and settings;
• Request public comment to identify incidents that have not been reported to
EPA;
• Review reported incidents of groundwater contamination and any follow-up
actions or investigations by other parties (state or local agencies, industry,
academia, etc.); and,
• Make a determination regarding whether further investigation is needed,
based on the analysis of information gathered through the Phase I effort.
EPA's approach for evaluating the potential threat to USDWs was an extensive
information collection and review of empirical and theoretical data. EPA reviewed
incidents of drinking water well contamination believed to be associated with hydraulic
fracturing and found no confirmed cases that are linked to fracturing fluid injection into
CBM wells or subsequent underground movement of fracturing fluids. Although
thousands of CBM wells are fractured annually, EPA did not find confirmed evidence
that drinking water wells have been contaminated by hydraulic fracturing fluid injection
into CBM wells.
EPA also evaluated the theoretical potential for hydraulic fracturing to affect USDWs
through one of two mechanisms:
1. Direct injection of fracturing fluids into a USDW in which the coal is located,
or injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
Regarding the question of injection of fracturing fluids directly into USDWs, EPA
considered the nature of fracturing fluids and whether or not coal seams are co-located
with USDWs. Potentially hazardous chemicals may be introduced into USDWs when
fracturing fluids are used in operations targeting coal seams that lie within USDWs. In
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-16
image:
EPA816-R-04-003 Executive Summary
particular, diesel fuel contains BTEX compounds, which are regulated under SDWA.
However, the threat posed to USDWs by the introduction of some fracturing fluid
constituents is reduced significantly by the removal of large quantities of groundwater
(and injected fracturing fluids) soon after a well has been hydraulically fractured. In fact,
CBM production is dependent on the removal of large quantities of groundwater. EPA
believes that this groundwater production, combined with the mitigating effects of
dilution and dispersion, adsorption, and potentially biodegradation, minimize the
possibility that chemicals included in the fracturing fluids would adversely affect
USDWs.
Because of the potential for diesel fuel to be introduced into USDWs, EPA requested,
and the three major service companies agreed to, the elimination of diesel fuel from
hydraulic fracturing fluids that are injected directly into USDWs for CBM production
(USEPA, 2003). Industry representatives estimate that these three companies perform
approximately 95 percent of the hydraulic fracturing projects in the United States.
In evaluating the second mechanism, EPA considered the possibility that hydraulic
fracturing could cause the creation of a hydraulic connection to an adjacent USDW. The
low permeability of relatively unfractured shale may help to protect USDWs from being
affected by hydraulic fracturing fluids in some basins. If sufficiently thick and relatively
unfractured shale is present, it may act as a barrier not only to fracture height growth, but
also to fluid movement. Shale's ability to act as a barrier to fracture height growth is
primarily due to the stress contrast between the coalbed and the shale. Another factor
controlling fracture height can be the highly cleated nature of some coalbeds. In some
cases, when the fracturing fluid enters the coal seam, it is contained within the coal
seam's dense system of cleats and the growth of the hydraulic fracture will be limited to
the coal seam (see Appendix A).
Some studies that allow direct observation of fractures (i.e., mined-through studies)
indicate many fractures that penetrate into, or sometimes through, one or more
formations overlying coalbeds can be attributed to the existence of pre-existing natural
fractures. However, given the concentrations and flowback of injected fluids, and the
mitigating effects of dilution and dispersion, adsorption, and potentially biodegradation,
EPA does not believe that possible hydraulic connections under these circumstances
represent a significant potential threat to USDWs.
It is important to note that states with primary enforcement authority (primacy) for their
UIC Programs implement and enforce their regulations, and have the authority under
SDWA to place additional controls on any injection activities that may threaten USDWs.
States may also have additional authorities by which they can regulate hydraulic
fracturing. With the expected increase in CBM production, the Agency is committed to
working with states to monitor this issue.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-17
image:
EPA 816-R-04-003 Executive Summary
REFERENCES
Alabama Oil and Gas Board. 2002. Public Comment OW-2001-0002-0029 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Andrews, R.D., BJ. Cardott, and T. Storm. 1998. The Hartshorne Play in Southeastern
Oklahoma: regional and detailed sandstone reservoir analysis and coalbed-
methane resources. Oklahoma Geological Survey, Special Publication 98-7.
Bostic, J.L., L.L. Brady, M.R. Howes, R.R. Burchett, and B.S. Pierce. 1993.
Investigation of the coal properties and the potential for coal-bed methane in the
Forest City Basin. US Geological Survey, Open File Report 93-576.
Close, Jay. C. 1993. Natural Fractures in Coal; Chapter 5 of AAPG Studies in Geology
38, "Hydrocarbons from Coal", pp. 119-133.
Colorado Oil and Gas Conservation Commission. 2001. http://www.oil-gas.state.co.us/
Condra, G.E. and B.C. Reed. 1959. The geological section of Nebraska. Nebraska
Geological Survey Bulletin 14A, 1959.
Cordova, R.M. 1963. Reconnaissance of the ground-water resources of the Arkansas
Valley Region, Arkansas. Contributions to the Hydrology of the United States,
Geological Survey Water-Supply Paper 1669-BB, 1963.
DASC website. 2001a. Kansas elevation map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html
DASC website. 2001b. Ozark Aquifer base map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html
DeBruin, R.H., R.M. Lyman, R.W. Jones, and L.W. Cook. 2000. Information Pamphlet
7. Wyoming State Geological Survey.
Dion, N.P. 1984. Washington Ground-Water Resources. In National Water Summary,
US Geological Survey Water-Supply Paper No. 2275, pp. 433-438.
Duigon, M.T. and M.J. Smigaj. 1985. First report on the hydrologic effects of
underground coal mining in Southern Garrett County, Maryland, US Geological
Survey Report of Investigations No. 41.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-18
image:
EPA816-R-04-003 Executive Summary
Elder, C.H. and M. Deul. 1974. Degasification of the Mary Lee coalbed near Oak
Grove, Jefferson county, Alabama, by vertical borehole in advance of mining; US
Bureau of Mines Report 7968.
Ellard, J.S., R.P. Roark, and W.B. Ayers. 1992. Geologic controls on coalbed methane
production: an example from the Pottsville formation, Black Warrior Basin,
Alabama USA. Symposium on Coalbed Methane Research and Development in
Australia. James Cook University, p. 45-61.
Eleventh Circuit Court of Appeals, 1997. LEAF v. EPA, 118F.3d 1467.
Flowerday, C.F., R.D. Kuzelka, andD.T. Pederson, compilers. 1998. The Ground
Water Atlas of Nebraska.
Foster, J.B. 1980. Fresh and saline ground-water map of West Virginia. US Geological
Survey, West Virginia Geological and Economic Survey, Map WV-12.
Friedman, S.A. 1982. Determination of reserves of methane from coalbeds for use in
rural communities in eastern Oklahoma. Oklahoma Geological Survey, Special
Publication 82-3, 1982.
Gas Technology Institute (GTI). 2001. Personal communication with GTI staff.
Graham, G. 2001. Colorado Division of Water Resources, personal communication with
staff.
Halliburton, Inc. 2003. Personal communication with Halliburton staff, fracturing fluid
expert, Steve Almond. April 2003.
Hinkel, J.J., K.H. Nimerick, K. England, J.C. Norton, and M. Roy. 1991, Design and
evaluation of stimulation and workover treatments in coal seam reservoirs;
Proceedings 1991 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), Tuscaloosa, p. 453-458.
Holditch, S.A., J.W. Ely, M.E. Semmelbeck, R.H. Carter, J. Hinkle, and R.G. Jeffrey.
1988. Enhanced recovery of coalbed methane through hydraulic fracturing; SPE
Paper 18250, Proceedings 1988 SPE Annual Technical Conference and
Exhibition (Production Operations and Engineering), p. 689.
Holditch, S.A., J.W. Ely, and R.H. Carter. 1989. Development of a coal seam fracture
design manual; Proceedings, 1989 Coalbed Methane Symposium, Tuscaloosa,
Alabama, pp. 299-320.
Holditch, S.A., 1993, Completion methods in coal-seam reservoirs; Journal of Petroleum
Technology, v.45 n.3 (March 1993), pp. 270-276.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-19
image:
EPA816-R-04-003 Executive Summary
Hopkins, Herbert T. 1966. Fresh-saline water interface map of Kentucky. US
Geological Survey, Kentucky Geological Survey, Series X.
Jeu, S.J., T.L. Logan, and R.A. McBane. 1988, Exploitation of deeply buried coalbed
methane using different hydraulic fracturing techniques; SPE paper 18253,
Proceedings 63rd Annual Technical Conference (Houston).
Jones, A.H., Bell, G.J., and Morales, R.H. 1987. Examination of potential mechanisms
responsible for the high treatment pressures observed during stimulation of
coalbed reservoirs; SPE Paper 16421, Proceedings, Department of Energy/SPE
Symposium: Gas from Low Permeability Reservoirs, p. 317.
Kaiser, W.R., Swartz, T.E., and Hawkins, GJ. 1994. Hydrologic framework of the
Fruitland formation, San Juan Basin. New Mexico Bureau of Mines and Minerals
Bulletin 146: Coalbed methane in the upper Cretaceous Fruitland formation, San
Juan Basin, New Mexico and Colorado, pp. 133-164.
Kelafant, J.R., D.E. Wicks, and V.A. Kuuskraa. March 1988. A geologic assessment of
natural gas from coal seams in the Northern Appalachian Coal Basin. Topical
Report - Final Geologic Report (September 1986 - September 1987).
Macfarlane, A. 2001. Kansas Geological Survey, personal communication.
Morales, R,H, McLennan, J.D., Jones, A.H., and Schraufnagel, R.A. 1990.
Classification of treating pressures in coal fracturing; Proceedings of the 31st U.S.
Symposium on Rock Mechanics, 31, pp. 687-694.
National Water Summary. 1984. Hydrologic events, selected water-quality trends, and
ground-water resources. United States Geological Survey Water-Supply Paper
No. 2275.
Nielsen, P. E. and Hanson, M. E. 1987. Analysis and Implications of Three Fracture
Treatments in Coals at the USX Rock Creek Site Near Birmingham, Alabama,
1987 Coalbed Methane Symposium, Tuscaloosa, AL (Nov. 16-19, 1987).
OCC (Oklahoma Corporation Commission), Depth to Base of Treatable Water Map
Series, 2001.
Palmer, ID., N.S. King, and D.P. Sparks. 1991. The character of coal fracture
treatments in Oak Grove field, Black Warrior basin, SPE paper no. 22914,
Proceedings, 1991 Society of Petroleum Engineers annual technical conference
and exhibition, pp.277-286.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-20
image:
EPA816-R-04-003 Executive Summary
Palmer, ID., N.S. King, and D.P. Sparks. 1993a. The character of coal fracture
treatments in the Oak Grove field, Black Warrior basin; In Situ, Journal of Coal
Research, v.l7 (3), pp. 273-309.
Palmer, ID., S.W. Lambert, and J.L. Spitler. 1993b. Coalbed methane well completions
and stimulations. Chapter 14 of AAPG Studies in Geology 38, pp. 303-341.
Pashin, J.C. and F. Hinkle. 1997. Coalbed Methane in Alabama. Geological Survey of
Alabama Circular 192, 71pp.
Pashin, J.C., W.E. Ward, R.B. Winston, R.V. Chandler, D.E. Bolin, K.E. Richter, W.E.
Osborne, and J.C. Sarnecki. 1991. Regional analysis of the Black Creek-Cobb
coalbed methane target interval, Black Warrior Basin, Alabama. Alabama
Geological Survey Bulletin 145, 127pp.
Platt, S. January, 2001. US EPA Region 3, personal communication.
Quarterly Review. 1993. Coalbed methane-state of the industry. Methane From Coal
Seams Technology, August, 1993.
Rice, C.A., M.S. Ellis, and J.H. Bullock, Jr. 2000. Water co-produced with coalbed
methane in the Powder River Basin, Wyoming: preliminary compositional data.
US Geological Survey Open-File Report 00-372.
Sedam, A.C. and R.B. Stein. 1970. Saline ground-water resources of Ohio. Hydrologic
Investigations Atlas HA-366, Department of the Interior, US Geological Survey.
Tessin, R. 2001. Colorado Oil and Gas Conservation Commission, personal
communication.
Tyler, R., A.R. Scott, and W.R. Kaiser. 1998. Defining coalbed methane exploration
fairways: An example from the Piceance Basin, Rocky Mountain Foreland.
Western United States, Conference Document, March 23-25.
http://georef.cos.com/cgi-bin/getRec?un=2001-012340
U.S. Department of Energy. 1999. Environmental Benefits of Advanced Oil and Gas
Exploration and Production Technology, Office of Fossil Energy, p 8.
U.S. Environmental Protection Agency. 2001. Personal communication with EPA
Regional staff.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-21
image:
EPA816-R-04-003 Executive Summary
US Environmental Protection Agency. 2003. A Memorandum of Agreement Between
The United States Environmental Protection Agency And BJ Services Company,
Halliburton Energy Services, Inc., and Schlumberger Technology Corporation
Elimination of Diesel Fuel in Hydraulic Fracturing Fluids Injected into
Underground Sources of Drinking Water During Hydraulic Fracturing of Coalbed
Methane Wells, December 12, 2003.
http://www.epa.gov/safewater/uic/pdfs/moa uic hyd-fract.pdf
United States Geological Survey (USGS). 1971. State of Ohio, 1:500,000 topographic
map.
USGS. 1973. State of Kentucky, 1:500,000 topographic map. National Water
Summary. 1984. Hydrologic events, selected water-quality trends, and ground-
water resources. United States Geological Survey Water-Supply Paper No. 2275.
Utah Department of Natural Resources. 2002. Public Comment OW-2001-0002-0090
to "Draft Evaluation of Impacts to Underground Sources of Drinking Water by
Hydraulic Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol.
63, No. 185. p. 33992, September 24, 2002.
Virginia Department of Mines, Minerals, and Energy (VDMME). 2001. Personal
communication with VDMME staff.
Wilson, R. February, 2001. Director, Virginia Division of Gas & Oil, Department of
Mines, Minerals, and Energy, personal communication.
Winston, R.B. 1990. Vitrinite reflectance of Alabama's bituminous coal; Alabama
Geological Survey Circular 139, 54 pp.
Zebrowitz, M.J., J.R. Kelafant, and C.M. Boyer. 1991. Reservoir characterization and
production potential of the coal seams in Northern and Central Appalachian
Basins. Proceedings of the 1991 Coalbed Methane Symposium, The University
of Alabama/Tuscaloosa, May 13-16, 1991.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs ES-22
image:
List of Acronyms and Abbreviations
ADEM
Bbl/min
Bcf
Bgs
BHP
BLM
BTEX
Btu
CBM
CDH
CCL
CDWR
CFR
CMHPG
COGCC
DASC
DNR
DOE
EPA
g
g/mL
GRI
GTI
GSA
HC1
Micrograms per gram
Micrograms per liter
Alabama Department of Environmental Management
Barrel per minute
Billion cubic feet
Below ground surface
Bottom hole pressure
Bureau of Land Management
Benzene, toluene, ethylbenzene, xylenes
British thermal unit
Coalbed methane
Colorado Department of Health
Contaminant Candidate List
Colorado Division of Water Resources
Code of Federal Regulations
Carboxymethylhydroxypropylguar
Colorado Oil and Gas Conservation Commission
Data Access and Support Center
Department of Natural Resources
Department of Energy
Environmental Protection Agency
Gram
Grams per milliliter
Gas Research Institute
Gas Technology Institute
Geological Survey of Alabama
Hydrochloric acid
xiii
image:
HEC
HPG
KC1
L
LEAF
Mcf
MCL
mil
mg/L
mL
MOA
MSDS
MTBE
NMOCD
NPDEA
0GB
OGWDW
P3D
PAH
POM
ppm
PRBRC
PRCMIC
psi
SDWA
SEO
SJRA
TBEG
Hydroxyethylcellulose
Hydroxypropylguar
Potassium chloride
Liter
Legal Environmental Assistance Foundation
Million cubic feet
Maximum contaminant level
Millidarcy
Milligrams per liter
Milliliter
Memorandum of Agreement
Material Safety Data Sheet
Methyl tert butyl ether
New Mexico Oil Conservation Division
National Pollution Discharge Elimination System
Oil and Gas Board
Office of Ground Water and Drinking Water
Pseudo 3 Dimensional
Polynuclear aromatic hydrocarbons
Polycyclic organic matter
Parts per million
Powder River Basin Resource Council
Powder River CoalbedMethane Information Council
Pounds per square inch
Safe Drinking Water Act
State Engineer's Office
San Juan Regional Authority
Texas Bureau of Economic Geology
xiv
image:
Tcf
TDS
TGD
UIC
USBM
USDW
USGS
VDMME
wt.
Trillion cubic feet
Total dissolved solids
Tennessee Geology Division
Underground Injection Control
United States Bureau of Mines
Underground Source of Drinking Water
United States Geological Survey
Virginia Division of Oil and Gas, within the Department of Mines, Minerals and Energy
Weight
xv
image:
Glossary
Adsorption
Alluvial aquifer
Amphoteric
Anaerobic Bacteria
Anisotropic
Annulus
Anticline
Aureole
Azimuth
Bedrock aquifer
Billion cubic feet
Bio genie
Bituminous
Breaker
Breccia
Brecciated
Btu
Butt Cleat
Adhesion of gas molecules, ions or molecules in solution to the surface of solid bodies
with which they are in contact.
A water-bearing deposit of unconsolidated material (e.g., sand and gravel) left behind by
a river or other flowing water.
Having both basic and acidic properties.
Bacteria that thrive in oxygen-poor environments.
Having some physical property that varies with direction from a given location.
The space between the casing (the material that is used to keep the well stable; typically
this material is steel) in a well and the wall of the hole, or between two concentric strings
of casing, or between casing and tubing.
A fold of layered, sedimentary rocks whose core contains stratigraphically older rocks,
the shape of the fold is generally convex upward.
A ring surrounding a volcanic intrusion where the surrounding rock has been altered.
The direction of a horizontal line as measured on an imaginary horizontal circle.
An aquifer located in the solid rock underlying unconsolidated surface materials (i.e.,
sediment). Solid rock can bear water when it is fractured.
A unit typically used to define gas production volumes in the coalbed methane industry; 1
Bcfis roughly equivalent to the volume of gas required to heat approximately 12,000
households for one year (based on the Department of Energy's average household energy
consumption statistic, 2001).
A direct product of the physiological activities of organisms.
From the base word bitumen, referring to a general term for various solid and semi-solid
hydrocarbons that are able to join together and are soluble in carbon bisulfide (e.g.,
asphalts).
A fracturing fluid additive that is added to break down the viscosity of the fluid.
A coarse-grained clastic rock composed of angular broken rock fragments held together
by a mineral cement or a fine-grained matrix.
Consisting of angular fragments cemented together.
British thermal unit; a unit of measure used to define energy.
The coal cleat set that abuts into face cleats.
Capture Zone
The portion of an aquifer that contributes water to a particular pumping well.
xvi
image:
Cavitation Cycling
Cleats
CMHPG
Craton
Crosslinker
Crosslinked Gel
Darcy
Desorption
Epiclastic
Evapotranspiration
Face Cleat
Flowback
Fracture Conductivity
Geophone
Graben
Guar
HC1
HEC
Hydraulic Conductivity
Injectate
Isopach
Also known as cavity completion, an alternative completion technique to hydraulic
fracturing, in which a cavity is generated by alternately pumping in nitrogen and blowing
down pressure.
Natural fractures in coal that often occur in systematic sets, through which gas and water
can flow.
Carboxymethyl hydroxypropylguar; a form ofguargel.
A part of the earth's crust that has attained stability and has been relatively undeformed
for a long time; the term is restricted to continents, and includes both shield and
platform.
An additive that when added to a linear gel, will create a complex, high viscosity,
pseduoplastic fracturing fluid.
A gel to which a crosslinker has been added (see crosslinker).
A measure of the permeability of rock or sediment.
Liberation of tightly held methane gas molecules previously bound to the solid surface of
the coal.
Formed from the fragments or particles broken away (by weathering and erosion) from
pre-existing rocks to form an altogether new rock in a new place.
The portion of precipitation returned to the air through evaporation and transpiration.
A coal cleat set that is through-going and continuous.
The process of causing fluid to flow back to the well out of a fracture after a hydraulic
fracturing event is completed.
The capability of the fracture to conduct fluids under a given hydraulic head difference.
A seismic detector, placed on or in the ground, that responds to ground motion at its
point of location.
An elongate, down-dropped block that is bounded by nearly parallel faults on both sides.
Organic powder thickener, typically used to make viscous fracturing fluids, completely
soluble in hot and cold water, insoluble in oils, grease and hydrocarbons.
Molecular formula for hydrochloric acid; can be used in diluted form in the hydraulic
fracturing process to fracture limestone formations and to clean up perforations in
coalbed methane fracturing treatments.
Hydroxyethylcellulose; a form ofguar gel.
(see permeability)
In relation to the coalbed methane industry, this is the fracturing fluid injected into a
coalbed methane well.
A line drawn on a map through points of equal true thickness of a designated
stratigraphic unit or group of stratigraphic units.
xvii
image:
Isotopic
Isotropic
KC1
Lacustrine
Laminar Flow
Leakoff
Lenticular
Linear Gel
Lithology
Millidarcy
Mcf
mg/L
Rocks formed in the same environment, i.e. in the same sedimentary basin or geologic
province.
A medium, such as unconsolidated sediments or a rock formation, whose properties are
the same in all directions.
Molecular formula for potassium chloride.
Pertaining to, produced by, or formed in a lake or lakes.
Water flow in which the stream lines remain distinct and the flow direction at every point
remains unchanged with time; non-turbulent flow.
The magnitude of pressure exerted on a formation that causes fluid to be forced into the
formation. In common usage, leakoffis often considered the movement of fluid out of
primary fractures and into a geologic formation, either through small existing permeable
paths (connected pores and natural fracture networks) or through small
pathways created or enlarged in the rock through the fracturing process.
Pertaining to a discontinuous, lens-shaped (saucer-shaped) stratigraphic body.
A simple guar-based fracturing fluid usually formulated using guar and water with
additives or guar with dieselfuel.
The description of rocks based on mineralogic composition and texture.
The customary unit of measurement of fluid permeability; equivalent to 0.001 Darcy.
Million cubic feet; a unit typically used to define gas production volumes in the coalbed
methane industry; 1 Mcf is roughly equivalent to the volume of gas required to heat
approximately 12 households for one year (based on the Department of Energy's
average household energy consumption statistic, 2001); Mcf can sometimes represent
1,000 cubic feet.
Milligrams per liter; typically used to define concentrations of a dissolved compound in a
fluid.
Mined-through studies Mined-through studies are projects in which coalbeds have been actually mined through
(i. e., the coal has been removed) so that remaining coal and surrounding rock can be
inspected, after the coalbeds have been hydraulically fractured. These studies provide
unique subsurface access to investigate coalbeds and surrounding rock after hydraulic
fracturing.
Moduli
Overthrust
Pad
Paleochannels
Plural of modulus (often referred to as bulk modulus), the ratio of stress to strain,
abbreviated as "k". The bulk modulus is an elastic constant equal to the applied stress
divided by the ratio of the change in volume to the original volume of a body.
A low-angle thrust fault of large scale, with total displacement (lateral or vertical)
generally measured in kilometers.
An initial volume of fluid that is used to initiate and propagate a fracture before a
proppant is placed.
Old or ancient river channels preserved in the subsurface as lenticular sandstones.
XVlll
image:
Permeability
Physiographic
Play
Potentiometric
ppm
Primacy
Primary porosity
Proppant
psi
Rank
Screen-out
Secondary porosity
Semianthracite
Stratigraphy
Subbituminous
Subgraywacke
Surficial
Syncline
Tcf
The capacity of a porous rock, sediment, or soil to transmit a fluid; it is a measure of the
relative ease of fluid flow under equal pressure and from equal elevations.
A region of which all parts are similar in geologic structure and climate and which has
had a unified geomorphic history; its relief features differ significantly from those of
adjacent regions.
A productive coalbed methane formation, or a productive oil or gas deposit.
The total head of ground water, defined by the level to which water will rise in a well.
Parts per million; typically used to define concentrations of a dissolved compound in a
fluid; equivalent to 1 mg/L.
The right to self-establish, self-enforce and self-regulate environmental standards; this
enforcement responsibility is granted by EPA to States and Indian Tribes.
The porosity preserved from some time between sediment deposition and the final rock-
forming process; (e.g., the spaces between grains of sediment).
Granules of sand, ceramic or other minerals that are wedged within the fracture and act
to "prop" it open after the fluid pressure from fracture injection has dissipated.
Pounds per square inch; a unit of pressure.
The degree of metamorphism in coal; the basis of coal classification into a natural series
from lignite to anthracite.
Term used to describe a fracturing job where proppant placement has failed.
The porosity created through alteration of rock, commonly by processes such as,
dissolution and fracturing.
Term used to identify coal rank; specifically representing coal that possesses a fixed-
carbon content of 86% to 92%.
The study of rock strata; concerning all characteristics and attributes of rocks and their
interpretation in terms of mode of origin and geologic history.
A black coal, intermediate in rank between lignite and bituminous.
A sedimentary rock (sandstone) that contains less feldspar, and more and better-rounded
quartz grains than graywacke; intermediate in composition between graywacke and
orthoquartzite; it is lighter-colored and better-sorted, and has less matrix than
greywacke.
Pertaining to or lying in or on a surface; specific to the surface of the earth.
A fold of layered, sedimentary rocks whose core contains stratigraphically younger
rocks; shape of fold is generally concave upward.
Trillion cubic feet; a unit typically used to define gas production volumes in the coalbed
methane industry; 1 Tcf is roughly equivalent to the volume of gas required to heat
approximately 12 million households for one year (based on the Department of Energy's
average household energy consumption statistic, 2001).
xix
image:
Thermogenic
Toughness
Transmissivity
Up-warp
Viscosity
Volcaniclastic
A direct product of high temperatures, (e.g. Thermogenic methane).
The point at which enough stress intensity has been applied to a rock formation, so that a
fracture initiates and propagates.
A measure of the amount of water that can be transmitted horizontally through a unit
width by the full saturated thickness of the aquifer under a hydraulic gradient of one.
The uplift of a region; usually a result of the release of isostatic pressure, e.g. the melting
of an ice sheet.
The property of a substance to offer internal resistance to flow; internal friction.
Composed of fragments or particles, and related to volcanic processes either by forming
as the result of explosive processes or due to the weathering and erosion of volcanic
rocks.
XX
image:
EPA816-R-04-003 Chapter 1
Introduction
Chapter 1
Introduction
Section 1421 of SDWA tasks EPA with protecting USDWs for all current and future
drinking water supplies across the country (see section 1.3 for the complete definition of
a USDW). EPA's UIC Program is responsible for ensuring that fluids injected into the
ground (for purposes including waste disposal, oil field brine disposal, enhanced
recovery of oil and gas, mining, and emplacement of other fluids) do not endanger
USDWs.
EPA, through its UIC Program, conducted a fact-finding effort based primarily on
existing literature. The goal of this study was to assess the potential for contamination of
USDWs due to the injection of hydraulic fracturing fluids into coalbed methane wells and
to determine, based on these findings, whether further study is warranted. For the
purposes of this study, EPA assessed USDW impacts by the presence or absence of
documented drinking water well contamination cases caused by coalbed methane
hydraulic fracturing, clear and immediate contamination threats to drinking water wells
from coalbed methane hydraulic fracturing, and the potential for coalbed methane
hydraulic fracturing to result in USDW contamination based on two possible mechanisms
as follows:
1. Direct injection of fracturing fluids into a USDW in which the coal is located,
or injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
EPA obtained information for this study from literature searches, field visits, a review of
reported groundwater contamination incidents in areas where coalbed methane is
produced, and solicitation of information from the public on any impacts to groundwater
believed to be associated with hydraulic fracturing.
EPA also reviewed 11 major coal basins throughout the United States to determine if
coalbeds are co-located with USDWs and to understand the coalbed methane activity in
the area (Figure 1-1). The basins shown in red have the highest coalbed methane
production volumes. They are the Powder River Basin in Wyoming and Montana, the
San Juan Basin in Colorado and New Mexico, and the Black Warrior Basin in Alabama.
Hydraulic fracturing is or has been used to stimulate coalbed methane wells in all basins,
although it has not frequently been used in the Powder River, Sand Wash, or Pacific Coal
Basins.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-1
image:
EPA816-R-04-003
Chapter 1
Introduction
Figure 1-1. Locus Map of Major United States Coal Basins
1.1 EPA's Rationale for Conducting This Study
Although coalbed methane has many environmental advantages over traditional energy
sources, concerns have been raised regarding the environmental impacts of coalbed
methane production. Coalbed methane production in certain areas has led to
groundwater depletion and production water discharge issues (i.e., issues that are not
associated with the quality of USDWs). Citizens, state agencies, producers, and the
regional EPA offices in those areas are working in concert to better understand and
mitigate these potential problems.
This study examines the potential for hydraulic fracturing fluid injection into coalbed
methane wells to contaminate USDWs. EPA conducted this study in response to
allegations that hydraulic fracturing of coalbed methane wells has affected the quality of
groundwater (i.e., issues that are associated with the mandates of the UIC Program).
State oil and gas agencies receiving such complaints have indicated that, based on their
investigations, hydraulic fracturing of coalbed methane wells has not contributed to water
quality degradation in USDWs.
In response to an Eleventh Circuit Court of Appeals (hereafter, "the Court") decision
[LEAFv. EPA, 118F.3d 1467 (11th Cir, 1997)], the State of Alabama recently
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
1-2
image:
EPA816-R-04-003 Chapter 1
Introduction
supplemented its rules governing the hydraulic fracturing of wells to include additional
requirements to protect USDWs during the hydraulic fracturing of coalbeds for methane
production. Prior to the Court's decision, EPA had not considered hydraulic fracturing as
an underground injection activity, because the Agency did not consider production well
stimulation as an activity subject to UIC regulations. Nevertheless, the Court held that
the injection of fluids for the purpose of hydraulic fracturing constitutes underground
injection as defined under SDWA, that all underground injection must be regulated, and
that hydraulic fracturing of coalbed methane wells in Alabama must be regulated under
Alabama's UIC program.
In the wake of the Eleventh Circuit Court decision, EPA decided to assess the potential
for hydraulic fracturing fluid injection into coalbed methane wells to contaminate
USDWs. EPA's decision to conduct this study was also based on concerns voiced by
individuals who may be affected by coalbed methane development, Congressional
interest, and the need for additional information before EPA could make any further
regulatory or policy decisions regarding hydraulic fracturing.
1.2 Overview of Hydraulic Fracturing
Hydraulic fracturing is a technique used by the oil and gas industry to improve the
production efficiency of oil and coalbed methane wells. The hydraulic fracturing process
uses high hydraulic pressures to initiate a fracture. A hydraulically induced fracture acts
as a conduit in the rock or coal formation that allows the oil or coalbed methane to travel
more freely from the rock pores to the production well that can bring it to the surface.
In the case of coalbed methane gas production, the gas is not structurally "trapped" under
pressure. Rather, most of the coalbed methane is adsorbed within small pores in the
"micro-porous matrix" of the coal (Koenig, 1989; Winston, 1990; Close, 1993). When
coalbed methane production begins, water is first pumped out (or "produced" in the
industry terminology) from the fractures, joints, and cleats (i.e., tiny, disconnected
clusters of fractures) in the coal until the pressure declines to the point that methane
begins to desorb from the coal matrix itself (Gray, 1987).
To extract the coalbed methane, a production well is drilled through rock layers to
intersect the coal seam that contains the coalbed methane. Next, a fracture is created or
enlarged in the coal seam to connect the well bore to the coalbed joint/cleat system. To
create such a fracture, a thick, water-based fluid is pumped into the coal seam at a
gradually increasing rate. At a certain point, the coal seam will not be able to
accommodate the fracturing fluid as quickly as it is being injected. When this occurs, the
pressure is high enough that the coal gives way, and a fracture is created or an existing
fracture is enlarged. To hold the fracture open, a propping agent, usually sand
(commonly known as "proppant"), is pumped into the fracture so that when the pumping
pressure is released, the fracture does not close completely because the proppant is
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-3
image:
EPA816-R-04-003 Chapter 1
Introduction
"propping" it open. The resulting fracture filled with proppant becomes a conduit
through which water can flow to the production well, thus depressurizing the coal matrix,
allowing for the desorption of methane and its flow towards the production well.
The extent of the fracture in a coalbed is controlled by the characteristics of the geologic
formation (including the presence of natural fractures), the fracturing fluid used, the
pumping pressure, and the depth at which the fracturing is performed. Whether the
fracture grows taller or longer is determined by the properties of the surrounding rock. A
hydraulically created fracture will always take the path of least resistance through the
coal seam and surrounding formations.
A more comprehensive discussion of the fracturing process and the fracturing
fluids/additives used in hydraulic fracturing of coalbed methane wells is presented in
Chapters 3 and 4, respectively.
1.3 EPA's Authority to Protect Underground Sources of Drinking Water
SDWA requires EPA and EPA-authorized states to have effective programs to prevent
underground injection of fluids from endangering USDWs (42 U.S.C. 300h et seq.).
Underground injection is the subsurface emplacement of fluids through a well bore (42
U.S.C. 300h(d)(l)). Underground injection endangers drinking water sources if it may
result in the presence of any contaminant in underground water which supplies or can
reasonably be expected to supply any public water system, and if the presence of such
contaminant may result in such system's noncompliance with any national primary
drinking water regulation (i.e., maximum contaminant levels) or may otherwise adversely
affect the health of persons (42 U.S.C. 300h(d)(2)). SDWA's regulatory authority
extends to underground injection practices; SDWA does not provide a general grant of
authority for EPA to regulate oil and gas production.
A USDW is defined in the UIC regulations at 40 CFR 144.3 as an aquifer or a portion of
an aquifer that:
"A. 1. Supplies any public water system; or
2. Contains sufficient quantity of groundwater to supply a
public water system; and
i. currently supplies drinking water for human
consumption; or
ii. contains fewer than 10,000 milligrams per liter
(mg/L) total dissolved solids (TDS); and
B. Is not an exempted aquifer."
The water quality standard for USDWs is more stringent than EPA's National Secondary
Drinking Water Standards for potable water, which cover aesthetic concerns such as taste
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-4
image:
EPA816-R-04-003 Chapter 1
Introduction
and odor. These secondary standards recommend a IDS limit of 500 mg/L (40 CFR
143.3).
An accurate understanding of the definition of USDW requires understanding of two
other terms: public water system and aquifer exemption.
A public water system is defined at 40 CFR 141.2 as:
"A system for the provision to the public of water for human consumption
through pipes or, after August 5, 1998, other constructed conveyances, if
such a system has at least 15 service connections or regularly serves an
average of at least twenty-five individuals daily at least 60 days out of the
year."
To better quantify the definition of USDW, EPA determined that any aquifer yielding
more than 1 gallon per minute can be expected to provide sufficient quantity of water to
serve a public water system and therefore falls under the definition of a USDW (U.S.
EPA Memorandum, 1993). EPA also assumes that all aquifers contain sufficient quantity
of groundwater to supply a public water system, unless proven otherwise through
empirical data.
An aquifer exemption may be granted under certain circumstances. According to 40
CFR 144.3, an exempted aquifer meets the definition of a USDW, but has been exempted
according to the procedures in 40 CFR 144.7. An aquifer, or portion thereof, can be
designated as an exempted aquifer, if it meets the following criteria (40 CFR 146.4):
1. It does not currently serve as a source of drinking water; and,
2. It cannot now and will not in the future serve as a source of drinking water
because it is:
• Mineral, hydrocarbon, or geothermal energy producing, or can be
demonstrated to be commercially producible; or
• Situated at a depth or location which makes recovery of water for drinking
water purposes economically or technologically impractical; or
• So contaminated that it would be economically or technologically
impractical to render that water fit for human consumption; or
• Located over a Class III well mining area subject to subsidence or
catastrophic collapse; or,
3. The TDS content of the groundwater is more than 3,000 and less than 10,000
mg/L and, it is not reasonably expected to supply a public water system.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-5
image:
EPA816-R-04-003 Chapter 1
Introduction
All requests for aquifer exemptions must be approved by the EPA Administrator or an
authorized representative. A list of exempted aquifers, for states where such exemptions
exist, is maintained by the state agency managing the UIC program or the regional EPA
office. A comprehensive list or map identifying all USDWs in every state does not exist.
Identification of USDWs is an ongoing effort, as is EPA's consideration of aquifer
exemptions. For example, coalbed methane production wells using hydraulic fracturing
to stimulate production may be located in areas that coincide with existing aquifer
exemptions.
Currently, injection associated with hydraulic fracturing of coalbed methane production
wells is regulated only in Alabama under the state UIC program, and that injection
activity falls under the category of Class II wells (Alabama Oil and Gas Board,
Administrative Code, Oil and Gas Report 1, 400-3). Class II wells include the injection
of brines and other fluids that are associated with oil and gas production.
1.4 Potential Effects of Hydraulic Fracturing of Coalbed Methane Wells on
USDWs
EPA identified two possible mechanisms by which hydraulic fracturing fluid injection
into coalbed methane wells might affect the quality of USDWs:
1. The direct injection of fracturing fluids into a USDW in which the coal is
located (Figure 1-2), or injection of fracturing fluids into a coal seam which is
already in hydraulic communication with a USDW (e.g., through a natural
fracture system).
2. The creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
Fracturing fluids can be directly or indirectly injected into a USDW, depending on the
location of the coalbed relative to a USDW. In many coalbed methane-producing regions,
the target coalbeds occur within USDWs, and the fracturing process injects stimulation
fluids directly into the USDWs (Figure 1-2 at the end of the chapter). In other production
regions, target coalbeds are adjacent to the USDWs, which are either higher or lower in the
geologic section. EPA investigated the potential for fractures to extend through
stratigraphic layers that separate coalbeds and USDWs and the potential for stimulation
fluids to indirectly enter a USDW during the fracturing process (Figure 1-3 at the end of
the chapter).
Local geologic conditions may interfere with the complete recovery of fracturing fluids
injected into a formation. As a result, some of the fracturing fluids may be "stranded" in
the USDW (Figures 1-2 and 1-3). Any hazardous constituents in the stimulation fluids
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-6
image:
EPA816-R-04-003 Chapter 1
Introduction
could potentially contaminate groundwater in a USDW and any drinking water supplies
that rely on the USDW.
1.5 Study Approach
Given the enormous variation in geology among and within coalbed basins in the United
States, any initial evaluation of potential impacts by hydraulic fracturing of coalbeds on
USDWs at a national level would necessarily be broadly focused. Based on public input,
EPA decided to carry out this study in discrete phases to better define its scope and to
determine if additional study is needed after assessing the results of the preliminary
phase(s). EPA designed the study to have three possible phases, changing the focus from
general to more specific as findings warrant.
Phase I of the study is a fact-finding effort based primarily on existing literature to
identify and assess the potential threat to USDWs posed by hydraulic fracturing fluid
injection into coalbed methane wells. It is designed to determine if site-specific detailed
studies, including collection of new data, are needed. An overview of the methodology
used for Phase I is provided below; a detailed discussion of this methodology is provided
in Chapter 2.
In Phase I, EPA:
• Conducted a literature review for information on hydraulic fracturing
processes, hydraulic fracturing fluids and additives, the geologic settings of
and the hydraulic fracturing practices used in the 11 major coal basins (Figure
1-1), and the identification of coal seams that are co-located with USDWs.
• Published a request in the Federal Register (66 FR 39396 (U.S. EPA, 2001))
for information from the public, as well as governmental and regulatory
agencies, regarding incidents of groundwater contamination believed to be
associated with hydraulic fracturing of coalbed methane wells.
• Reviewed reported incidents of groundwater contamination and any follow-up
actions or investigations by other parties such as state or local agencies,
industry, and academia.
• Conducted field visits in three states.
In addition, EPA collaborated with the Department of Energy (DOE) to produce a
document that details the technical aspects of hydraulic fracturing in the oil and gas
industry. This document is included as Appendix A to this report.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-7
image:
EPA816-R-04-003 Chapter 1
Introduction
EPA also provided support for a site-specific study, which was conducted by the
Geological Survey of Alabama (GSA). This study attempts to address a concern that is
central to USDW contamination and drawdown issues: the degree to which flow is
confined within coalbeds in coalbed methane fields. Information on the GSA study is
available at http://www.gsa.state.al.us/gsa/3DFracpage/3Dfracstudy.htm.
1.6 Stakeholder Involvement
EPA took several steps to fully involve the public and all stakeholders during the study.
These steps included:
• Publishing Federal Register notices:
- requesting comments on the study plan (65 FR 45774 (USEPA, 2000));
requesting information from the public on any impacts to groundwater
believed to be associated with hydraulic fracturing of coalbed methane
wells (66 FR 39396 (USEPA, 2001));
- Requesting comments on the August 2002 draft of the study (67 FR 55249
(USEPA, 2002)).
• Holding a public meeting to obtain additional stakeholder input on the
proposed study plan published in the July 2000 Federal Register notice (65
FR 45774 (USEPA, 2000))
• Providing periodic updates for stakeholders in the form of written
communication.
• Maintaining a Web site where stakeholders can view the project documents
and provide information to EPA.
EPA also received and reviewed comments from 105 commenters submitted in response
to the August 2002 Federal Register notice (67 FR 55249 (USEPA, 2002)), which
announced the availability of the August 2002 version of this Phase I study report. EPA
incorporated many of these comments into this final Phase I report. A summary of the
public comments and EPA's responses is provided in, "Public Comment and Response
Summary for the Study on the Potential Impacts of Hydraulic Fracturing of Coalbed
Methane Wells on Underground Sources of Drinking Water" (EPA 816-R-04-004),
available on EPA's electronic docket.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-8
image:
EPA816-R-04-003 Chapter 1
Introduction
1.7 Information Contained within This Report
This Phase I report is composed of an executive summary, 7 chapters, 11 attachments,
and 2 appendices. The main chapters address the following topics:
• Chapter 2, Study Methodology, discusses in detail EPA's method for
collecting information under Phase I of the study.
• Chapter 3, Characteristics of Coalbed Methane Production and Associated
Hydraulic Fracturing Practices, discusses the hydraulic fracturing process as it
applies to coalbed methane production.
• Chapter 4, Hydraulic Fracturing Fluids, describes the use and nature of
hydraulic fracturing fluids and their additives. It also discusses EPA's
evaluation of the fate and transport of fracturing fluids that are injected into
targeted coal layers during the hydraulic fracturing process.
• Chapter 5, Summary of Coalbed Methane Basin Descriptions, briefly
describes each of the 11 major coal basins in the United States and discusses
the potential for impacts to USDWs in these basins.
• Chapter 6, Water Quality Incidents, in response to stakeholders'
recommendations, summarizes water quality and quantity complaints received
from citizens pertaining to hydraulic fracturing, coalbed methane production,
and well stimulation.
• Chapter 7, Summary of Findings, summarizes the major findings presented in
Chapters 3 through 6.
In addition, Chapters 3 through 6 contain numerous figures and tables to help readers
visualize the hydraulic fracturing process and to help summarize some of the key
information in the report.
The attachments to the report are a collection of in-depth hydrologic investigations of the
11 coal basins, focusing primarily on the coalbed methane production activities and the
relationship between coalbed and USDW locations within these 11 basins. The
attachments expand the discussions of Chapter 5 with greater details on the specific
geology and gas production activities for the 11 basins.
Appendix A, Hydraulic Fracturing, contains DOE's technical report on hydraulic
fracturing. Appendix B, Quality Assurance Protocol, explains the quality assurance and
quality control measures EPA used to conduct this study.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 1-9
image:
EPA816-R-04-003
Chapter 1
Introduction
Figure 1-2. Hypothetical Mechanism - Direct Fluid Injection Into a USDW (Coal
within USDW)
Fractun- f'uii n Irtjarted iito (.rallied
Frmrturt> Created
San* Find Stun did During Rodurton
, . - - .
3jrandai3 Ruid
(_=^niicwi
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
1-10
image:
EPA816-R-04-003
Chapter 1
Introduction
Figure 1-3. Hypothetical Mechanism - Fracture Creates Connection to USDW
Franco Fluid n lnj«t«d inlo Coaltasd ftnans
RudSlrwTfed Dunnn P^nourtcr
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
1-11
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
Chapter 2
Study Methodology
This chapter outlines EPA's approach for completing Phase I of the study. This chapter
describes the development of the study, the information collection and review process
that EPA used, and the internal and external review process for the report.
2.1 Overview of the Study Methods
EPA developed the Phase I study methodology to assess the potential for contamination
of USDWs due to the injection of hydraulic fracturing fluids into coalbed methane wells,
and to determine, based on these findings, whether further study is warranted.
On July 25, 2000, EPA published a Federal Register notice (65 FR 45774 (USEPA,
2000)) requesting comment on a conceptual study design in order to receive stakeholder
input on how an EPA study should be structured. On August 24, 2000, EPA held a
public meeting to obtain additional stakeholder input on the proposed study design. EPA
received more than 80 sets of comments from industry, state oil and gas agencies,
environmental groups, and individual citizens in response to the Federal Register notice
and public meeting. A summary of these comments can be viewed on EPA's Web site at
www.epa.gov/safewater/uic/cbmstudy.
EPA revised its study approach in response to the comments it received on the conceptual
study design. The final study design, "Study Design for Evaluating the Impacts to
Underground Sources of Drinking Water by Hydraulic Fracturing of Coalbed Methane
Reservoirs", was released in April 2001 and is available on the website referenced above.
One significant change in the final study design was EPA's decision to complete the
study in a phased approach to efficiently address the stated project objectives. This
phased approach, similar in design to that used in other complex studies, would allow
EPA to use information gained in one phase to focus on the need for, and direction of,
subsequent phases.
Phase I of the study was intended as a limited-scope assessment that would enable the
Agency to determine if hydraulic fracturing of coalbed methane wells clearly poses little
or no threat to USDWs, or if the practice may pose a threat. In Phase I, EPA:
• Gathered existing information to review hydraulic fracturing processes,
practices and settings;
• Requested public comment to identify incidents that had not been reported to
EPA; and
• Reviewed reported incidents of groundwater contamination and any follow-up
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-1
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
actions or investigations by other parties (state or local agencies, industry,
academia, etc.).
In addition, as recommended by commenters, EPA decided to compile accounts of
personal experiences with coalbed methane impacts on drinking water wells. These
experiences are summarized in Chapter 6.
In its final study design, EPA indicated that the Agency would make a determination
regarding whether further investigation was needed after analyzing the Phase I
information. Specifically, EPA determined that it would not continue into Phase II of the
study if the investigation found that no hazardous constituents were used in fracturing
fluids, hydraulic fracturing did not increase the hydraulic connection between previously
isolated formations, and reported incidents of water quality degradation could be
attributed to other, more plausible causes.
EPA identified two possible mechanisms by which hydraulic fracturing fluid injection
into coalbed methane wells could potentially affect the quality of USDWs:
1. Direct injection of fracturing fluids into a USDW in which the coal is located,
or injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
To determine if contamination might occur through either mechanism, EPA collected
information on:
1. Hydraulic fracturing practices.
2. Hydraulic fracturing fluids and additives to determine whether these
substances contain hazardous constituents.
3. The hydrogeology of the coalbed methane basins, including the identification
of coal seams that are located in USDWs.
4. Water quality incidents potentially associated with hydraulic fracturing.
EPA anticipated that sufficient information would be available to evaluate the impacts of
direct injection into USDWs because the main considerations are the location of the coal
formations relative to USDWs and the chemical constituents in hydraulic fracturing
fluids. The Agency further anticipated that documenting USDW impacts via the creation
of a hydraulic connection between the coalbed formation and adjacent USDW(s) would
be more difficult. This is because more detailed, site-specific, geological information or
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-2
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
data for specific fracturing events needed to definitively document such a hydraulic
communication are not readily available. Site-specific data include:
• Water quantity and quality conditions in a USDW (or a well) both before and
after a fracturing event;
• Location, dimensions, and conductivity of fractures created during the
coalbed stimulation event;
• Measured changes in groundwater flow between the USDW and coalbeds or
other aquifers; and
• Impacts of other, unrelated, hydrologic and water quality processes that could
also be affecting the USDW.
2.2 Information Sources
EPA obtained available literature and information through:
• Literature reviews.
• Coordination with DOE.
• Interviews with companies that perform hydraulic fracturing and interviews
with citizens, local and state authorities, the Bureau of Land Management and
EPA Region 8 personnel.
• Field visits.
• Responses to EPA's Federal Register request (66 FR 3 93 96 (U. S. EPA,
2001)) for information on incidents of groundwater contamination believed to
be associated with hydraulic fracturing of coalbed methane wells.
EPA researched more than 200 peer-reviewed publications, interviewed approximately
50 employees from industry and state or local government agencies, and communicated
with approximately 40 citizens and groups who are concerned that CBM production
affected their drinking water wells.
The procedure that EPA used to obtain information from each of these sources is
discussed in more detail below. A copy of the quality assurance protocol that EPA
employed to verify all the sources of data used to write this report is provided as
Appendix B.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-3
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
2.2.1 Literature Revi ews
EPA conducted a review of existing literature and information on hydraulic fracturing for
coalbed methane production. The focus of the literature review was to obtain
information on topics 1 through 3 listed in Section 2.1, above.
The degree to which information was available for each of the 11 coalbed basins in the
report was variable. The amount of information available depended on the extent of
exploration and production in each basin.
EPA conducted an extensive literature search, using the Engineering Index and GeoRef
on-line reference databases, for abstracts from technical articles, books, and proceedings.
EPA also conducted Internet-based searches to locate additional information using
relevant Web sites located using various search engines, including Google™, Yahoo®,
and Alta Vista®. EPA used specialized search engines, such as those provided on state
geological survey Web sites and by the Gas Technology Institute (GTI) for specific
queries. All relevant Web sites were logged in project books and referenced in this report
when cited.
EPA conducted these literature searches by subject topics and using the following key
words, either separately or in combination: coal basin, coalbed methane, cross-linked
gel, fracturing fluid additives, fracturing fluid technology, fracturing fluid performance,
fracturing fluids, ground water, hydraulic fracturing, hydraulic fracture dimension,
hydraulic fracture growth, hydrology, linear gel, methane gas production, nitrogen foam,
underground sources of drinking water, and USDWs. EPA printed, catalogued, and
surveyed all results of searches for pertinent journal articles, books, and conference
proceedings containing information that might meet the specific data needs of this report.
EPA acquired most of the pertinent articles, which were identified from the Engineering
Index and GeoRef on-line reference databases, from the University of Texas Library in
Austin because this library's holdings include an extensive collection of publications
related to oil and gas production. EPA researched references from the University of
Texas documents and acquired those documents that were relevant to the study. Only a
small fraction of the pertinent articles, specifically proprietary articles and articles
published for overseas conferences were unavailable. EPA also acquired articles from
GTI. EPA has archived, by topic, all papers collected for the study.
To verify key information extracted from the literature, EPA contacted by phone relevant
organizations such as state regulatory agencies, state geological surveys, natural gas
companies, GTI, and service companies. The Agency used telephone logs to document
all communications. Personal conversations with the employees of the various
organizations yielded additional information in the form of reports, figures, and maps, as
well as statements based on best professional judgment and experience. These were
collected, documented, and referenced in conjunction with the literature identified in the
literature searches. The majority of the literature pertaining to coalbed methane basins
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-4
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
and hydraulic fracturing was written in the early to mid 1990s. According to the Texas
Bureau of Economic Geology (TBEG) (personal communication, TBEG Staff, 2000),
this period of intense activity was a result of the emphasis placed on gas exploration by
the Section 29 Tax Credit of the Crude Oil Windfall Profit Tax Act of 1980 and research
grants to industry, academia, and government agencies. The Section 29 credit does not,
however, apply to coal gas wells drilled after December 31, 1992.
2.2.2 Department of Energy
EPA reviewed information from DOE's "White Paper" on hydraulic fracturing practices.
This paper addresses the following topics:
• Objectives of hydraulic fracturing.
• How candidate wells are selected for hydraulic fracturing.
• How fracture treatments are designed.
• Field operation considerations.
• Physics of fracture formation in coalbeds.
• Fracturing fluids.
• Stimulation techniques used for developing coalbeds.
• Instrumentation/methods for tracking fractures.
The complete DOE paper is included as Appendix A, and excerpts from this paper are
included in Chapter 3, Characteristics of Coalbed Methane Production and Associated
Hydraulic Fracturing Practices.
2.2.3 Interviews
EPA contacted hydraulic fracturing service companies including BJ Services Company,
Halliburton Energy Services, Inc., and Schlumberger Technology Corporation, as well as
a fracturing fluids producer, Hercules, Inc., to obtain information regarding the content of
hydraulic fracturing fluids and additives they use or manufacture. Two companies,
Halliburton and Schlumberger, provided EPA with material safety data sheets (MSDSs)
for several hydraulic fracturing fluids and additives. The MSDSs were reviewed to
determine the nature of the constituents in fracturing fluids used to stimulate coalbed
methane production. These topics are discussed in Chapter 4, Hydraulic Fracturing
Fluids.
EPA also evaluated reports from individuals and organizations that are concerned that
their drinking water supplies were affected by hydraulic fracturing. These reported
personal experiences came from Colorado, New Mexico, Wyoming, Alabama, and
Virginia. In response to these reports, EPA conducted telephone interviews with citizens,
local and state authorities, the Bureau of Land Management and EPA Region 8
personnel. EPA also evaluated state agency responses to any complaints received by EPA
or state agencies. The Agency also evaluated the available data to determine whether
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-5
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
there is sufficient information to reveal the source of the alleged water quality
contamination.
2.2.4 Field Visits
EPA conducted field visits in Colorado, Kansas, and Virginia to better understand how
local coalbed methane production activities may vary from basin to basin. In addition,
during the field visits, EPA was able to meet with concerned local citizens and state
agencies to discuss coalbed methane production issues. A summary of these field visits
is outlined below.
In August 2000, EPA met with a group of concerned citizens, officials from the Colorado
Oil and Gas Conservation Commission, and representatives of the La Plata County
government. EPA witnessed a fracturing event, reviewed records including temperature
logs of past fracturing events conducted on coalbed methane wells, and performed a
reconnaissance of the area allegedly affected by coalbed methane production.
In August 2001, EPA met with the Virginia Department of Mines, Minerals and Energy,
the agency that regulates the coalbed methane production industry in Virginia. The
Department provided information about the state's investigation of water quality
incidents potentially associated with coalbed methane production in the Central
Appalachian Valley. The Department also submitted water quality incident reports for
review by EPA. During this visit, EPA met with concerned citizens in Virginia. Citizens
groups from Buchannan and surrounding counties were invited to meet with EPA and
DOE staff to discuss water quality issues believed to be related to local hydraulic
fracturing of coalbed methane wells. Notes from the meeting are referenced in Chapter
EPA also organized a field visit with Consol Energy, Inc. and Halliburton to witness a
hydraulic fracturing event. Halliburton performed a hydraulic fracture job on a coalbed
methane well in western Virginia using equipment, fracturing fluids, and techniques,
which are typical of those described in the literature. EPA was able to observe the
fracturing process and collect information, including MSDSs from the service company
and gas company engineers. The information from this field visit was used to
supplement the data on hydraulic fracturing fluids in Chapter 4.
In November 2001, EPA witnessed a fracturing event in Wilson County, Kansas, to gain
a better understanding of the regional geology and hydraulic fracturing practices in the
area. In attendance were Colt Energy (the well operator); Consolidated Industrial
Services, Inc. (the service company conducting the fracture job); and two state agencies,
the Kansas Corporation Commission, and the Missouri Department of Natural Resources.
MSDSs for fracturing fluids typically used in the area were also provided to EPA by the
Kansas Corporation Commission.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-6
image:
EPA 816-R-04-003 Chapter 2
Study Methodology
2.2.5 Federal Register Notice to Identify Reported Incidents
EPA provided an opportunity for the public to submit information on any impacts to
groundwater believed to be associated with hydraulic fracturing through a request for
public comment (66 FR 39396 (USEPA, 2001)). EPA also sent copies of the Federal
Register notice with a cover letter to county-level public health and/or environmental
officials in counties that may be producing coalbed methane. In addition, letters were
sent to stakeholders informing them that the Federal Register notice had been published.
Responses to the Federal Register notice are available at EPA's water docket (docket
number W-01-09; Water Docket (MC 4101); Rm EB 57; U.S. Environmental Protection
Agency; 1200 Pennsylvania Avenue, NW; Washington, DC 20460; phone number: (202)
566-2426). A summary of the comments is provided in Chapter 6.
2.3 Review Process
This report has benefited from a series of internal and external technical reviews. EPA
verified information through telephone interviews with state and local officials, as well as
through the Agency's internal quality assurance process. EPA conducted a quality
assurance review of the data collection procedures as well as a review of the individual
literature sources cited in the report. In addition, more than nine EPA offices reviewed
and commented on the draft report. Other federal agencies that reviewed the draft report
included DOE and the U.S. Geological Survey (USGS).
In 2001, EPA also submitted the draft report to a scientific peer-review panel consisting
of experts from industry, academia, and government agencies. The panel's task was to
review the draft report and provide comments on the descriptions and conclusions
developed by EPA. The panel also provided information about additional data sources to
supplement those used in the report. Following receipt of comments on the draft report,
EPA made the appropriate changes to the document prior to its publication and release.
EPA made the report available for public comment by an announcement in the Federal
Register on August 28, 2002 (67 FR 55249 (USEPA, 2002)). The 60-day public
comment period officially ended on October 28, 2002. The Agency received and
reviewed comments from 105 commenters and incorporated many of these comments
into this final Phase I report. A summary of the public comments and EPA's responses is
provided in, "Public Comment and Response Summary for the Study on the Potential
Impacts of Hydraulic Fracturing of Coalbed Methane Wells on Underground Sources of
Drinking Water" (EPA 816-R-04-004), available on EPA's electronic docket.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 2-7
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
Chapter 3
Characteristics of Coalbed Methane Production
and Associated Hydraulic Fracturing Practices
Understanding the practice of hydraulic fracturing as it pertains to coalbed methane
production is an important first step in evaluating its potential impacts on the quality of
USDWs. This chapter presents the following: an overview of the geologic processes
leading to coal formation, an overview of coalbed methane production practices, a
discussion of fracture behavior, a review of the literature on the use and recovery of
fracturing fluids, a discussion of mechanisms affecting fluid recovery, and a summary of
the methods used for measuring and predicting fracture dimensions and fracturing fluid
movement. In addition, several diagrams have been included at the end of this chapter to
help illustrate many of these topics. Specifically, Figures 3-1 through 3-8 show the
location of the coal basins, the geography of a peat-forming system, the geometry of
natural cleats and hydraulically induced fractures, an overview of the hydraulic fracturing
process, the relationship between water and gas production rates, and side and plan views
of vertical hydraulic fractures.
3.1 Introduction
Coalbed methane is a gas formed as part of the geological process of coal generation, and
is contained in varying quantities within all coal. Coalbed methane is exceptionally pure
compared to conventional natural gas, containing only very small proportions of "wet"
compounds (e.g., heavier hydrocarbons such as ethane and butane) and other gases (e.g.,
hydrogen sulfide and carbon dioxide). Coalbed gas is over 90 percent methane, and is
suitable for introduction into a commercial pipeline with little or no treatment (Rice,
1993; Levine, 1993).
From the earliest days of coal mining, the flammable and explosive gas in coalbeds has
been one of mining's paramount safety problems. Over the centuries, miners have
developed several methods to extract the coalbed methane from coal and mine workings.
Coalbed methane well production began in 1971 and was originally intended as a safety
measure in underground coalmines to reduce the explosion hazard posed by methane
(Elder and Deul, 1974).
In 1980, the United States Congress enacted a tax credit for "Non-conventional energy
production." In 1984, there were only several hundred coalbed methane wells in the
United States and most were used for mine de-methanization. By 1990, the anticipated
expiration of the tax credit contributed to a dramatic increase in the number of coalbed
methane wells nationwide. In addition, DOE and GTI supported extensive research into
coalbed exploration and production methods. Federal tax credits and State Severance
Tax exemptions served to subsidize the development of coalbed methane resources (Soot,
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-1
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
1991; Pashin and Hinkle, 1997). The federal tax credits and incentives expired at the end
of 1992, but coalbed methane exploration, development, and reserves have remained
stable or increased (Stevens et al., 1996). At the end of 2000, coalbed methane
production from 13 states totaled 1.353 trillion cubic feet, an increase of 156 percent
from 1992. During 2000, a total of 13,973 coalbed methane wells were in production
(GTI; EPA Regional Offices, 2001). By the end of 2000, coalbed methane production
accounted for about 7 percent of the total United States dry gas production and 9 percent
of proven dry gas reserves (EIA, 2001).
Coal is defined as a rock that contains at least 50 percent organic matter by weight. The
precursor of coal is peat, plant matter deposited over time in fresh-water swamps
associated with coastal deltaic rivers. The coal resources from which coalbed methane is
derived have similar geologic origins. In the United States, they are usually found in
geologic formations that are approximately 65-325 million years old. Coal formation
occurred during a time of moderate climate and broad inland oceans. Sea level rose and
fell in conjunction with tectonic forces (i.e., subsidence and uplift of land masses) and
melting/freezing cycles of decreases and increases in the polar ice masses. As a result,
coastal environments such as coastal deltas and peat swamps migrated landward when
sea levels rose and moved seaward when sea levels fell, marked by cycles of
submergence and emergence. With these cycles of rising and falling sea levels, what was
a peat swamp at one time would later be under 100 feet of water. The cycle of sea level
rising and falling is marked in the geologic record as cycles of inter-layered deep and
shallow water sediments.
The type of sediments deposited at a given location varied with the depth of submergence
(Figure 3-2). Generally, carbon-rich organic plant matter was deposited in shallow peat
swamps, sand was deposited along beaches and other near-shore, shallow marine
environments, and silts and clays and calcium-rich muds were deposited further off-shore
in deeper marine environments. Subjected to high pressure over considerable time (due
to burial under subsequent sediments), the peats were transformed into coal, the sand into
sandstone, the silts and clays into shales, siltstones, and mudstones, and the calcium-rich
muds were transformed into limestones. These coal-bearing inter-layered sedimentary
sequences are sometimes referred to as "coal cycles." The idealized coal cycle consists
of repeated sequences of very fine-grained sediments (shales and limestones) overlain by
coarser sediments (siltstones and sandstones), and then capped by coal. The sequence
repeats with shales and limestones over the coal, followed by siltstones and sandstones,
then more coal, and so on. Sometimes certain formations are missing from the
sequences, so coal is often, though not always, overlain by shales and limestones.
The sedimentation patterns in these fluctuating coastal environments over geologic time
scale determined the presence, thickness, and geometry of present-day coalbeds. The
number of coal cycles determines the number of resulting coalbeds. For example, the
Black Warrior Basin of Alabama includes up to 10 cycles, whereas the San Juan Basin
(New Mexico and Colorado) contains as few as 3. The short, rising and falling sea level
cycles reflected in the Black Warrior Basin geology produced many thin coalbeds,
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-2
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
ranging from less than 1 inch to as much as 4 feet thick (Carrol et al., 1993; Pashin,
1994a and 1994b), whereas the stable, long-term cycles of the San Juan Basin produced
fewer, but thicker coalbeds, with single coalbeds up to 70 feet thick (Kaiser and Ayers,
1994).
Peat is transformed into coal when it is buried by accumulating sediment and heated in
the subsurface over geologic time. The "rank" of coal describes the amount of energy
(measured in British thermal units or Btus) it contains, and is a function of the proportion
and type of organic matter, the length and temperature of burial, and the influences of
subsequent hydrogeologic and tectonic processes (Carrol et al., 1993; Levine, 1993; Rice,
1993). Methane is generated as part of the process whereby peat is transformed into coal.
The origin of methane in coal of low rank, such as bituminous coal, is primarily biogenic
(i.e., the result of bacterial action on organic matter) (Levine 1993, as cited by the
Alabama Oil and Gas Board, 2002). Low rank coals tend to have lower gas content than
high rank coals such as anthracite. Anthracite can have extremely high gas content, but
the gas tends to desorb so slowly that anthracite is an insignificant source of coalbed
methane (Levine, 1993, as cited by the Alabama Oil and Gas Board, 2002). Commercial
coalbed methane production takes place in coals of mid-rank, usually low- to high-
volatile bituminous coals (Levine, 1993; Rice, 1993).
A network of fractures, joints, and a sub-network of small joints called cleats commonly
characterize the physical structure of coalbeds. Joints are larger, systematic, near-vertical
fractures within the coal, generally spaced from several feet to several dozen feet apart
(Close, 1993; Levine, 1993). There are two types of cleats: the primary, more
continuous cleats are called face cleats, while the abutting cleats are called butt cleats
(Laubach and Tremain 1991; Close, 1993; Levine, 1993) (Figure 3-3). The butt cleats
appear as the rungs on a ladder that are bounded on each side by the face cleats. The
spacing between cleats is often roughly proportional to the thickness of coal cut by the
cleats; thin coals have more closely spaced cleats and thick coals more widely spaced
cleats (Laubach et al., 1998, as cited by Olson, 2001).
The primary (or natural) permeability of coal is very low, typically ranging from 0.1 to
30 millidarcies (md) (McKee et al., 1989). According to Warpinski (2001), because coal
is a very weak (low modulus) material and cannot take much stress without fracturing,
coal is almost always highly fractured and cleated. The resulting network of fractures
commonly gives coalbeds a high secondary permeability (despite coal's typically low
primary permeability). Groundwater, hydraulic fracturing fluids, and methane gas can
more easily flow through the network of fractures. Because hydraulic fracturing
generally enlarges pre-existing fractures and rarely creates new fractures (Steidl, 1993;
Diamond, 1987a and b; Diamond and Oyler, 1987), this network of natural fractures is
very important to the extraction of methane from the coal.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-3
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
3.2 Hydraulic Fracturing
This section provides an overview of the hydraulic fracturing process, and the factors that
affect fracture behavior and fracture orientation. Figure 3-4 provides a simplified
graphical representation of a hypothetical hydraulic fracturing event in a coalbed methane
well. This diagram shows the fracture initiation and propagation stages, as well as the
proppant placement and fracturing fluid recovery stages. Only horizontal fractures are
shown in this diagram, although hydraulically induced fractures are often vertically
oriented.
3.2.1 The Hydraulic Fracturing Process
Hydraulic fracturing is a technique used by the oil and gas industry to improve the
production efficiency of oil and coalbed methane wells. The extraction of coalbed
methane is enhanced by hydraulically enlarging and/or creating fractures in the coal
zones. The resulting fracture system facilitates pumping of groundwater from the coal
zone, thereby reducing pressure and enabling the methane to be released from the coal
and more easily pumped through the fracture system back to the well (and then through
the well to the surface). To initiate the process, a production well is drilled into the
targeted coalbeds. Fracturing fluids containing proppants are then injected under high
pressure into the well and specifically into the targeted coalbeds in the subsurface.
The fracturing fluids are injected into the subsurface at a rate and pressure that are too
high for the targeted coal zone to accept. As the resistance to the injected fluids
increases, the pressure in the injecting well increases to a level that exceeds the
breakdown pressure of the rocks in the targeted coal zone, and the rocks "breakdown"
(Olson, 2001). In this way, the hydraulic fracturing process "fractures" the coalbeds (and
sometimes other geologic strata within or around the targeted coal zones). This process
sometimes can create new fractures, but most often opportunistically enlarges existing
fractures, increasing the connections of the natural fracture networks in and around the
coalbeds (Steidl 1993; Diamond 1987a and b; Diamond and Oyler, 1987). The pressure-
induced fracturing serves to connect the network of fractures in the coalbeds to the
hydraulic fracturing well (which subsequently will serve as the methane extraction or
production well). The fracturing fluids pumped into the subsurface under high pressure
also deliver and emplace the "proppant." The most common proppant is fine sand; under
pressure, the sand is forced into the natural and/or enlarged fractures and acts to "prop"
open the fractures even after the fracturing pressure is reduced. The increased
permeability due to fracturing and proppant emplacement facilitates the flow and
extraction of methane from coalbeds.
Methane within coalbeds is not structurally "trapped" by overlying geologic strata, as in
the geologic environments typical of conventional gas deposits. Only about 5 to 9
percent of the coalbed methane is present as "free" gas within the joints and cleats of
coalbeds. Most of the coalbed methane is contained within the coal itself (adsorbed to
the sides of the small pores in the coal) (Koenig, 1989; Winston, 1990; Close, 1993).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-4
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
Before coalbed methane production begins, groundwater and injected fracturing fluids
are first pumped out (or "produced" in industry terminology) from the network of
fractures in and around the coal zone. The fluids are pumped until the pressure declines
to the point that methane begins to desorb from the coal (Gray, 1987).
Coalbed methane production initially requires pumping and removing significant
amounts of water to sufficiently reduce the hydrostatic pressure in the subsurface so that
methane can desorb from the coal before methane extraction can begin. Coalbed
methane is produced at close to atmospheric pressure (Ely et al., 1990; Schraufnagel,
1993). The proportion of water to methane pumped is initially high and declines with
increasing coalbed methane production (Figure 3-5). In contrast, in the production of
conventional petroleum-based gas, the production of gas is initially high, and as gas
production continues over time and the gas resources are progressively depleted, gas
production decreases and the amount of water pumped increases.
Almost every coalbed targeted for methane production must be hydraulically fractured to
connect the production well bore to the coalbed fracture network (Hoiditch et al., 1988).
Although the general hydraulic fracturing process (described above) is generally similar
across the country, the details of the process can differ significantly from location to
location depending on the site-specific geologic conditions. For example, although most
hydraulic fracturing wells are completely cased except for openings at the targeted coal
zone, many wells in the San Juan Basin are fractured by creating a cavity in the open-
hole section. Also, in contrast to the typical fracturing job, many wells in the Black
Warrior Basin are stimulated more than once. Here, when wells are open to multiple coal
seams, the hydraulic fracturing process may involve several or multiple fracturing events,
using from 2 to 5 hydraulic fracture treatments per well (depending on number of seams
and spacing between seams). Many coalbed methane wells are re-fractured at some time
after the initial treatment in an effort to re-connect the wellbore to the production zones to
overcome plugging or other well problems (Holditch, 1990; Saulsberry et al., 1990;
Palmer et al., 1991a and 1991b; Holditch, 1993). Also, in response to site-specific coal
geology and the economics of coalbed methane production where coal seams are thin and
vertically separated by up to hundreds of feet of intervening rock) operators might design
fracture treatments to enhance the vertical dimension and perform several fracture
treatments within a single well to produce methane in an economically viable fashion,
(Ely, et al., 1990; Holditch, 1990; Saulsberry et al., 1990; Spafford, 1991; Holditch,
1993).
3.2.2 Factors Affecting Fracture Behavior
Fracture behavior is of interest because it contributes to an understanding of the potential
impact of fracturing fluid injection on USDWs; the opportunities for fracture connections
within or into a USDW are affected by the extent to which a hydraulically induced
fracture grows. Specifically, when hydraulic fracturing fluids are injected into
formations that are not themselves USDWs, the following scenarios are of potential
concern:
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-5
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
• The hydraulically induced fracture may extend from the target formation into
a USDW.
• The hydraulically induced fracture may connect with natural (existing)
fracture systems and/or porous and permeable formations, which may
facilitate the movement of fracturing fluids into a USDW.
Fracture behavior through coal and other geologic formations commonly present above
and below coalbeds depends on site-specific factors such as the following:
1. Physical properties, types, thicknesses, and depths of the targeted coalbeds as
well as those of the surrounding geologic formations.
2. Presence of existing natural fracture systems and their orientation in the
coalbeds and surrounding formations.
3. Amount and distribution of stress (i.e., in-situ stress), and the stress contrasts
between the targeted coalbeds and surrounding formations.
4. Hydraulic fracture stimulation design including volume of fracturing fluid
injected into the subsurface as well as the fluid injection rate and fluid
viscosity.
Many of these factors are interrelated and together will influence whether and how far
hydraulic fractures will propagate into or beyond coalbeds targeted for fracturing. These
factors are discussed below.
Properties of Coalbeds and Surrounding Formations
Coalbed depth and rock types in the coal zone have important fundamental influences on
fracture dimensions and orientations. According to Nielsen and Hansen (1987, as cited in
Appendix A: DOE, Hydraulic Fracturing), generally, at depths of less than 1,000 feet,
the direction of least principal stress tends to be vertical and, therefore, at these relatively
shallow depths fractures typically have more of a horizontal than a vertical component.
Here, horizontal fractures tend to be created because the hydraulically induced pressure
forces the walls of the fracture to open in the direction of least stress (which is vertical),
creating a horizontal fracture. At these shallower depths, the horizontal fractures result
from the low vertical stress due to the relatively low weight of overlying geologic
material (due to the shallow depth). Shallow vertical fractures are most likely due to the
presence of natural (existing) vertical fractures, from which hydraulically induced
vertical fractures can initiate. Generally, in locations deeper than 1,000 feet, the least
principal stress tends to be horizontal so vertical fractures tend to form. Vertical fractures
created in these greater depths can propagate vertically to shallower depths and develop a
horizontal component (Nielsen and Hansen, 1987 as cited in Appendix A: DOE,
Hydraulic Fracturing). In the formation of these "T-fractures," the fracture tip may fill
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-6
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
with coal fines or intercept a zone of stress contrast, causing the fracture to turn and
develop horizontally, sometimes at the contact of the coalbed and an overlying formation.
In many coalbed methane basins, the depths, lithologic properties, and stress fields of the
coal zones result in hydraulic fractures that are higher than they are long ("length" refers
to horizontal distance from the well bore) (Diamond, 1987a; Morales et al., 1990; Zuber
et al., 1990; Holditch et al., 1989; Palmer and Sparks, 1990; Jones and Schraufnagel,
1991; Steidl, 1991; Wright, 1992; Palmer et al., 1991a and 1993a). Almost all of the sites
studied by Diamond (1987a and b) had vertical fractures, and about half had horizontal
fractures.
Naceur and Touboul (1990) state that the primary mechanisms controlling fracture height
are contrasts in the physical properties of the rock strata within and surrounding the coal
zone being fractured. Contrasts in strata stresses, moduli, leakoff, and toughness affect
fracture growth, with stress contrasts being the most important mechanism controlling
fracture height (Naceur and Touboul 1990). (Stress is discussed in more detail later in
this section.) Moduli are the ratios of stress to strain in various formations. Leakoff is
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. Toughness can be
defined as the point at which enough stress intensity has been applied to a rock formation,
so that a fracture initiates and propagates. Coal is generally very weak (with low
modulus) and easily fractures. Siltstones, sandstones, and mudstones (other rock types
commonly occurring in coal zones) tend to have higher moduli, greater toughness and
fracture less easily (Warpinski, 2001). Thick shales, which commonly overlie coalbeds,
often act as a barrier to fracture growth (see Appendix A).
Another factor controlling fracture height can be the highly cleated nature of some
coalbeds. In some cases, highly cleated coal seams will prevent fractures from growing
vertically. When the fracturing fluid enters the coal seam, it is contained within the coal
seam's dense system of cleats and the growth of the hydraulic fracture will be limited to
the coal seam (see Appendix A).
The low permeability of relatively unfractured shale may help to protect USDWs from
being affected by hydraulic fracturing fluids in some basins. If sufficiently thick and
relatively unfractured shales are present, they may act as a barrier not only to fracture
height growth, but also to fluid movement. The degree to which any formation overlying
targeted coalbeds will act as a hydraulic barrier will depend on site-specific factors.
The lithology of coalbeds and surrounding formations is variable in the basins where
coalbed methane is produced. Although common, the idealized coal cycle (with shales
overlying coalbeds) is not always present in all coal basins or necessarily in all parts of
any basin. Although Holditch (1993) states that fracture heights can grow where the coal
seam is bounded above or below by sandstone, Warpinski (2001) states that highly
layered formations or very permeable strata, such as some sandstones, can act to inhibit
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-7
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
fracture growth. Some of the coal seams of the San Juan Basin are bounded below by
sandstone. In some locations in each of the other basins, coalbeds are underlain by,
overlain by, or interbedded with sandstones. Additional detail on the stratigraphy within
each basin is provided in the attachments to this study.
Differences in fracture behavior may also be due in part to very small (but influential)
layers or irregularities that exist in the rocks as part of the sedimentation process that
created them. Therefore, a valid measurement of rock properties relevant to fracture
behavior at one location may not adequately represent the properties of similar rock at
another location (Hanson et al., 1987; Jones et al., 1987a and 1987b; Palmer et al., 1989;
Morales et al., 1990; Naceur and Touboul, 1990; Jones and Schraufnagel, 1991; Palmer
et al., 1993b; Elbel, 1994). For example, the presence of a shallow clay layer as thin as
10 millimeters at the upper contact of a coal seam can cause a vertically propagating,
shallow hydraulic fracture to "turn" horizontal and fail to penetrate the next overlying
coal seam (Jones et al., 1987a; Palmer et al., 1989; Morales et al., 1990; Palmer et al.,
1991b and 1993b). In other cases, hydraulic fractures may penetrate into or even, as
shown in the case of some thin shales, completely through overlying shale layers
(Diamond, 1987a and b; Diamond and Oyler, 1987). Warpinski et al. (1982) found that
even microscopically-thin ash beds can influence hydraulic fracture propagation. In other
words, the site-specific geology can play a key role in influencing fracture behavior. In
addition to the effects of the rock type and sometimes even thin layers within strata,
natural fractures also play a role in fracture behavior and fracture propagation.
Natural Fracture Systems
Steidl (1993), based on his "mined-through" studies, concluded that high coalbed
methane production depends greatly on the presence of pre-existing natural fracture
systems. Hydraulic fracturing tends to widen naturally occurring planes of weakness and
rarely creates new fractures, as based on observations by Diamond (1987a and b) and
Diamond and Oyler (1987) in their mined-through studies. ("Mined-through" studies
provide unique subsurface access to investigate coalbeds and surrounding rock after
hydraulic fracturing. Mined-through studies are reviewed in more detail in section 3.4.1.)
Diamond and Oyler (1987) also noted that this opportunistic enlarging of preexisting
fractures appears to account for those cases where hydraulic fractures propagate from the
targeted coalbeds into overlying rock, and their studies found penetration into overlying
layers in nearly half of the fractures intercepted by underground mines.
Importantly, in several locations in the Diamond (1987a and b) study sites, fluorescent
paint was injected along with the hydraulic fracturing fluids and the paint was found in
natural fractures from 200 to slightly more than 600 feet beyond the sand-filled
("propped") portions of hydraulically induced or enlarged fractures. This suggests that
the induced/enlarged fractures link into the existing fracture network system and that
hydraulic fracturing fluids can move beyond, and sometimes significantly beyond, the
propped, sand-filled portions of hydraulically induced fractures (Steidl 1993; Diamond
1987a and b; Diamond and Oyler, 1987). The mined-through studies did not conduct
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-8
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
systematic assessments of the extent of the fractures into or through the roof rock shales
that were immediately above the mined coal (the rock strata immediately above a mined
coal layer is referred to as the "roof rock").
In-Situ Stress and Stress Contrasts
In-situ stress and the relative stress of neighboring geologic strata are important
influences on fracture behavior. A discussion of in-situ stress is provided in DOE's paper
"Hydraulic Fracturing" (provided as Appendix A). In-situ stress is described as:
"Underground formations are confined and under stress... [The graphic
below] illustrates the local stress state at depth for an element of
formation. The stresses can be divided into 3 principal stresses... [In the
graphic below,] ai is the vertical stress, 02 is the maximum horizontal
stress, while 03 is the minimum horizontal stress, where <Ji>a2>a3. This
is a typical configuration for coalbed methane reservoirs. However,
depending on geologic conditions, the vertical stress could also be the
intermediate (a2) or minimum stress (a3). These stresses are normally
compressive and vary in magnitude throughout the reservoir, particularly
in the vertical direction (from layer to layer). The magnitude and direction
of the principal stresses are important because they control the pressure
required to create and propagate a fracture, the shape and vertical extent of
the fracture, the direction of the fracture, and the stresses trying to crush
and/or embed the propping agent during production."
a
a 4 > a o > a
Local in-situ stress at depth.
According to (Naceur and Touboul 1990), the contrast in stress between adjacent rock
strata within and surrounding the targeted coal zone is the most important mechanism
controlling fracture height. Stress contrast is important in determining whether a fracture
will continue to propagate in the same direction when it hits a geologic contact between
two different rock types. Often, a high stress contrast results in a barrier to fracture
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-9
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
propagation. An example of this would be where there is a geologic contact between a
coalbed and an overlying, thick, higher-stress shale.
Hydraulic Fracture Stimulation Design
The procedures and fracturing fluids used to stimulate coalbed methane wells can differ
from operator to operator in a single basin due to local characteristics of geology and
depth and to perceived advantages of cost, effectiveness, production characteristics, or
other factors. On a larger scale, although fracture stimulations in coalbed methane
projects in different basins may share common rock types and characteristics, fracture
behavior can differ significantly. Discussions on hydraulic fracturing practices in 11
individual coal basins are included in Chapter 5 and in Attachments 1 through 11.
Aspects of fracture behavior, such as fracture dimensions (height, length, and width), are
affected by the different fracturing approaches taken by the operator during a hydraulic
fracturing event. Generally, the larger the volume of fracturing fluids injected, the larger
the potential fracture dimensions. Fluid injection rates and viscosity can also affect
fracture dimensions (Olson, 2001; Diamond and Oyler, 1987). Large injection volumes
also often result in extensive networks of induced fractures. Gelled water treatments may
result in the widest and longest fractures, but this occurrence cannot be concluded with
certainty from the mined-through studies (Diamond and Oyler, 1987; Diamond 1987a
and b).
The effects of these operator-controlled actions interact with and are influenced by the
physical properties, depths, and in-situ stress of the geologic formations being fractured
(as listed above). For example, if a hydraulically induced fracture has a relatively
constant height due to a geologic layer acting as a barrier to fracture propagation, and the
fracture is forced to grow and increase in volume (through an increased volume of
fracturing fluid), the fracture will mainly grow in length. Also, increasing fluid viscosity
can increase the pressure due to injection, resulting in greater fracture width, and thus
often shorter fractures (Olson, 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-10
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
3.3 Fracturing Fluids
The fluids used for fracture development are pumped at high pressure into the well. They
may be "clear" (most commonly water, but may include acid, oil, or water with friction-
reducer additives) or "gelled" (viscosity-modified water, using guar or other gelling
agents). Some literature indicates that coalbed fracture treatments use from 50,000 to
350,000 gallons of various stimulation and fracturing fluids, and from 75,000 to 320,000
pounds of sand as proppant (Holditch et al., 1988 and 1989; Jeu et al., 1988; Hinkel et al.,
1991; Holditch, 1993; Palmer et al., 1991b, 1993a, and 1993b). More typical injection
volumes, based on average injection volume data provided by Halliburton for six coalbed
methane locations, indicate a maximum average injection volume of 150,000 gal/well
and a median average injection volume of 57,500 gal/well (Halliburton, Inc., 2003).
Depending on the basin and treatment design, the composition of these fluids varies
significantly, from simple water and sand to complex polymeric substances with a
multitude of additives. Types of fracturing fluids are discussed in greater detail in
Chapter 4.
3.3.1 Quantifying Fluid Recovery
Several studies have evaluated the recovery rates of hydraulic fracturing treatment fluids
in coal and non-coal formations as discussed in more detail below. Non-coal formations
were evaluated to augment the available flowback data.
Coal Formation
Palmer et al. (1991a) measured flowback rates in 13 hydraulic fracturing wells to
compare the gas production resulting from the use of water versus gel-based fracturing
fluids. This study was conducted in a coal seam with permeabilities from 5 to 20 md.
Ten samples collected over a 19-day flowback period indicated a recovery rate of 61
percent. Palmer et al. (199la) predicted total recovery to be from 68 percent to as much
as 82 percent.
Non-Coal Formations
Willberg et al. (1997) conducted a flowback analysis in 10 wells in a heterogeneous
sandstone and shale environment that was highly impermeable (i.e., with a permeability
of 0.01 md). The fluids used in this study were recovered at an average efficiency of 35
percent during the 4 to 5 day flowback period. Three wells were then sampled every 4 to
8 hours during the subsequent gas production phase to assess long-term polymer
recovery, which was found to be minimal (3 percent). Sampling of injected fluid and
chloride concentrations indicated that as the flowback and gas production periods
progressed, decreasing proportions of the extracted water consisted of the injected fluid,
while increasing proportions were natural formation water. In other words, natural
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-11
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
formation water was able to bypass viscous gel trapped in the formation and flow into the
production wells.
The authors further cited laboratory studies indicating that water may flow past the gel in
sand such as that used as proppant in these studies (Willberg et al., 1997). Because the
gel is more viscous than water, it is easier for water to respond to pumping and flow
through the formation towards the production well. As Willberg et al. (1997) writes,
"Production of formation water effectively competes and eventually supersedes residual
fracturing fluid recovery, thereby limiting the overall cleanup efficiency." Given that the
environments in which coalbed methane is produced are also generally saturated with
water, and similar sands are used as proppants, it is possible that gel recovery is impeded
in much the same way in coalbed methane stimulations.
Willberg et al. (1998) conducted another flowback analysis and described the effect of
flowback rate on cleanup efficiency in an initially dry, very low permeability (0.001 md)
shale. Some wells in this study were pumped at low flowback rates (less than 3 barrels
per minute (bbl/min). Other wells were pumped more aggressively at greater than 3
bbl/min. Thirty-one percent of the injected fluids were recovered when low flowback
rates were applied over a 5-day period. Forty-six percent of the fluids were recovered
when aggressive flowback rates were applied in other wells over a 2-day period.
Additional fluid recovery (10 percent to 13 percent) was achieved during the subsequent
gas production phase, resulting in a total recovery rate of 41 percent to 59 percent.
Willberg speculated that the lower recovery rate in the 1997 study was due to the
pumping of large amounts of formation water during the recovery process, compared to
the 1998 study that was conducted in a relatively dry environment.
3.3.2 Mechanisms Affecting Fluid Recovery
A variety of site-specific factors will influence the recovery efficiency of fracturing
fluids. These factors are summarized as follows:
• Fluids can "leakoff' (flow away) from the primary hydraulically induced
fracture into smaller secondary fractures. The fluids then become trapped in
the secondary fractures and/or pores of porous rock.
• Fluids can become entrapped due to the "check-valve effect," wherein
fractures narrow again after the injection of fracturing fluid ceases, formation
pressure decreases, and extraction of methane and groundwater begins.
• Some fluid constituents can become adsorbed to coal or react chemically with
the formation.
• Some volume of the fluids, moving along the hydraulically induced fracture,
may move beyond the capture zone of the pumping well, and thus cannot be
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-12
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
recovered during fluid recovery. The capture zone of the production well is
that portion of the aquifer that contributes water to the well.
• Some fluid constituents may not completely mix with groundwater and
therefore would be difficult to recover during production pumping.
Each of these mechanisms is discussed in greater detail in this section.
Fluid Leakoff
Fluids can be "lost" (i.e., remain in the subsurface unrecovered) due to "leakoff' into
connected fractures and the pores of porous rocks (Figure 3-7). Fracturing fluids injected
into the primary hydraulically induced fracture can intersect and flow (leakoff) into
preexisting smaller natural fractures. Some of the fluids lost in this way may occur very
close to the well bore after traveling minimal distances in the hydraulically induced
fracture before being diverted into other fractures and pores. The volume of fracturing
fluids that may be lost in this way depends on the permeability of the rocks and the
surface area of the fracture(s).
The high injection pressures of hydraulic fracturing can force the fracturing fluids to be
transported deep into secondary fractures. The cleats in coal are presumed to play an
important role in leakoff (Olson, 2001). Movement into smaller fractures and cleats can
be to a point where flowback efforts will not recover them. The pressure reduction
caused by pumping during subsequent production is not sufficient to recapture all the
fluid that has leaked off into the formation. The loss of fluids due to leakoff into small
fractures and pores is minimized by the use of cross-linked gels, discussed in more detail
in Chapter 4.
Check-Valve Effect
A check-valve effect occurs when natural or propagating fractures open and allow fluids
to flow through when fracturing pressure is high, but subsequently prevent the fluids
from flowing back towards the production well as they close after fracturing pressure
decreases (Warpinski et al., 1988; Palmer et al., 1991a). A long fracture can be pinched
off at some distance from the well. This reduces the effective fracture length available to
transport methane from various locations within the coalbed to the production well.
Fluids trapped beyond the "pinch point" are unlikely to be recovered during flowback.
In most cases, when the fracturing pressure is released, the fracture closes in response to
natural subsurface compressive stresses. Because the primary purpose of hydraulic
fracturing is to increase the effective permeability of the target formation and connect
new or widened fractures to the well, a closed fracture is of little use. Therefore, a
component of coalbed methane production well development is to "prop" the fracture
open, so that the enhanced permeability from the pressure-induced fracturing persists
even after fracturing pressure is terminated. To this end, operators use a system of fluids
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-13
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
and "proppants" to create and preserve a high-permeability fracture-channel from the
well into the formation.
The check-valve effect takes place in locations beyond the zone where proppants have
been emplaced (or in smaller secondary fractures that have not received any proppant).
Because of the heterogeneous, stratified, and fractured nature of coal deposits, it is likely
that some volume of stimulation fluid cannot be recovered due to its movement into
zones that were not completely "propped."
Adsorption and Chemical Reactions
Adsorption and chemical reactions can prevent the fluid from being recovered.
Adsorption is the process by which fluid constituents adhere to a solid surface (i.e., the
coal, in this case) and are thereby unavailable to flow with groundwater. Adsorption to
coal is likely; however, adsorption to other surrounding geologic material (e.g., shale,
sandstone) is likely to be minimal. Another possible reaction affecting the recovery of
fracturing fluid constituents is the neutralization of acids (in the fracturing fluids) by
carbonates in the subsurface.
Movement of Fluids Outside the Capture Zone
Fracturing fluids injected into the target coal zone flow into fractures under very high
pressure. The hydraulic gradients driving fluid flow away from the well during injection
are much greater than the hydraulic gradients pulling fluid flow back towards the
production well during flowback and production pumping. Some portion of the coalbed
methane fracturing fluids could be forced along the hydraulically induced fracture to a
point beyond the capture zone of the production well. The size of the capture zone will
be affected by the regional groundwater gradients, as well as by the drawdown caused by
the well. If fracturing fluids have been injected to a point outside of the well's capture
zone, they will not be recovered through production pumping and, if mobile, may be
available to migrate through an aquifer. Site-specific geologic, hydrogeologic, injection
pressure, and production pumping details would provide the information needed to
estimate the dimension of the production well capture zone and the extent to which the
fracturing fluids might travel, disperse, and dilute.
Incomplete Mixing of Fracturing Fluids with Water
Steidl (1993) documented the occurrence of a gelling agent that did not dissolve
completely and formed clumps at 15 times the injected concentration in the fracture
induced by one well. Steidl (1993) also directly observed, in his mined-through studies,
gel hanging in stringy clumps in many other fractures induced by that one well. As
Willberg et al. (1997) noted, laboratory studies indicate that fingered flow of water past
residual gel may impede fluid recovery. Therefore, some fracturing fluid gels appear not
to flow with groundwater during production pumping and remain in the subsurface
unrecovered. Such gels are unlikely to flow with groundwater during production, but
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-14
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
may present a source of gel constituents to flowing groundwater during and after
production.
3.4 Measuring and Predicting the Extent of Fluid Movement
Because fractures can possibly connect with or even extend into USDWs, fracture height
is relevant to the issue of whether hydraulic fracturing fluids can affect USDWs. Current
methods of measuring or predicting fracture growth, including mathematical models, are
described. The models are effective in setting parameters for a given hydraulic fracture
operation. Coalbed methane well operators have a financial incentive to keep the
hydraulically induced fracture generally within the target coal zone so that expenditures
on hydraulic horsepower, fracturing fluids, and proppants are minimized for commercial
extraction of methane from the coal. In addition, a detailed review is included on
"mined-through" studies that were conducted primarily by the U.S. Bureau of Mines.
These studies provide unique information on the direct measurement of the dimensions
and other characteristics of fractures created in coal seams and surrounding strata by
hydraulic fracturing. Paint, injected with the fracturing fluids, was used as a tracer in
some of these studies, enabling one of the most direct measurements of the extent of fluid
movement due to hydraulic fracturing.
The particular stratigraphy of a fracturing site will determine what fracture heights are
significant with respect to USDWs. That is, a given fracture height may be considered
small at a particular site in one basin, but may be more significant in another basin where
there is a smaller vertical separation between hydraulically fractured coalbeds and a
USDW. The extent of fracturing is controlled by the characteristics of the geologic
formations (including the presence of shales or natural fractures), the volume and type of
fracturing fluid used, the pumping pressure, and the depth at which the fracturing is
performed. Several methods are available to operators to measure or predict the extent to
which fracture stimulation fluid moves and the related values of maximum induced
fracture extension and "propped" fracture height. Propped height (i.e., height in the
fracture to which proppant has been distributed) was found to be 60 percent to 75 percent
of total vertical fracture height (Mavor et al., 1991; Rahim and Holditch, 1992; Nolte and
Smith, 1981; Nolte and Economides, 1991; Zuber et al., 1991). Furthermore, in cases
where proppant "screens out" or emplacement partially fails, proppant may exist in 20
percent or less of fracture height.
Both the current and some older methods for estimating fracture dimensions are
discussed below. In general, these methods fall into three areas: direct measurements;
indirect measurements; and model estimates. Terminology in the literature regarding
fracture dimensions is sometimes inconsistent; some articles describe "measured"
fracture dimensions when referring to indirect measures or even model estimates.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-15
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
3.4.1 Direct Measurements
Direct measures include mined-through (or mineback) studies (where mining of
subsurface coal seams that were previously hydraulically fractured allows direct access to
fractures for measurement); dye tracing conducted in conjunction with mined-through
studies; downhole cameras (used to visually inspect fractures in the borehole), including
borehole image logging and downhole video logging; surface and downhole tiltmeters;
and microseismic monitoring (or imaging). Fracture geometry is most dependably
measured by microseismic monitoring or downhole tiltmeters (Warpinski, 2001), or by
tracers (Diamond and Oyler, 1987). Downhole cameras can be used only in open bore
holes (uncased wells), so fracture measurements using cameras do not reflect
conventional coalbed methane fracturing that typically occurs in cased wells. Both
downhole cameras and mined-through approaches to fracture measurements are limited
to areas exposed by the wellbore and mining activities, respectively. Nonetheless, the
mined-through studies provide the most direct approach for estimating fracture
dimensions.
Mined-Through Studies
Twenty-two coalbeds were hydraulically fractured, subsequently mined-through, and
investigated several months to several years later in Pennsylvania, Alabama, West
Virginia, Illinois, Virginia, and Utah (Diamond 1987a and b; Diamond and Oyler 1987).
Similar studies have been conducted by Jeffrey et al. (1993) in Queensland, Australia,
and Steidl (1991a; 1991b; 1993) in the Black Warrior Basin, Alabama. The Diamond
studies were designed to evaluate the effect of the hydraulic fracturing treatment on
mining safety. All the mined-through studies enabled direct observation of induced
fractures and surrounding material and evaluation of the movement of sand proppant and
fracturing fluids through both induced and natural fractures. Eight of the treatments
included fluorescent paint in the injected fluid to aid in mapping fluid movement
(Diamond 1987aandb).
Steidl (1993) found that fracture widths were typically 0.1 inch, but could be as wide as 4
inches. Measured sand-filled (propped) fractures were 2 to 526 feet in length (Steidl
1993, Jeffrey et al., 1993), although Steidl found a sand-free extension of a sand-filled
fracture 870 feet from the borehole. Diamond (1987a and b) found treatment fluids
beyond the sand-filled portions of the fractures using paint injected with the fracturing
fluids. In most of the wells where paint was injected, the paint was found 200 to 300 feet
beyond the sand-filled portions of fractures. However in one borehole, paint extended
out from the well bore for 630 feet, although the sand-filled portion of the fracture was
only 95 feet in extent (Diamond, 1987a and b). These paint-coated fractures were
produced using typical hydraulic fracturing processes in fairly typical coalbed methane
geologic conditions.
Fluorescent paint was observed in locations that indicated fluids did not travel in a direct
linear path from the induced fracture. Fluids often followed a stair-step pathway through
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-16
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
the coalbed (Diamond and Oyler, 1987). The fluorescent paint was also useful for
identifying small fractures penetrated by treatment fluids but not by sand proppant.
Multiple small, parallel fractures were penetrated by treatment fluids at many of the
locations studied. Given that treatment fluids have been documented to travel more than
six times farther than sand proppant, studies looking at the dimensions of sand-filled
fractures alone are unlikely to capture the extent of fluid movement within and beyond
coalbed methane reservoirs (Diamond, 1987a and b).
About half of the sites studied by Diamond (1987a and b) and Diamond and Oyler (1987)
had fractures penetrating beyond the coalbeds into the roof rock (the rock overlying the
coal in the mined areas). Jeffrey et al. (1993) found that most of the proppant in three of
their four treatments was found in the roof rock. Thus, mined-through studies in
Australia and in six states in the United States found that hydraulic fracturing fluids
penetrated into, and, when shales were very thin, through strata surrounding coalbeds in
50 percent of stimulations in the United States and 75 percent of the stimulations in
Australia. The mined-through studies, however, generally cannot provide measures of
how far the fractures actually extend, since mining did not extend beyond the coal and
into the roof rock.
Other Direct Measurements
A discussion of other fracture diagnostic methods is provided in DOE's paper "Hydraulic
Fracturing" (provided as Appendix A).
"Fracture diagnostics involves analyzing the data before, during and after
a hydraulic fracture treatment to determine the shape and dimensions of
both the created and propped fracture. Fracture diagnostic techniques
have been divided into several groups (Cipolla and Wright, 2000).
Direct far field techniques
Direct far field methods are comprised of tiltmeter fracture mapping and
microseismic fracture mapping techniques. These techniques require
delicate instrumentation that has to be emplaced in boreholes surrounding
and near the well to be fracture treated. When a hydraulic fracture is
created, the expansion of the fracture will cause the earth around the
fracture to deform. Tiltmeters can be used to measure the deformation and
to compute the approximate direction and size of the created fracture.
Surface tiltmeters are placed in shallow holes surrounding the well to be
fracture treated and are best for determining fracture orientation and
approximate size. Downhole tiltmeters are placed in vertical wells at
depths near the location of the zone to be fracture treated. As with surface
tiltmeters, downhole tiltmeter data can be analyzed to determine the
orientation and dimensions of the created fracture, but are most useful for
determining fracture height. Tiltmeters have been used on an
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-17
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
experimental basis to map hydraulic fractures in coal seams (Nielson and
Hanson, 1987).
Microseismic fracture mapping relies on using a downhole receiver array
of accelerometers or geophones to locate microseisms or micro-
earthquakes that are triggered by shear slippage in natural fractures
surrounding the hydraulic fracture. ... In essence, noise is created in a
zone surrounding the hydraulic fracture. Using sensitive arrays of
instruments, the noise can be monitored, recorded, analyzed and mapped.
...Microseismic monitoring has traditionally been too expensive to be
used on anything but research wells, but its cost has dropped dramatically
in the past few years, so although still expensive (on the order of $50,000
to $100,000), it is being used more commonly throughout the industry. ...
If the technology is used at the beginning of the development of a field,
however, the data and knowledge gained are often used on subsequent
wells, effectively spreading out the costs.
Direct near-wellbore techniques
Direct near-wellbore techniques are run in the well that is being fracture
treated to locate or image the portion of fracture that is very near (inches)
the wellbore. Direct near-wellbore techniques... [include] borehole image
logging [and] downhole video logging, and caliper logging. If a hydraulic
fracture intersects the wellbore, these direct near-wellbore techniques can
be of some benefit in locating the hydraulic fracture.
However, these near-wellbore techniques are not unique and cannot
supply information on the size or shape of the fracture once the fracture is
2-3 wellbore diameters in distance from the wellbore. In coal seams,
where multiple fractures are likely to exist, the reliability of these direct
near-wellbore techniques are even more speculative. As such, very few of
these direct near-wellbore techniques are used on a routine basis to look
for a hydraulic fracture."
3.4.2 Indirect Measurements
Indirect measures of fracture dimensions include pressure analyses (sometimes referred
to as net, treating, or bottom hole pressure analyses that are sometimes analyzed in
conjunction with proppant volume assessments) and radioactive tracing. (Radioactive
tracing can be conducted on either fracturing fluids or proppants. It is sometimes referred
to as a "tagged" study, and is typically measured through gamma ray logging.) Pressure
analyses generally monitor bottom hole pressures (BHPs) over time to infer fracture
propagation. For example, declining net pressure during water/gel pumping stages
indicates rapid fracture height growth (Saulsberry, et al., 1990). Proppant volumes and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-18
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
historical fracturing and methane production data are used to improve estimates based on
pressure analyses. Fracture heights and lengths that are inferred by pressure analyses are
commonly described in the literature as "measured." Radioactive tracers provide only
approximate estimates of fracture dimensions because they are measured in near-wellbore
environments.
3.4.3 Model Estimates
The other main category of indirect measures of fracture dimensions is hydraulic fracture
modeling. The basic elements of fracture modeling were developed between 1955 and
1961 (Nolte and Economides, 1991). Many modeling studies were conducted to aid in
the design of fracture stimulation treatments (i.e., to determine the volume and pump rate
of fluids and proppants that are required to achieve a desired fracture geometry).
Model estimates of fracture heights and lengths are common, including estimates using
three-dimensional (and quasi-three dimensional) models. Modeling capabilities have
advanced considerably in the last several 15 years, and the newest PSD (pseudo 3
dimensional) models simultaneously predict height, width, and length based on treatment
input data and reservoir parameters (Olson, 2001). A discussion of indirect fracture
modeling techniques is provided by DOE in the "Hydraulic Fracturing" paper (provided
as Appendix A). An excerpt from that paper is provided below.
"The indirect fracture techniques consist of hydraulic fracture modeling of net
pressures, pressure transient test analyses, and production data analyses. Because
the fracture treatment data and the post-fracture production data are normally
available on every well, the indirect fracture diagnostic techniques are the most
widely used methods to estimate the shape and dimensions of both the created and
the propped hydraulic fracture.
The fracture treatment data can be analyzed with a PSD fracture propagation
model to determine the shape and dimensions of the created fracture. The PSD
model is used to history match the fracturing data, such as injection rates and
injection pressures. Input data, such as the in-situ stress and permeability in key
layers of rock can be varied (within reason) to achieve a history match of the field
data.
Post-fracture production and pressure data can be analyzed using a 3D reservoir
simulator to estimate the shape and dimensions of the propped fracture. Values of
formation permeability, fracture length and fracture conductivity can be varied in
the reservoir model to achieve a history match of the field data.
The main limitation of these indirect techniques is that the solutions are not
unique and require as much fixed data as possible. For example, if the engineer
has determined the formation permeability from a well test or production test
prior to the fracture treatment, so that the value of formation permeability is
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-19
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
known and can be fixed in the models, the solution concerning values of fracture
length become more unique. Most of the information in the literature concerning
post-fracture analyses of hydraulic fractures has been derived from these indirect
fracture diagnostic techniques."
There are several caveats regarding the use and interpretation of model estimates. In-situ
stress values of the target coal seams and surrounding strata are important model inputs.
Actual in-situ stress measurements are very difficult to obtain and are rarely conducted
(Warpinski, 2001). Therefore, almost all modeling is conducted using inferred stress
values (as estimated, for example, from the mechanical and lithological properties of
rocks from, or similar to those in, the target coal zone). Given the geologic variability
and site-specific influences on fracture behavior described above, the reliability of
fracture height and length estimates obtained from various models is obviously
influenced by the quality of the inferred model inputs regarding geologic factors.
Models also necessarily rely on simplifying assumptions to simulate fracture propagation
and behavior through sometimes complex geologic zones. As with all modeling, the
reliance on inferred input variables and some assumptions introduces some subjectivity to
the modeling process. Dependable modeling requires knowledge of and allowance for
the detailed stratigraphy of the geologic strata throughout the coal zone. (It was noted in
section 3.2.2 that thin clay layers or ash beds can influence fracture behavior.) Simplified
geologic models might represent the subsurface as 2 to 3 distinct geologic layers, to
reduce computing and data requirements, when a 30- or 50-layer model may be necessary
to accurately predict fracture height (Rahim et al., 1998). Nevertheless, models are
necessary simplifications of fracture behavior in the geologic subsurface, and significant
research has been conducted in the last several decades so that model estimates of
fracture behavior in methane-producing coalbeds are now an invaluable tool for industry.
3.4.4 Limitations of Fracture Diagnostic Techniques
Warpinski (1996) discussed many of these same fracture diagnostic techniques. In
general, the best fracture diagnostics techniques are expensive and used only in research
wells. Fracture diagnostic techniques can provide important data when entering a new
production area or a new formation. However, for coalbed methane wells, where costs
must be minimized to maintain profitability, the best fracture diagnostic techniques are
rarely used and are often considered to be prohibitively expensive.
Warpinksi (2001) further provided other general conclusions regarding estimates of
fracture dimensions:
• Fracture heights inferred from pressure data are almost always greater than the
corresponding heights measured with the more dependable microseismic
monitoring or tiltmeters.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-20
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
• Actual fracture lengths may be greater or less than the lengths estimated from
models or inferred from pressure analyses, depending on many site-specific
geologic factors.
• Fracture geometry can be accurately measured using microseismic monitoring
and measured somewhat using downhole tiltmeters. These technologies have
been found to be invaluable for determining how fractures actually behave.
Table 3-1 lists certain diagnostic techniques and their limitations.
Table 3-1. Limitations of Fracture Diagnostic Techniques (Appendix A: DOE,
Hydraulic Fracturing)
Parameter
Fracture Height
Fracture Height
Fracture Height
Fracture Height
Fracture Height
Fracture Height
Fracture Length
Fracture Length
Fracture Length
Fracture Length
Fracture Azimuth
Fracture Azimuth
Fracture Azimuth
Fracture Azimuth
Technique
Tracer logs
Temperature logs
Stress profiling
PSD models
Microseismic
Tiltmeters
PSD models
Well testing
Microseismic
Tiltmeters
Core techniques
Log techniques
Microseismic
Tiltmeters
Limitation
Shallow depth of investigation; shows height only near the wellbore
Difficult to interpret; shallow depth of investigation; shows height only near wellbore
Does not measure fracture directly; must be calibrated with in-situ stress tests
Does not measure fracture directly; estimates vary depending on which model is used
Optimally requires nearby offset well; difficult to interpret; expensive
Difficult to interpret; expensive and difficult to conduct in the field
Length inferred, not measured; estimates vary greatly depending on which model is
used
Large uncertainties depending upon assumptions and lack of prefracture well test data
Optimally requires nearby offset well; difficult to interpret; expensive
Difficult to interpret; expensive and difficult to conduct in the field
Expensive to cut core and run tests; multiple tests must be run to assure accuracy
Requires open hole logs to be run; does not work if natural fractures are not present
Analysis intensive; expensive for determination of azimuth
Useful only to a depth of 5,000 feet; requires access to large area; expensive
From: Appendix A, DOE, Hydraulic Fracturing
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-21
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
3.5 Summary
Coalbed methane development began as a safety measure to extract methane, an
explosion hazard, from coal prior to mining. Since 1980, coalbed methane production
has grown rapidly, spurred by tax incentives to develop non-conventional energy
production. At the end of 2000, coalbed methane production from 13 states totaled 1.353
trillion cubic feet, an increase of 156 percent from 1992. At year-end 2000, coalbed
methane production accounted for about 7 percent of the total United States dry gas
production and 9 percent of proven dry gas reserves (EIA, 2001).
Methane within coalbeds is not "trapped" under pressure as in conventional gas
scenarios. Only about 5 to 9 percent of the methane is present as "free" gas within the
natural fractures, joints, and cleats. Almost all coalbed methane is adsorbed within the
micro-porous matrix of the coal (Koenig, 1989; Winston, 1990; Close, 1993).
Coalbed methane production starts with high-pressure injection of fracturing fluids and
proppant into targeted coal zones. The resulting induced or enlarged fractures improve
the connections of the production well to the fracture networks in and around the coal
zone. When production begins, water is pumped from the fractures in the coal zone to
reduce pressure in the formation. When pressures are adequately reduced, methane
desorbs from the coal matrix, moves through the network of induced and natural fractures
in the coal toward the production well, and is extracted through the well and to the
surface.
Fractures that are created at shallow depths (less than approximately 1,000 feet) typically
have more of a horizontal than a vertical component. Vertical fractures created at deeper
depths can propagate vertically to shallower depths where they may develop a horizontal
component. These "T-fractures" may involve the fracture "turning" and developing
horizontally at a coalbed-mudstone interface.
Fracture behavior through coal, shale, and other geologic strata commonly present in coal
zones depends on site-specific factors such as relative thicknesses and in-situ stress
differences between the target coal seam(s) and the surrounding geologic strata, as well
as the presence of pre-existing natural fractures. Often, a high stress contrast between
adjacent geologic strata results in a barrier to fracture propagation. This occurs in coal
zones where there is a geologic contact between a high-stress coal seam and an overlying,
thick, relatively low-stress shale.
The fluids used for fracture development are injected at high pressure into the targeted
coal zone in the subsurface. These fluids may be "clear" (primarily consisting of water,
but may include acid, oil, or water with friction-reducer additives) or "gelled" (viscosity-
modified water using guar or other gelling agents). Hydraulic fracturing in coalbed
methane wells may require 50,000 to 350,000 gallons of fracturing fluids and 75,000 to
320,000 pounds of sand as proppant to prop or maintain the opening of fractures after the
injection (fracturing) pressure is reduced (Holditch et al., 1988 and 1989; Jeu et al., 1988;
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-22
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
Hinkel etal., 1991; Holditch, 1993; Palmer et al., 1991b, 1993a, and 1993b). More
typical injection volumes, based on average injection volume data provided by
Halliburton for six coalbed methane locations, indicate a maximum average injection
volume of 150,000 gal/well and a median average injection volume of 57,500 gal/well
(Halliburton, Inc., 2003).
In any fracturing job, some fracturing fluids cannot be recovered and are said to be "lost"
to the formation. Palmer (199la) observed that for fracture stimulations in multi-layered
coal formations, 61 percent of stimulation fluids were recovered during a 19-day
production sampling of a coalbed methane well in the Black Warrior Basin. He further
estimated that from 68 percent to possibly as much as 82 percent would eventually be
recovered. A variety of site-specific factors, including leakoff into the coal seams and
surrounding strata, the check-valve effect, adsorption and other geochemical processes,
and flow through the hydraulic fracture beyond the well's capture zone will serve to
reduce recovery of hydraulic fracturing fluids injected into subsurface coal zones to
promote coalbed methane extraction.
The mined-through studies by the U.S. Bureau of Mines (see Diamond, 1987a and b) and
others provide important directly-measured characteristics of hydraulic fracturing in coal
seams and surrounding strata. Further, paint tracer studies conducted as part of
Diamond's (1987a and b) mined-through studies can provide estimates on the extent of
hydraulic fracturing fluid movement, which may be greater than the extent of sand-filled
(propped) hydraulic fracture heights or lengths given fluid movement through natural
fractures. These estimates of the extent of fluid movement are usually limited by the area
exposed to mining.
A significant amount of diagnostic research has been conducted in the last decade
enabling industry to develop a practical, applied understanding of general fracture
behavior as it relates to methane production. Operators use a number of techniques to
estimate fracture dimensions to design fracture stimulation treatments to minimize
expenditures on hydraulic horsepower, fracturing fluids, and proppants. Modeling is
increasingly more sophisticated, but still commonly depends on at least some inferred
(and subjective) input data. Reliable fracture height and length can be measured
accurately by microseismic monitoring and tiltmeters (Warpinksi, 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-23
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-1. Major United States Coal Basins
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-24
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-2. Geography of an Ancient Peat-Forming System
•***-3ft- -' ^-n^st **• '^1»* •:>/=• w -X= - 4™-- ~-
GitKjraphy -af an Ancient Peat-fanning Sjslt-m (e*am\itii torn Biaefc.''Arar»ci Basin, Alabama)
(P aslim a IK! H infc c. 1 fl a,' i
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-25
image:
EPA 816-R-04-003 Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-3. Schematic Representation of "Face Cleat" (F) and "Butt Cleat" (B)
v,
X
F
\
V"
.
X* -jf r-~ ~~ "'». •-•'""" i
, .-->' \^""""-'B "" •
"
•. -,.
V '
O O,5 In
Schematic Kepfcsenlaliai) af 't-am C«at' (\-1 and 'Bill C eal' (B) (Ayas ct a:., 199*1
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 3-26
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-4. A Graphical Representation of the Hydraulic Fracturing Process in
Coalbed Methane Wells
FratftBto p HuKJ Injection
- Fluid fjuaai .a pmisu* buUup mai
*| 1 j.-^r^lHS III* 1f*e»lll* *
JUT cufirllwl!" iij IX1 'irecUon
f radar* J-'n>jiaQ SHor1:
Z. ' kid nrolnly inljimtH h ** dractlan a" ttn
UcOtd llfll [anylna a prcppMtl
' p, - ~ 1 1-~ *t~^ '•> i ""^'in r~' pr jT
ilu Tiicum apon. rmckn prcoajiicfl
ara ct» ccnMnuam pr ncsni.
arc Frcppjnt Irj&rtlwi
I
liildc Wntar IT afec a«1^irf sd to TOIUGB thn
pnavuni In til ^rnvahmi 1-3 (hal g» flaw am ccrnrwxn
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-27
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-4. A Graphical Representation of the Hydraulic Fracturing Process in
Coalbed Methane Wells (Continued)
Wrthane
Mdhww Productan
B.TtnfiMftaBdt process (names *c
.in.l fiii-Ai'lirl'Uvirty
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-28
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-5. Water And Gas Production Over Time
c
o
2
0_
Gas
Water
Time
Caa Production Over THne lSauisaer?v el tn.. 19861
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-29
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-6. Side-View of a Vertical Hydraulic Fracture Typical of Coalbeds
S;sdff-V"ww of a Vwlirat Hydraulic. Fracture Typtcal of
hytlrai.lir. Ir*dijr«s pflnfirato Mxiara1 cruil aaanw,
(Palmer etal..
3-6
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-30
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-7. Plan View (Looking Down the Wellbore) of Vertical, Two-Winged
Coalbed Methane Fracture Showing the Reservoir Region Invaded by Fracturing
Fluid Leakoff
\
X
L = fracture half length
L = 150 to 300 feet
Leakoff
Reservoir Region Q
D = 50 to 100ft
\
\
Plan View of a Vertical, Two-Winged Coalbed Methane Fracture Showing the Reservoir Region Invaded
by Fracturing Fluid Leakoff (Palmer et al., 1991).
Figure 3-7
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-31
image:
EPA816-R-04-003
Chapter 3
Characteristics of CBM Production and Associated HF Practices
Figure 3-8. Plan View of a Vertical Hydraulic Fracture
Perforation / Fracture
Junction
Perforation
Cement
Vertical
Fracture
C
ear-Wellbore
onstriction
i of .5 V^
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
3-32
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
Chapter 4
Hydraulic Fracturing Fluids
This chapter summarizes the information EPA collected on the types and volumes of
fracturing fluids and additives that may be used for hydraulic fracturing of coalbed
methane wells. This chapter also provides EPA's evaluation of the fate and transport of
fracturing fluids that are injected into targeted coal layers during the hydraulic fracturing
process. This evaluation was conducted to provide the Agency with information on
whether a Phase II study is warranted. Captioned photographs in this chapter show the
use of fracturing fluids at a coalbed methane well (Figures 4-1 through 4-11 at the end of
this chapter).
4.1 Introduction
The types and use of fracturing fluids have evolved greatly over the past 60 years and
continue to evolve. The U.S. oil and gas industry has used fluids for fracturing geologic
formations since the early 1940s (Ely, 1985). The Handbook of Stimulation Engineering
(Ely, 1985), a comprehensive history of the evolution of hydraulic fracturing fluids in the
oil and gas industry, was used as a source of information for this chapter. In addition,
EPA identified fluids and fluid additives commonly used in hydraulic fracturing through
literature searches, reviews of relevant MSDSs provided by service companies, and
discussions with field engineers, service company chemists, and state and federal
employees.
Available scientific literature indicates that hydraulic fracturing fluid performance
became a prevalent research topic in the late 1980s and the 1990s. Most of the literature
pertaining to fracturing fluids relates to the fluids' operational efficiency rather than their
potential environmental or human health impacts. There is very little documented
research on the environmental impacts that result from the injection and migration of
these fluids into subsurface formations, soils, and USDWs. Some of the existing
literature does offer information regarding the basic chemical components present in
most of these fluids. The composition of fracturing fluids and additives is discussed in
detail in the next section.
The main goal of coalbed hydraulic fracturing is to create a highly conductive fracture
system that will allow flow through the methane-bearing coal zone to the production well
used to extract methane (and groundwater). Hydraulic fracturing fluids are used to
initiate and/or expand fractures, as well as to transport proppant into fractures in coalbed
formations. Proppants are sand or other granular substances injected into the formation
to hold or "prop" open coal formation fractures created by hydraulic fracturing. The
viscosity of fracturing fluids is considered when they are formulated, to provide for
efficient transport and placement of proppant into a fracture. Most of the fracturing
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-1
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
fluids injected into the formation are pumped back out of the well along with
groundwater and methane gas (see section 3.3 in Chapter 3 for a more detailed discussion
of fracturing fluid recovery).
4.2 Types of Fracturing Fluids and Additives
Service companies have developed a number of different oil- and water-based fluids and
treatments to more efficiently induce and maintain permeable and productive fractures.
The composition of these fluids varies significantly, from simple water and sand to
complex polymeric substances with a multitude of additives. Each type of fracturing
fluid has unique characteristics, and each possesses its own positive and negative
performance traits. For ideal performance, fracturing fluids should possess the following
four qualities (adapted from Powell et al., 1999):
• Be viscous enough to create a fracture of adequate width.
• Maximize fluid travel distance to extend fracture length.
• Be able to transport large amounts of proppant into the fracture.
• Require minimal gelling agent to allow for easier degradation or "breaking"
and reduced cost.
Water-based fracturing fluids have become the predominant type of coalbed methane
fracturing fluid (Appendix A: DOE, Hydraulic Fracturing). However, fracturing fluids
can also be based on oil, methanol, or a combination of water and methanol. Methanol is
used in lieu of, or in conjunction with, water to minimize fracturing fluid leakoff and
enhance fluid recovery (Thompson et al., 1991). Polymer-based fracturing fluids made
with methanol usually improve fracturing results, but require 50 to 100 times the amount
of breaker (e.g., acids used to degrade the fracturing fluid viscosity, which helps to
enhance post-fracturing fluid recovery) (Ely, 1985). In some cases, nitrogen or carbon
dioxide gas is combined with the fracturing fluids to form foam as the base fluid. Foams
require substantially lower volumes to transport an equivalent amount of proppant.
Diesel fuel is another component of some fracturing fluids although it is not used as an
additive in all hydraulic fracturing operations. A variety of other fluid additives (in
addition to the proppants) may be included in the fracturing fluid mixture to perform
essential tasks such as formation clean up, foam stabilization, leakoff inhibition, or
surface tension reduction. These additives include biocides, fluid-loss agents, enzyme
breakers, acid breakers, oxidizing breakers, friction reducers, and surfactants such as
emulsifiers and non-emulsifiers. Several products may exist in each of these categories.
On any one fracturing job, different fluids may be used in combination or alone at
different stages in the fracturing process. Experienced service company engineers will
devise the most effective fracturing scheme, based on formation characteristics, using the
fracturing fluid combination they deem most effective.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-2
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
The main fluid categories are:
• Gelled fluids, including linear or cross-linked gels.
• Foamed gels.
• Plain water and potassium chloride (KC1) water.
• Acids.
• Combination treatments (any combination of 2 or more of the aforementioned
fluids).
Some of the fluids and fluid additives may contain constituents of potential concern.
Table 4-1, at the end of section 4.2.6, lists examples of chemicals found in hydraulic
fracturing fluids according to the MSDSs provided by service companies, and potential
human health effects associated with the product. It is important to note that information
presented in MSDSs is for pure product. Each of the products listed in Table 4-1 is
significantly diluted prior to injection.
EPA also obtained two environmental impact statements that were prepared by the
Bureau of Land Management (BLM). In these impact statements, BLM identified
additional chemical compounds that may be in fracturing fluids including methyl tert
butyl ether (MTBE) (U.S. Department of the Interior, CO State BLM, 1998).
However, EPA was unable to find any indications in the literature, on MSDSs, or in
interviews with service companies that MTBE is used in fracturing fluids to stimulate
coalbed methane wells.
4.2.1 Gelled Fluids
Water alone is not always adequate for fracturing certain formations because its low
viscosity limits its ability to transport proppant. In response to this problem, the industry
developed linear and cross-linked fluids, which are higher viscosity fracturing fluids.
Water gellants or thickeners are used to create these gelled fluids. Gellant selection is
based on formation characteristics such as pressure, temperature, permeability, porosity,
and zone thickness. These gelled fluids are described in more detail below.
Linear Gels
A substantial number of fracturing treatments are completed using thickened, water-
based linear gels. The gelling agents used in these fracturing fluids are typically guar
gum, guar derivatives such as hydroxypropylguar (HPG) and
carboxymethylhydroxypropylguar (CMHPG), or cellulose derivatives such as
carboxymethylguar or hydroxyethylcellulose (HEC). In general, these products are
biodegradable. Guar is a polymeric substance derived from the seed of the guar plant
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-3
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
(Ely, 1985). Guar gum, on its own, is non-toxic and, in fact, is a food-grade product
commonly used to increase the viscosity and elasticity of foods such as ice cream.
To formulate a viscous fracturing gel, guar powder or concentrate is dissolved in a carrier
fluid such as water or diesel fuel. Increased viscosity improves the ability of the
fracturing fluid to transport proppant and decreases the need for more turbulent flow.
Concentrations of guar gelling agents within fracturing fluids have decreased over the
past several years. It was determined that reduced concentrations provide better and
more complete fractures (Powell et al., 1999).
Diesel fuel has been frequently used in lieu of water to dissolve the guar powder because
its carrying capacity per unit volume is much higher (Halliburton, Inc., 2002). "Diesel is
a common solvent additive, especially in liquid gel concentrates, used by many service
companies for continuous delivery of gelling agents in fracturing treatments" (GRI,
1996). Diesel does not enhance the efficiency of the fracturing fluid; it is merely a
component of the delivery system (Halliburton, Inc., 2002). Using diesel instead of water
minimizes the number of transport vehicles needed to carry the liquid gel to the site
(Halliburton, Inc., 2002).
The percentage of diesel fuel in the slurried thickener can range between 30 percent and
almost 100 percent, based on the MSDSs summarized in Table 4-1. Diesel fuel is a
petroleum distillate and may contain known carcinogens. One such component of diesel
fuel is benzene, which, according to literature sources, can make up anywhere between
0.003 percent and 0.1 percent by weight of diesel fuel (Clark and Brown, 1977; R.
Morrison & Associates, Inc., 2001). Slurried diesel and gel are diluted with water prior
to injection into the subsurface. The dilution is approximately 4 to 10 gallons of
concentrated liquid gel (guar slurried in diesel) per 1,000 gallons of make-up water to
produce an adequate polymer slurry (Halliburton, Inc., Virginia Site Visit, 2001;
Schlumberger, Ltd., 2001; Consolidated Industrial Services, Inc., Virginia Site Visit,
2001; BJ Services, 2001).
Cross-linked Gels
One major advance in fracturing fluid technology was the development of cross-linked
gels. The first cross-linked gels were developed in 1968 (Ely, 1985). When cross-
linking agents are added to linear gels, the result is a complex, high-viscosity fracturing
fluid that provides higher proppant transport performance than do linear gels (Messina,
Inc. Web site, 2001; Ely, 1985; Halliburton Inc., Virginia Site Visit, 2001). Cross-linking
reduces the need for fluid thickener and extends the viscous life of the fluid indefinitely.
The fracturing fluid remains viscous until a breaking agent is introduced to break the
cross-linker and, eventually, the polymer. Although cross-linkers make the fluid more
expensive, they can considerably improve hydraulic fracturing performance, hence
increasing coalbed methane well production rates.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-4
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
Cross-linked gels are typically metal ion-cross-linked guar (Ely, 1985). Service
companies have used metal ions such as chromium, aluminum, titanium, and other metal
ions to achieve cross-linking (Ely, 1985). In 1973, low-residue (cleaner) forms of cross-
linked gels, such as cross-linked hydroxypropylguar, were developed (Ely, 1985).
According to MSDSs summarized in Table 4-1, cross-linked gels may contain boric acid,
sodium tetraborate decahydrate, ethylene glycol, and monoethylamine. These
constituents are hazardous in their undiluted form and can cause kidney, liver, heart,
blood, and brain damage through prolonged or repeated exposure. According to a BLM
environmental impact statement, cross-linkers may contain hazardous constituents such
as ammonium chloride, potassium hydroxide, zirconium nitrate, and zirconium sulfate
(U.S. Department of the Interior, CO State BLM, 1998). Concentrations of these
compounds in the fracturing fluids were not reported in the impact statement. The final
concentration of cross-linkers is typically 1 to 2 gallons of cross-linker per 1,000 gallons
of gel (Halliburton, Inc., Virginia Site Visit, 2001; Schlumberger, Ltd., 2001).
4.2.2 Foamed Gels
Foam fracturing technology uses foam bubbles to transport and place proppant into
fractures. The most widely used foam fracturing fluids employ nitrogen or carbon
dioxide as their base gas. Incorporating inert gases with foaming agents and water
reduces the amount of fracturing liquid required. Foamed gels use fracturing fluids with
higher proppant concentrations to achieve highly effective fracturing. The gas bubbles in
the foam fill voids that would otherwise be filled by fracturing fluid. The high
concentrations of proppant allow for an approximately 75-percent reduction in the overall
amount of fluid that would be necessary using a conventional linear or cross-linked gel
(Ely, 1985; Halliburton, Inc., Virginia Site Visit, 2001). Foaming agents can be used in
conjunction with gelled fluids to achieve an extremely effective fracturing fluid
(Halliburton, Inc., Virginia Site Visit, 2001).
Foam emulsions experience high leakoff; therefore, typical protocol involves the addition
of fluid-loss agents, such as fine sands (Ely, 1985; Halliburton, Virginia Site Visit, 2001).
Foaming agents suspend air, nitrogen, or carbon dioxide within the aqueous phase of a
fracturing treatment. The gas/liquid ratio determines if a fluid will be true foam or
simply a gas-energized liquid (Ely, 1985). Carbon dioxide can be injected as a liquid,
whereas nitrogen must be injected as a gas to prevent freezing (Halliburton, Inc., Virginia
Site Visit, 2001).
According to the MSDSs summarized in Table 4-1, foaming agents can contain
diethanolamine and alcohols such as isopropanol, ethanol, and 2-butoxyethanol. They
can also contain hazardous substances including glycol ethers (U.S. Department of the
Interior, CO State BLM, 1998). One of the foaming agent products listed in Table 4-1
can cause negative liver and kidney effects, although the actual component causing these
effects is not specified on the MSDS. The final concentration is typically 3 gallons of
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-5
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
foamer per 1,000 gallons of gel (Halliburton, Inc., Virginia Site Visit, 2001;
Schlumberger, Ltd., 2001).
4.2.3 Water & Potassium Chloride Water Treatments
Many service companies use groundwater pumped directly from the formation or treated
water for their fracturing jobs. In some coalbed methane well stimulations, proppants are
not needed to prop fractures open, so simple water or slightly thickened water can be a
cost-effective substitute for an expensive polymer or foam-based fracturing fluid with
proppant (Ely, 1985). Hydraulic fracturing performance is not exceptional with plain
water, but, in some cases, the production rates achieved are adequate. Plain water has a
lower viscosity than gelled water, which reduces proppant transport capacity.
Similar to plain water, another fracturing fluid uses water with potassium chloride (KC1)
in addition to small quantities of gelling agents, polymers, and/or surfactants (Ely, 1985).
Potassium chloride is harmless if ingested at low concentrations.
4.2.4 Acids
Acids are used in limestone formations that overlay or are interbedded within coals to
dissolve the rock and create a conduit through which formation water and coalbed
methane can travel (Ely, 1985). Typically, the acidic stimulation fluid is hydrochloric
acid or a combination of hydrochloric and acetic or formic acid. For acid fracturing to be
successful, thousands of gallons of acid must be pumped far into the formation to etch the
face of the fracture (Ely, 1985). Some of the cellulose derivatives used as gelling agents
in water and water/methanol fluids can be used in acidic fluids to increase treatment
distance (Ely, 1985). As discussed in section 4.2.5, acids may also be used as a
component of breaker fluids.
In addition, acid can be used to clean up perforations of the cement surrounding the well
casing prior to fracturing fluid injection (Halliburton, Inc., Virginia Site Visit, 2001;
Halliburton, Inc., 2002). The cement is perforated at the desired zone of injection to ease
fracturing fluid flow into the formation (Halliburton, Inc., Virginia Site Visit, 2001;
Halliburton, Inc., 2002).
Table 4-1 provides information on formic and hydrochloric acids. Acids are corrosive,
and can be extremely hazardous in concentrated form. Acids are substantially diluted
with water-based or water-and-gas-based fluids prior to injection into the subsurface.
The injected concentration is typically 1,000 times weaker than the concentrated versions
presented in the product MSDSs (Halliburton, Inc., Virginia Site Visit, 2001;
Schlumberger, Ltd., 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-6
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
4.2.5 Fluid Additives
Several fluid additives have been developed to enhance the efficiency and increase the
success of fracturing fluid treatments. The major categories of these additives are defined
and briefly described in the following sections.
Breakers
Breaker fluids are used to degrade the fracturing fluid viscosity, which helps to enhance
post-fracturing fluid recovery, or flowback. Breakers can be mixed with the fracturing
fluid during pumping, or they can be introduced later as an independent fluid. There are
a variety of breaker types including time-release and temperature-dependent types. Most
breakers are typically acids, oxidizers, or enzymes (Messina, Inc. Web site, 2001).
According to a BLM environmental impact statement, breakers may contain hazardous
constituents, including ammonium persulfate, ammonium sulphate, copper compounds,
ethylene glycol, and glycol ethers (U.S. Department of the Interior, CO State BLM,
1998). Concentrations of these compounds in the fracturing fluids were not presented in
the environmental impact statement.
Biocides
One hydraulic fracturing design problem that arises when using organic polymers in
fracturing fluids is the incidence of bacterial growth within the fluids. Due to the
presence of organic constituents, the fracturing fluids provide a medium for bacterial
growth. As the bacteria grow, they secrete enzymes that break down the gelling agent,
which reduces the viscosity of the fracturing fluid. Reduced viscosity translates into poor
proppant placement and poor fracturing performance. To alleviate this degradation in
performance, biocides, bactericides, or microbicides are added to the mixing tanks with
the polymeric gelling agents to kill any existing microorganisms (e.g., sulfate-reducing
bacteria, slime-forming bacteria, algae), and to inhibit bacterial growth and deleterious
enzyme production. Bactericides are typically hazardous by nature (Messina, Inc. Web
site, 2001). They may contain hazardous constituents, including poly cyclic organic
matter (POM) and polynuclear aromatic hydrocarbons (PAHs) (U.S. Department of the
Interior, CO State BLM, 1998).
Information from MSDSs for a biocide and a microbicide is summarized in Table 4-1.
These concentrated products are substantially diluted prior to injection into the
subsurface. Typical dilution in the make-up water is 0.1 to 0.2 gallons of microbicide in
1,000 gallons of water (Halliburton, Inc., Virginia Site Visit, 2001; Schlumberger, Ltd.,
2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-7
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
Fluid-Loss Additives
Fluid-loss additives restrict leakoff of the fracturing fluid into the exposed rock at the
fracture face. Because the additives prevent excessive leakoff, fracturing fluid
effectiveness and integrity are maintained. Fluid-loss additives of the past and present
include bridging materials such as 100 mesh sand, 100 mesh soluble resin, and silica
flour, or plastering materials such as starch blends, talc silica flour, and clay (Ely, 1985).
Friction Reducers
To optimize the fracturing process, water-based fluids must be pumped at maximum rates
and fluids must be injected at maximum pressures. Increasing flow velocities and
pressures in this manner can lead to undesirable levels of friction within the injection well
and the fracture itself. In order to minimize friction, friction reducers are added to water-
based fracturing fluids. The friction reducers are typically latex polymers or copolymers
of acrylamides. They are added to slick water treatments (water with solvent) at
concentrations of 0.25 to 2.0 pounds per 1,000 gallons (Ely, 1985). Some examples of
friction reducers are oil-soluble anionic liquid, cationic polyacrilate liquid, and cationic
friction reducer (Messina, Inc. Web site, 2001).
Acid Corrosion Inhibitors
Corrosion inhibitors are required in acid fluid mixtures because acids will corrode steel
tubing, well casings, tools, and tanks. The solvent acetone is a common additive in
corrosion inhibitors (GRI, 1996). Information from MSDSs for acid inhibitors is
summarized in Table 4-1. These products can affect the liver, kidney, heart, central
nervous system, and lungs. They are quite hazardous in their undiluted form. These
products are diluted to a concentration of 1 gallon per 1,000 gallons of make-up water
and acid mixture (Halliburton, Inc., Virginia Site Visit, 2001; Schlumberger, Ltd., 2001).
Acids and acid corrosion inhibitors are used in very small quantities in coalbed methane
fracturing operations (500 to 2,000 gallons per treatment).
4.2.6 Proppants
The purpose of a proppant is to prop open a hydraulic fracture. An ideal proppant should
produce maximum permeability in a fracture. Fracture permeability is a function of
proppant grain roundness, proppant purity, and crush strength. Larger proppant volumes
allow for wider fractures, which facilitate more rapid flowback to the production well.
Over a period of 30 minutes, 4,500 to 15,000 gallons of fracturing fluid will typically
transport and place approximately 11,000 to 25,000 pounds of proppant into the fracture
(Powell et al., 1999).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-8
image:
EL a EL
a- 3. e
2. B- C
a i a
—
i!
a c0
B O
OQ B
O O
l-b ft
c/>
Tililf J-|: S."ltjr»--lfriiilo i.lUuliirtwl t'H-diii-ih F-ini h Mv/riillr hrsi-iiirins Huii,Ut"l.is*«f ur.
itlnrrv wa-tni
Wsdw
IffM
ftoottf Jp'tt
Praniij jjprtt
IqrdrocUorit
f-i-Ji * M
Chcntcjl
j-i s'jtjiR
I' IlJ-.M.^ **. EluLi.
i;-10 J** b» *- . E^a',1
','-!'*, b« me 5»l« »t>
% r <r
Will
•
its
I M.r ilte*.
r«-*t
SiH,*!1
t B»,t >1 U ili
ion ,".;|i.i. ; il- iiiuii • i
BCJ^,*,FI -..'fiont,
1^ hi rrs
n it .i
..«-j. 4 "
s rmj* t tr*t»-
S-£«. al ill n*!l a
Elttitil
!• i' ir-irn,«
oo
K^
ON
3
o
OQ
3
S
image:
EL a EL
a- 3. e
2. B- C
a i a
3. oo
B o
OQ £
o ri
Pr.dnd
IHrtCjJc
T ihlt 4-1: CkiracKrisdei rf tndiliatd Ckiricili F-wnl li ftrfradic Ft*:1nnnj Fluidi 1^4wd <•» MSDSt.i
ChfDrjl ( niipiaKuD la
a 4 *, Ijj x
-J J
Tta,i«4))|i( J lafortnad-jn"
» M»T •%«« fji Mi.ia ittsiln
v hi i<i»; r1 r«l«'*-j
ii4tj r'Ul («f mi 31 ia a>
teats
ff Jj "^ iff
tew. tin rHkfr *i 'KB* 4zsns»
»TltK" AjW,
* t
oo
K^
ON
3
o
OQ
3
S
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
4.3 The Fate and Transport of Stimulation Fluids Injected into Coal and
Surrounding Rock During Hydraulic Fracturing of Coalbed Methane
Reservoirs (with a Special Focus on Diesel Fuel)
Diesel fuel is sometimes a component of gelled fluids. Diesel fuel contains constituents of
potential concern regulated under SDWA - benzene, toluene, ethylbenzene, and xylenes
(i.e., BTEX compounds). The use of diesel fuel in fracturing fluids poses the greatest
threat to USDWs because BTEX compounds in diesel fuel exceed the MCL at the point-of-
injection (i.e. the subsurface location where fracturing fluids are initially injected).
The remainder of this section presents EPA's qualitative evaluation of the fate and
transport of fracturing fluids injected into targeted coal layers in the subsurface during
hydraulic fracturing. Although EPA's MO A with the three major service companies has
largely eliminated diesel fuel from fracturing fluids injected directly into any USDWs,
there may still be rare instances in which diesel fuel is used by other service companies or
operators (USEPA, 2003). Therefore an evaluation of the use of diesel fuel in fracturing
fluid, which also provides follow-up on the draft of this report published in August, 2002,
is included in this chapter.
EPA revised its procedure for assessing the potential effects of fracturing fluid constituents
on USDWs from the procedure presented in the August 2002 draft of this report as follows:
• EPA has revised the fraction of BTEX compounds in diesel used to estimate the
point-of-injection concentrations from a single value to a documented broader
range of values for the fraction of BTEX in diesel fuel. For example, the
fraction of benzene in diesel was revised from 0.00006gbenzene/gdiesei to a range
with a minimum value of 0.000026 gbenzene/gdiesei and a maximum value of 0.001
gbenzene/gdiesei- If the maximum value for benzene in diesel is used to estimate
the concentration of benzene at the point-of-injection, the resulting estimate is
17 times higher than that presented in the Draft Report.
• In this report, EPA used more current values for two of the parameters used to
estimate the point-of-injection concentrations of BTEX compounds.
Specifically, the estimates in this report use a density of the diesel fuel-gel
mixture of 0.87 g/mL compared to 0.84 g/mL in the Draft Report, and a fraction
of diesel fuel in gel of 0.60 gdiesei/ggei compared to 0.52 gdiesei/ggei in the Draft
Report. The use of these more current values does not affect the order of
magnitude of the revised point-of-injection calculations.
• The August 2002 Draft Report included estimates of the concentration of
benzene at an idealized, hypothetical edge of the fracture zone located 100 feet
from the point-of-injection. Based on new information and stakeholder input,
EPA concluded that the edge of fracture zone calculation is not an appropriate
model for reasons including:
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-11
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
Mined-through studies reviewed by EPA indicated that hydraulic
fracturing injection fluids had traveled several hundred feet beyond the
point-of-injection.
The assumption of well-mixed concentrations within the idealized
fracture zone is insufficient. One mined-through study indicated an
observed concentration of gel in a fracture that was 15 times the injected
concentration, with gel found to be hanging in stringy clumps in many
fractures. The variability in gel distribution in hydraulic fractures
indicates that the gel constituents are unlikely to be well mixed in
groundwater.
Based on more extensive review of the literature, the width of a typical
fracture was estimated to be much thinner than that used in the Draft
Report (0.1 inch versus 2 inches). The impact of the reduced width of a
typical fracture is that the calculated volume of fluid that can fit within a
fracture is less. After an initial volume calculation using the new width,
EPA found that the volume of the space within the fracture area may not
hold the volume of fluid pumped into the ground during a typical
fracturing event. Therefore, EPA assumes that a greater volume of
fracturing fluid must "leakoff' to intersecting smaller fractures than
what was assumed in the Draft Report, or that fluid may move beyond
the idealized, hypothetical "edge of fracture zone." This assumption is
supported by field observations in mined-through studies, which indicate
that fracturing fluids often take a stair-step transport path through the
natural fracture system.
• In the Draft Report, EPA approximated the edge of fracture zone concentrations
considering only dilution. Based on new information and stakeholder input on
the Draft Report, EPA does not provide estimates of concentrations beyond the
point-of-injection in the final report. Developing such concentration values
with the precision required to compare them to MCLs would require the
collection of significant amounts of site-specific data. This data in turn would
be used to perform a formal risk assessment, considering numerous fate and
transport scenarios. These activities are beyond the scope of this Phase I study.
The remainder of this section includes a discussion of the following components of EPA's
analysis:
• The concentrations of BTEX at the point-of-injection.
• The percentage of fracturing fluids recovered during the recovery process.
• The influence of the capture zone.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-12
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
• Factors that would increase or decrease the concentrations of BTEX remaining
in the subsurface.
The first step in EPA's analysis of the potential threat to USDWs from the injection of
fracturing fluids was calculating the point-of-injection concentrations of BTEX introduced
from diesel fuel in the gelling agent. In Step 2, EPA considered factors that affect the
degree to which hydraulic fracturing fluids are recovered. Steps 3, 4, and 5 provide
analyses of physical/chemical, hydrogeological, and biological processes that could affect
the fate and transport of hazardous chemicals introduced into coal seams. These steps are
summarized in Table 4-2.
4.3.1 Point-of-Injection Calculation
The formulations or "recipes" for fracturing fluids differ among service companies and
among sites; the amount of fracturing fluid used will also vary. Thus, a range of point-of-
injection concentrations likely exists. According to field paperwork obtained during EPA's
site visits (Consolidated Industrial Services, Inc., 2001; Halliburton, 2001) and information
provided by three service company scientists (BJ Services, 2001; Halliburton, 2001;
Schlumberger, Ltd., 2001), between 4 and 10 gallons of diesel-containing gelling agent are
added to each 1,000 gallons of water used in hydraulic fracturing, when diesel is used. In
addition, the fraction of BTEX in diesel may range by up to two orders of magnitude
(Potter and Simmons, 1998). The lower and upper ranges of the values presented in Potter
and Simmons (1998), as well as the three different values cited for gelling agent, were used
to estimate point-of-injection concentrations for each of three fracturing fluid recipes (i.e.,
the ratio of fracturing gel to water). The resulting 24 point-of-injection calculations are
provided in Table 4-2. These estimates provide the basis for a qualitative assessment
regarding whether a Phase II study is warranted.
The following example illustrates how EPA estimated the concentrations of BTEX at the
point-of-injection. Due to the variations in the recipe used by service companies, EPA's
analysis begins with three different possible scenarios, as follows:
• Low ratio: 4 gallons of gel per 1,000 gallons of water
• Medium ratio: 6 gallons of gel per 1,000 gallons of water
• High ratio: 10 gallons of gel per 1,000 gallons of water
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-13
image:
EPA 8 16-R-04-003 Chapter 4
_ Hydraulic Fracturing Fluids
The concentration of benzene in fracturing fluid at the point-of-injection ([benzene]inj) can
be calculated using the following equation:
[benzene]mj = (rgw) x (p^) x (f^) x (fbd) x (3,785 mLgei/galgei) x (1 galwater/3.785 Lwater) x (106 ,wg/g)
Where:
rgw = the ratio of diesel fuel-gel mixture to injection water (galgei/l,000 galwater)
(4 galgei/l,000galwater, 6galgei/l,000 galwater, and 10 galgei/l,000 galwater represent the low, medium,
and high ratios, respectively)
pdg = the density of the diesel fuel-gel mixture (ggei/mLgei) = 0.84 ggei/mLgei (Halliburton, 2002)
fdg = the fraction of diesel fuel in the gel (gdiesei/ggei) = 0-52 gdiesei/ggei (Halliburton, 2002)
fbd = the fraction of benzene in diesel fuel (gbenzene/gdiesei) = 0.000026 to 0.001 gbenzene/gdiesei (Potter and
Simmons, 1998)
3,785 mLgei/galgei = volume conversion factor
1 galwater/3.785 Lwater = volume conversion factor
106 /j,g/g = mass conversion factor
The concentration of benzene at the point-of-injection is calculated for the three gel/water
ratios and the minimum and maximum concentrations of benzene in diesel fuel.
Using rgw = 4 galgei/l,000galwater and fbd = 0.000026 gbenzene/gdiesei as an example,
[benzene]inj is calculated as follows:
[benzene]mj = (4 galgei/l,000galwater) x (0.84 ggei/mLgei) x (0.52 gfcei/ggei) x
(0.000026 gbenzene/gdiesei) x (3,785 mLgel/galgel) x (1 galwater/3.785 Lwater) x (1,000 mL/L) x (106 ^g/g) = 45 ,wg
Table 4-2 summarizes the estimated injection concentrations of each BTEX constituent for
the three assumed gel/water ratios and the minimum and maximum concentrations of
BTEX in diesel fuel. It also presents the MCL for each compound. Many of the estimated
concentrations of BTEX exceed the MCL at the point-of-injection.
Table 4-2 and the remainder of this section provide a qualitative assessment of the fate and
transport processes that could attenuate the concentrations of BTEX in groundwater.
Factors that would influence the availability of constituents of potential concern in
fracturing fluids and decrease their concentrations include:
• Fluid Recovery - much of the fluid is eventually pumped back to the surface.
• Adsorption and entrapment - some of these constituents will undergo adsorption
to the coal or become entrapped in the formation.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-14
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
• Biodegradation - some fracturing fluid constituents, such as benzene, may
undergo partial biodegradation.
4.3.2 Fracturing Fluid Recovery
Following the injection of fracturing fluids into the subsurface through coalbed methane
wells (i.e., production wells), considerable amounts of the fracturing fluids are removed.
During the recovery process, the injected fluids and ambient groundwater are pumped out
of the formation through the production well to reduce formation pressure, enabling
methane desorption and extraction. Palmer et al. (199la) found that 61 percent of
fracturing fluids were recovered based on samples collected from coalbed methane wells
over a 19-day period. Their study predicted total recovery to be between 68 and 82
percent.
Palmer et al. (1991a) also discussed the possibility that a "check-valve effect" could trap
some of the fracturing fluid on one side (i.e., upgradient, during production) of a collapsed
or narrowed fracture, preventing the fluid from flowing back to the production well. This
check-valve effect can occur in both natural and induced fractures when the fractures
narrow again after the injection of fracturing fluid ceases, formation pressure decreases,
and extraction of methane and groundwater begins.
Another factor preventing full recovery of injected fluids is the high injection pressure used
during hydraulic fracturing operations. Fracturing fluids are forced into the subsurface
under high pressure to enlarge and propagate existing fractures. The hydraulic gradients
that cause fluids to flow away from the well during injection are much greater than the
hydraulic gradients that occur during fluid recovery. As a result, some of the fracturing
fluids will travel beyond the capture zone of the production well. The capture zone of a
production well is the portion of the aquifer that contributes water to the well. The size of
this zone will be affected by regional groundwater gradients, and by the drawdown caused
by the well (USEPA, 1987). Fluids that flow beyond the capture zone of the production
well generally are not recovered during the flowback process.
Gel contained in fracturing fluids may be unrecovered because its properties differ from
that of water and highly soluble constituents of fracturing fluids. One mined-through study
reviewed by EPA described evidence of gel clumps within many fractures (Steidl, 1993).
One observed concentration of gel in a fracture was 15 times the injected concentration.
When the fluids exist as undissolved gel, they may remain attached to the sides of the
fractures or be trapped within smaller fractures or pores present in formations that surround
the coalbed. The mined-through studies suggest that such fluids are unlikely to flow with
groundwater during production, but they may present a source of gel constituents to
flowing groundwater subsequent to fluid recovery. Fate and transport processes discussed
later in this section can serve to reduce gel constituent concentrations that may result from
trapped fluids. Mechanisms that may affect the recovery of fracturing fluids are discussed
in section 3.3.2 of Chapter 3.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-15
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
4.3.3 The Influence of the Capture Zone
The recovery process typically lasts approximately 10-20 years. During that time,
groundwater within the production well's capture zone flows toward the production well.
Assuming complete mixing, the predicted recovery of injected BTEX is between 68 and 82
percent (Palmer et al., 1991a). Thus, between 20 and 30 percent of the BTEX injected is
expected to remain in the formation. It is reasonable to expect that most of the unrecovered
fluid lies outside the capture zone and that the residual concentrations of BTEX within the
capture zone are substantially less than the injection concentrations. Chemicals such as
BTEX that are not recovered from within the capture zone during groundwater production
may be diluted by groundwater that flows into the formation to replace production water.
Additional attenuation from sorption and biodegradation may occur. Subsequent to
production, dispersion and diffusion may serve to reduce residual BTEX concentrations.
The fracturing fluids that flow beyond the capture zone are affected by regional
groundwater flow and may be diluted by groundwater.
4.3.4 Fate and Transport Considerations
BTEX that has moved beyond the production well's capture zone is of the greatest concern.
The fate and transport mechanisms that may affect BTEX concentrations beyond the
capture zone are evaluated in this section. Factors that would likely decrease exposure
concentrations and/or availability of BTEX include attenuation through groundwater flow
dynamics, biological processes, and adsorption.
BTEX outside of the capture zone will likely be transported by groundwater flowing
according to regional hydraulic gradients. This flow and transport are not influenced by
production pumping. Nevertheless, mechanical dispersion will cause BTEX to spread
horizontally and vertically in the aquifer, thereby reducing the concentrations. The degree
of mechanical dispersion depends in part on the velocity of flow and increases with
increased travel distance. Dilution can have a significant effect on the BTEX
concentrations that could migrate to drinking water wells, especially if these wells are
hundreds to thousands of feet from a hydraulically induced fracture. The process of
molecular diffusion (i.e., the movement of BTEX from areas of higher to lower
concentration due to the concentration differences) will further reduce BTEX
concentrations. Collectively, mechanical dispersion and molecular diffusion are referred to
as hydrodynamic dispersion (Fetter, 1994).
The biodegradation of diesel fuel constituents, including BTEX, has been studied in other
geologic settings and laboratory studies and may lead to reductions in concentrations in
coalbeds given the appropriate site conditions. No information was found about the
occurrence of biodegradation or biodegradation rates of BTEX in coalbeds or surrounding
rock. In order for biodegradation to occur, organisms capable of using BTEX as a food
source must be present and conditions such as favorable pH, salinity, and sometimes the
availability of oxygen, nitrogen, and phosphorous must be met to ensure bacterial survival.
Generally, substantial benzene degradation occurs in aerobic environments. The levels of
oxygen in a particular formation vary widely depending primarily on the depth of coalbeds
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-16
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
from the surface. Data regarding biodegradation of benzene in an anaerobic environment
indicates a range from no degradation to relatively slow degradation (USEPA, 1999).
As groundwater flows through a formation, chemicals such as BTEX may be retarded by
adsorption. Although adsorption in coalbeds is likely, quantification of adsorption is
difficult in the absence of laboratory or site-specific studies (due to competition for
adsorption between BTEX and more lipophilic and less soluble constituents of diesel fuel
and coal, and fracture thickness). Other processes, such as desorption of BTEX from the
coal surface, and dissolution of BTEX from the gel phase may play a role in BTEX
transport. Entrapment of gel in pore spaces and fractures may also influence the degree to
which BTEX is available to groundwater. In some cases, the gel may be entrapped in such
a way that it is neither available to flow back towards the production well nor flow towards
a USDW in response to regional groundwater gradients.
According to the information listed on MSDSs provided to EPA, several of the constituents
of potential concern listed in Table 4-1 can have toxic effects when people are exposed to
sufficiently high concentrations through the susceptible route(s) of exposure (i.e.,
inhalation, ingestion, skin contact). However, only the BTEX compounds originating from
diesel fuel are regulated under SDWA. None of the other constituents in Table 4-1 appear
on the Agency's draft Contaminant Candidate List (CCL). The drinking water CCL is the
primary source of priority contaminants for evaluation by EPA's drinking water program.
Contaminants on the CCL are known or anticipated to occur in public water systems and
may require regulations under SDWA. Information on the GSA study is available at
http://www.epa.gov/fedrgstr/EPA-WATER/2004/April/Dav-02/w7416.htm.
Further, EPA does not believe that the other Table 4-1 constituents potentially contained in
fracturing fluids are introduced through coalbed methane fracturing in concentrations high
enough to pose a significant threat to USDWs. First, it is EPA's understanding, based on
conversations with field engineers and on witnessing three separate fracturing events, that
fracturing fluids used for coalbed methane fracturing do not contain most of the
constituents listed in Table 4-1. Second, if the Table 4-1 constituents were used, EPA
believes some of the same hydrodynamic phenomena listed in steps 2 and 4 (flowback,
dilution and dispersion), step 3 (adsorption and entrapment), and potentially step 5
(biodegradation) would minimize the possibility that chemicals included in the fracturing
fluids would adversely affect USDWs.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-17
image:
EL a EL
C? 3. G
en ^ en
o p. ed
5'l|
"'OQ
I |
3. CT)
OQ g
O O
en
1
o
o
Table 4-2: Estimated Concentijtions of Diesel Contanninls in f rKturing Fluids 4 the Point-of-lnjection and
Factors Affecting That CMtetnfrtf an and Movanertt in OrounAiatef
PoiM'iiMnjictl(ri{POI] CakylaHpr
9ep1
Rt;ip*
!
. i ' I
— - •
-
f;
^J
=~ "
Ml
£
'
Ch-smid!
|M&.
rt^l
S ." i" i"
rftWArt;
I I Hi
i i.
FrKtwi af Chtntc ri
In Dusri
Hn.
, , ,
i" ri"('(f-r
Mix.
j ft':;'
'I
•'
PsBt^tlnJKtlan
CwKentntion
to*i-
K(n.
-!••
— — —
i-yi
:'*U
I_ii
1:0
M)
:<|1
' 1 1
m^
.
— ___„. — r
~ *
8 f
0
,'
Faclcfs AffKtinglht C-3nc«rtraliD[n and Mov»m«nt af Fluids in Ornundwatet
Step 2
Iniusnei al Pumping
Wrf Wfiin tfat Captum
Z«r»
",,--•" • -i r ;.,' »-
!-* ' ^-' ii U ' ' * ^
' • « • I, ' i :'
s' i ~ ' ~! i ', I
a«p3
AiliiHfftlan and
Entrapnswrt
,' .,-„ V, ,< _, ,,<• ,
- -• : v - • * ,!••
• •> •!..-- » -> '
1 ,. t ' .«! , .J'll»
1, ., 1 '
,•'.<• " ' r • i •
» . 7 ' V .-
Step!
Mid Ckip>inian
••,"•'• J i ' . ,
•• ' ' : 4
StepS
Blalaglcil Di^idatiatt
1 • ;j_- , - ,,-,- .
,"!^ 1 ... , , ,i „ J
'• * I-. • ,•;• ; .<r
° s »f , %
oo
K^
ON
o
B*
rq "73
3 3
. •>
image:
EPA 816-R-04-003 Chapter 4
Hydraulic Fracturing Fluids
4.4 Summary
Fracture engineers select fracturing fluids based on site-specific characteristics including
formation geology, field production characteristics, and economics. Hydraulic fracturing
operations vary widely in the types of fracturing fluids used, the volumes of fluid
required, and the pump rates at which they are injected. Based on the information EPA
collected, water or nitrogen foam frequently constitutes the solute in fracturing fluids
used for coalbed methane stimulation. Other components of fracturing fluids used to
stimulate coalbed methane wells may contain only benign ingredients, but in some cases,
they contain constituents such as diesel fuel that can be hazardous in their undiluted
forms. Fracturing fluids are significantly diluted prior to injection.
Water with a simple sand proppant can be adequate to achieve a desired fracture at some
sites. In some cases, water must be thickened to achieve higher proppant transport
capabilities. Thickening can be achieved by using linear or cross-linked gelling agents.
Cross-linkers are costly additives compared to simple linear gels, but a fluid's fracturing
efficiency can be greatly improved using cross-linkers. Foam fracturing fluids can be
used to considerably reduce the amount of injected fluid required. The reduced water
volume requirement translates into a space and cost savings at the treatment site because
fewer water tanks are needed. Foam fracturing fluids also promote rapid flowback and
reduced volumes of flowback water requiring disposal.
The use of diesel fuel in fracturing fluids poses the greatest potential threat to USDWs
because the BTEX constituents in diesel fuel exceed the MCL at the point-of-injection.
Given the concerns with the use of diesel fuel, EPA recently entered into agreements with
three major service companies to eliminate diesel fuel from hydraulic fracturing fluids
injected directly into USDWs to stimulate coalbed methane production. Industry
representatives estimate that these three companies perform approximately 95 percent of
the hydraulic fracturing projects in the United States.
In situations when diesel fuel is used in fracturing fluids, a number of factors would
decrease the concentration and/or availability of BTEX. These factors include fluid
recovery during flowback, adsorption, dilution and dispersion, and potentially
biodegradation of constituents. For example, Palmer et al. (1991a) documented that only
about one-third of fracturing fluid that is injected is expected to remain in the formation.
EPA expects fate and transport considerations would minimize the possibility that
chemicals included in fracturing fluids would adversely affect USDWs.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 4-19
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figures 4-1 and 4-2.
Liquid nitrogen tanker
trucks transport gas to the
site for nitrogen foam
fracturing. Nitrogen will
travel through pipes to be
mixed with water and a
foaming agent at the
wellhead prior to
injection. The foam is
used to create and
propagate the fracture
deep within the targeted
coal seam.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-20
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figures 4-3 and 4-4.
Chemicals are stored on site in a support truck. Fracturing fluid additives such
as the foaming agent can be pumped directly from storage containers to mix
tanks.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-21
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figure 4-5.
The fracturing fluid (water with additives) is stored on site in large, upright storage tanks.
Each tank contains mix water imported from off-site, or formation water extracted directly
from the gas well.
Figure 4-6.
Gelled water is pre-mixed in a truck-mounted mixing tank. Photograph shows a batch of
linear, guar-based gel. This gel is used to transport the sand proppant into the fracture
propagated by the nitrogen foam treatment.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-22
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figure 4-7.
The fracturing fluids, additives, and proppant are pumped to the wellhead and mixed
just prior to injection. The flow rate of each injected component is monitored
carefully from an on-site control center.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-23
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figures 4-8 and 4-9.
Electronic monitoring systems provide constant feedback to the service company's operators.
Fluid flow rates and pressure buildup within the formation are monitored to ensure that fracture
growth is safe and controlled.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-24
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-25
image:
EPA816-R-04-003
Chapter 4
Hydraulic Fracturing Fluids
Figures 4-10 and 4-11.
Fluid that is extracted
from the well is
sprayed through a
diffuser and stored in a
lined trench until it is
disposed of off-site or
discharged.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
4-26
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
Chapter 5
Summary of Coalbed Methane Basin Descriptions
As part of the Phase I study EPA conducted an extensive literature review to collect information
regarding the major coal basins in the United States. Eleven major coal basins were identified in
the United States and are shown in Figure 5-1 at the end of the chapter (those basins shaded in
red have the highest coalbed methane production volumes). The goals of this review were to
assess the following for each of the 11 major coal basins:
• The physical relationship between the coalbeds and the USDWs.
• Whether hydraulic fracturing is or has been used to stimulate coalbed methane wells
in production basins.
• The types of fluids used to create the fractures.
• If possible, whether the potential for contaminants to enter a USDW exists.
This information is necessary to evaluate whether hydraulic fracturing is practiced within a basin
and the types of fluids used in the fracturing process. More importantly, this information
establishes whether the coal formations lie within a USDW, creating the potential for hydraulic
fracturing fluid injection to threaten USDWs. A USDW is not necessarily currently used for
drinking water and may contain groundwater unsuitable for drinking without treatment. In some
cases, very little information was uncovered by EPA regarding certain topics for some of the
basins.
Each of the 11 major basins is described in this chapter and in Table 5-1 (immediately following
section 5.12 of this chapter). In addition, a more comprehensive description of the geology,
hydrology, and coalbed methane production activity for each basin is provided in Attachments 1
through 11 of this report.
5.1 The San Juan Basin
The San Juan Basin covers an area of about 7,500 square miles straddling the Colorado-New
Mexico state line in the Four Corners region (Figure 5-1). It measures roughly 100 miles long
north to south and 90 miles wide. The Continental Divide trends north to south along the east
side of the basin.
The major coal-bearing unit in the San Juan Basin is known as the Fruitland Formation. Coalbed
methane production occurs primarily in coals of the Fruitland Formation, but some coalbed
methane is trapped in the underlying and adjacent Pictured Cliffs sandstone. Many wells are
completed in both zones. The coals of the Fruitland Formation are very thick compared to
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-1
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
coalbeds in eastern basins: the thickest coals range from 20 to over 40 feet. Total net thickness
of all coalbeds ranges from 20 to over 80 feet throughout the San Juan Basin, compared to 5 to
15 feet in eastern basins.
Coalbed methane wells in the San Juan Basin range from 550 to 4,000 feet in depth, and about
2,550 wells were operating in 2001 (Colorado Oil and Gas Conservation Commission and New
Mexico Oil Conservation Division, 2001). The San Juan Basin is the most productive coalbed
methane basin in North America. In 1996, coalbed methane production there averaged about
800 thousand cubic feet per day per well and totaled over 800 billion cubic feet (Bcf) for that
year (Stevens et al., 1996). This total rose to 925 Bcf in 2000 (GTI, 2002)).
The majority of coalbed methane development and hydraulic fracturing in the northern
portion of the San Juan Basin takes place within a USDW. The waters in parts of the Fruitland
Formation usually contain less than 10,000 mg/L TDS, which is the water quality criterion for a
USDW. In the northern half of the formation, most waters contain less than 3,000 mg/L, and
wells near the outcrop produce water that contains less than 500 mg/L TDS.
Fracturing fluids used in the San Juan Basin include hydrochloric acid; slick water (water mixed
with solvent); linear and crosslinked gels; and, since 1992, nitrogen- or carbon dioxide-based
foams (Harper et al., 1985; Jeu et al., 1988; Holditch et al., 1988; Palmer et al., 1993b; Choate et
al., 1993). Data are not readily available concerning fracture growth and height within the
Fruitland Formation.
5.2 The Black Warrior Basin
The Black Warrior Basin is the southernmost of the three basins that compose the Appalachian
Coal Region of the eastern United States. The basin covers about 23,000 square miles in
Alabama and Mississippi. It is approximately 230 miles long from west to east and
approximately 188 miles wide from north to south (Figure 5-1). Basin coalbed methane
production is limited to the bituminous coalfields of west-central Alabama, primarily in Jefferson
and Tuscaloosa Counties.
Coalbed methane production in the Black Warrior Basin is confined to the Pennsylvanian-aged
Pottsville Formation. The ancient coastline of prehistoric Alabama was characterized by 8 to 10
"coal-deposition cycles" of rising and falling sea levels. Each cycle features mudstone at the
base of the cycle (deeper water) and coalbeds at the top (emergence). Most coalbed methane
wells tap the Black Creek/Mary Lee/Pratt cycles and range from 350 to 2,500 feet deep
(Holditch, 1990).
Coalbed methane production in the Black Warrior Basin is among the highest in the United
States. In 1996, about 5,000 coalbed methane wells were permitted in Alabama. In 2000, this
number increased to over 5,800 wells (Alabama Oil and Gas Board, 2002). Coalbed methane
wells have production rates that range from less than 20 to more than 1 million cubic feet (Mcf)
per day per well (Alabama Oil and Gas Board, 2002). Between 1980 and 2000, coalbed methane
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-2
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
wells in Alabama produced roughly 1.2 trillion cubic feet (Tcf) of gas. According to GTI,
annual gas production was 112 Bcf in 2000 (GTI, 2002).
Some portions of the Pottsville Formation contain waters that meet the quality criterion of less
than 10,000 mg/L IDS for a USDW. According to the Alabama Oil and Gas Board, some
waters in the Pottsville Formation do not meet the definition of a USDW and have TDS levels
considerably higher than 10,000 mg/L.
Early literature indicates that most of the wells in production in the early 1990s have been
hydraulically fractured an average of two to six times to achieve acceptable production rates
(Holditch et al., 1988; Holditch, 1990; Palmer et al., 1993a and 1993b).
5.3 The Piceance Basin
The Piceance Coal Basin is entirely within the northwest corner of the Colorado (Figure 5-1).
The coalbed methane reservoirs are found in the Upper Cretaceous Mesaverde Group, which
covers about 7,225 square miles of the basin.
The Mesaverde Group ranges in thickness from about 2,000 feet on the west to about 6,500 feet
on the east side of the basin (Johnson, 1989). The depth to the methane-bearing Cameo-
Wheeler-Fairfield coal zone is about 6,000 feet. Two-thirds of the coalbed methane occurs in
coals deeper than 5,000 feet, and the Piceance Basin is one of the deepest coalbed methane areas
in the United States (Quarterly Review, August 1993).
The depth of the coals in the Piceance Basin inhibits permeability, making it difficult to extract
the coalbed methane. This, in turn, has slowed coalbed methane development in the basin.
However, it is estimated that 80 trillion to 136 Tcf of coalbed methane are contained in the
Cameo-Wheeler-Fairfield coal zone of the basin (Tyler et al., 1998). Total coalbed methane
production was 1.2 Bcf in 2000 (GTI, 2002).
The Piceance Basin contains both alluvial and bedrock aquifers. Unconsolidated alluvial
aquifers (narrow and thin deposits of sand and gravel formed primarily along stream courses) are
the most productive aquifers in the Piceance Basin. The bedrock aquifers are known as the
upper and lower Piceance Basin aquifer systems. The upper aquifer system is about 700 feet
thick, and the lower aquifer system is about 900 feet thick. Water at depth in the Piceance Basin
appears to be of poor quality, minimizing its chance of being designated a USDW. In general,
the potable water wells in the Piceance Basin extend no further than 200 feet in depth, based on
well records maintained by the Colorado Division of Water Resources. A composite water
quality sample taken from 4,637 to 5,430 feet deep in the Cameo coal zone exhibited a TDS
level of 15,500 mg/L (Graham, 2001).
Hydraulic fracturing is practiced in this basin. A variety of fluids are used for fracturing,
including water with sand proppant and gelled water and sand. In some cases, hydraulic
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-3
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
stimulations created multiple short (100-foot), fractures around the wells (Quarterly Review,
August 1993).
It is unlikely that any USDWs and coals targeted for methane production (generally currently
located at great depth, such as 4,000 feet below the ground surface and deeper) would coincide in
this basin. The thousands of feet of stratigraphic separation between the coal gas bearing Cameo
Zone and the lower aquifer system in the Green River Formation should prevent any of the
effects from the hydrofracturing of gas-bearing strata from reaching either the upper or the lower
bedrock aquifers.
Research suggests that exploration may target areas where groundwater circulation may enhance
gas accumulation in the coal and associated sandstones (Tyler et al., 1998). Under these
exploration and development conditions, a USDW located in shallower Cretaceous rocks near
the margins of the basin could be affected by hydraulic fracturing. The depth of methane-
bearing coals (about 6,000 feet) seems to indicate that, in the Piceance Basin, the chances of
contaminating any overlying, shallower USDWs (no deeper than about 1,000 feet) from injection
of hydraulic fracturing fluids and subsequent subsurface fluid transport are minimal. The
coalbed methane producing Cameo Zone and the deepest known aquifer, the lower bedrock
aquifer, have a stratigraphic separation of over 6,000 feet.
5.4 The Uinta Basin
The Uinta Coal Basin is mostly within eastern Utah; a very small portion of the basin is in
northwestern Colorado (Figure 5-1). The basin covers approximately 14,450 square miles
(Quarterly Review, August 1993). The Uinta Basin is stratigraphically continuous with the
Piceance Basin of Colorado, but is structurally separated from it by the Douglas Creek Arch, an
uplift near the Utah - Colorado state line.
Coal seams occur in the Cretaceous Mancos Shale and the Upper Cretaceous Mesaverde Group
(Quarterly Review, 1993). Two major formations targeted for coalbed methane exploration are
the Mancos Shale's Ferron Sandstone Member, which include the coals most targeted
(approximately 90 percent of the time) for exploration (Petzet, 1996) and the Mesaverde Group's
Blackhawk Formation, which contains about 14 coal zones (Petzet, 1996). The Ferron Coals are
interbedded with sandstone and form a wedge of clastic sediment 150 to 750 feet thick. Depths
to coal in the Ferron Sandstone range from 1,000 to over 7,000 feet below ground surface
(Garrison et al., 1997). The Blackhawk Formation consists of coal seams interbedded with
sandstone and a combination of shale and siltstone. Coals tapped in the Blackhawk Formation
are 4,200 to 4,400 feet below the surface (Gloyn and Sommer, 1993).
Full-scale exploration in the Uinta Basin began in the 1990s (Quarterly Review, 1993). The
database covering the Uinta Basin indicates that there are about 1,255 coalbed methane wells in
production in the basin (Osborne, 2002). The coalbed methane potential of the Uinta Basin,
revised by the Utah Geological Survey in the early 1990s, has been estimated to be between 8
trillion and more than 10 Tcf (Gloyn and Sommer, 1993).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-4
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
At some locations, the groundwater in the Perron Coals and Blackhawk Formations would not
qualify as USDWs. According to the Utah Department of Natural Resources (DNR), Division of
Oil, Gas and Mining, the water there varies greatly by location, each location having some TDS
levels below and some above 10,000 mg/L (Utah DNR, 2002). In general, the quality of
Blackhawk water is higher than Perron water. For example, the most recent UIC application
noted the combined quality of input water to be approximately 31,000 mg/L TDS for the
Drunkards Wash Field (Ferron) and 9,286 mg/L TDS for the Castlegate Field (Blackhawk).
Fracturing fluid use is documented in the literature pertaining to the Uinta Basin. One company
reported performing hydraulic fracturing stimulations using cross-linked borate gel with 250,000
pounds of proppant (Quarterly Review, 1993). Others report that they fractured wells with low-
residue gel fracturing fluids and foams (Quarterly Review, 1993). GTI places the annual coalbed
methane production in the Uinta Basin at 75.7 Bcf in 2000 (GTI, 2002).
The Blackhawk Formation is underlain by 300 feet of shale and sandstone, which separate it
from the Castlegate Sandstone aquifer. It is underlain by similar geologic strata, which separate
it from the Star Point Sandstone. Only in highly faulted areas is there a reasonable possibility
that hydraulic fracturing fluids could migrate down to the Star Point Sandstone.
5.5 The Powder River Basin
The Powder River Basin is in northeastern Wyoming and southern Montana (Figure 5-1). The
basin covers approximately 25,800 square miles (Larsen, 1989), approximately 75 percent of
which is in Wyoming. Fifty percent of the Powder River Basin is believed to have the potential
for coalbed methane production (Powder River Coalbed Methane Information Council, 2000).
Annual production volume was estimated at 147 Bcf in 2000 (GTI, 2002). In 2002, wells in the
Powder River Basin produced about 823 Mcf per day of coalbed methane (DOE, 2002).
Coalbeds in this region are interspersed at varying depths with sandstones, mudstone,
conglomerate, limestone, and shale. The majority of the potentially productive coal zones range
from about 450 feet to over 6,500 feet below ground surface (Montgomery, 1999). The
uppermost formation is the Wasatch Formation, extending from land surface to 1,000 feet deep.
Most coal seams in the Wasatch Formation are continuous and thin (6 feet or less). The Fort
Union Formation lies directly below the Wasatch Formation and can be as much as 6,200 feet
thick (Law et al., 1991). The coalbeds in this formation are typically most abundant in the upper
portion, called the Tongue River member. This member is typically 1,500 to 1,800 feet thick, of
which up to a composite total of 350 feet of coal can be found in various beds. The thickest of
the individual coalbeds is over 200 feet (Flores and Bader, 1999). Recent estimates of coalbed
methane reserves in the Powder River Basin range from 7 trillion to 40 Tcf (Montgomery, 1999;
PRCMIC, 2000).
The Fort Union Formation that supplies municipal water to the City of Gillette is the same
formation that contains the coals that are developed for coalbed methane. The coalbeds contain
and transmit more water than the sandstones. The sandstones and coalbeds have been used for
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-5
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
the production of both water and coalbed methane. The water produced from the coalbeds meets
the quality criterion for USDWs of less than 10,000 mg/L IDS.
EPA's understanding is that hydraulic fracturing currently is not widely used in this region due
to concerns about the potential for increased groundwater flow into the coalbed methane
production wells (due to fracturing of impermeable formations adjacent to the coal, and the
creation of a hydraulic connection to adjacent aquifers) and the collapse of open hole wells in
coal upon dewatering. According to the available literature, where hydraulic fracturing has been
used in this basin, it has not been an effective method for extracting methane. Hydraulic
fracturing has been done primarily with water, or gelled water and sand, although recorded use
of a solution of potassium chloride was identified in the literature.
5.6 The Central Appalachian Basin
The Central Appalachian Coal Basin is the middle of three basins that compose the Appalachian
Coal Region of the eastern United States. It includes parts of Kentucky, Tennessee, Virginia,
and West Virginia (Figure 5-1) and covers approximately 23,000 square miles. The greatest
potential for methane development is in a small, 3,000-square-mile area in southwest Virginia
and south central West Virginia (Kelafant, et al., 1988).
The coal basin consists of six Pennsylvanian age coal seams (Zebrowitz et al., 1991, and Zuber,
1998). These coal seams typically occur as multiple coalbeds or seams that are widely
distributed (Zuber, 1998). The coal seams, from oldest to youngest (West Virginia/Virginia
name), are the Pocahontas No. 3, Pocahontas No. 4, Fire Creek/Lower Horsepen, Beckley/War
Creek, Sewell/Lower Seaboard, and lager/Jawbone (Kelafant et al., 1988). The Pocahontas coal
seams include the Squire Jim and Nos. 1 to 7; Nos. 3 and 4 are the thickest and cover the most
area. Most of the coalbed methane (2.7 Tcf) occurs in the Pocahontas seams (Kelafant et al.,
1988). In southwest Virginia and south central West Virginia, target coal seams achieve their
greatest thickness and occur at depths of about 1,000 to 2,000 feet (Kelafant et al., 1988).
The Nora Field in southwestern Virginia is one of the better-known coalbed methane production
fields. According to the Virginia Division of Gas and Oil, over 700 coalbed methane wells were
drilled in the Nora Field in 2002 (Virginia Division of Gas and Oil, 2002). The Virginia
Division of Gas and Oil also indicated that, in 2002, more than 1,800 coalbed methane wells
were drilled in southwestern Virginia's Buchanan County (VA Division of Gas and Oil, 2002.)
GTI reported that the entire basin produced 52.9 Bcf of gas in 2000 (GTI, 2002).
Because most of the coal strata dip, a coalbed methane well's location in the basin may
determine if hydraulic fracturing during the well's development will affect the water quality of
surrounding USDW. For instance, on the northeastern side of the basin, the depth to the
Pocahontas No. 3 coalbed is less than 500 feet. This depth gradually increases to over 2,000 feet
farther westward across this portion of the basin, in the direction of the dip of the coal seam.
Therefore, a well tapping this seam in the eastern portion of the basin may be within a USDW,
but a well tapping the seam in the western portion of the basin may be below the base of a
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-6
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
USDW. In addition, the base of the freshwater is not flat, but rather undulating. These factors
indicate that the relationship between a coalbed and a USDW must be determined on a site-
specific basis.
Hydraulic fracturing is a common practice in this region. Foam and water are the fracturing
fluids of choice, and sand serves as the proppant. Additives can include hydrochloric acid, scale
inhibitors, and microbicides. Pocahontas Oil & Gas, a subsidiary of Consol Energy, Inc., invited
EPA personnel to a well where a hydraulic fracturing treatment was being performed by
Halliburton Energy Services, Inc. Halliburton staff said that typical fractures extend from 300 to
600 feet from the well bore in either direction, but that fractures have been known to extend from
as few as 150 feet to as many as 1,500 feet in length (Halliburton Inc., Virginia Site Visit, 2001).
According to the fracturing engineer on-site, fracture widths range from one-eighth of an inch to
almost one and one-half inches (Halliburton, Inc., Virginia Site Visit, 2001).
Since some coalbed methane exploration has moved to shallower seams, the Commonwealth of
Virginia has instituted a voluntary program concerning depths at which hydraulic fracturing may
be performed (Virginia Division of Oil and Gas, 2002). The program involves an operator's
determination of the elevation of the lowest topographic point and the elevation of the deepest
water well within a 1,500-foot radius of any proposed extraction well (Wilson, 2001). Hydraulic
fracturing should occur at least 500 feet beneath than the lower of these two points.
5.7 The Northern Appalachian Basin
The Northern Appalachian Coal Basin is the northernmost of the three basins that make up the
Appalachian Coal Region of the eastern United States. It includes parts of Pennsylvania, West
Virginia, Ohio, Kentucky, and Maryland (Figure 5-1). The basin lies completely within the
Appalachian Plateau geomorphic province and covers approximately 43,700 square miles
(Adams et al., 1984, as cited by Pennsylvania Department of Conservation and Natural
Resources, 2002). The Northern Appalachian basin trends northeast to southwest. The Rome
Trough, a major graben structure, forms the southeastern and southern structural boundaries.
The basin is bounded on the northeast, north, and west by outcropping Pennsylvanian-aged
sediments (Kelafant et al., 1988).
The six Pennsylvanian-aged coal zones composing the Northern Appalachian Coal Basin are the
Brookville-Clarion, Kittanning, Freeport, Pittsburgh, Sewickley, and Waynesburg. These coal
units are within the Pottsville, Allegheny, and the Monongahela Groups (Kelafant et al., 1988).
Coal seam depths range from surface outcrops to as much as 2,000 feet below ground surface,
with most coal occurring at depths shallower than 1,000 feet (Quarterly Review, 1993). These
depth differences arise due to the dip of the coalbeds. The total thickness of the Pennsylvanian-
aged coal system averages 25 feet; however, better developed seams within the coal zones can
increase in thickness by up to twice the average (Quarterly Review, 1993).
Coalbed methane has been produced in commercial quantities from the Pittsburgh coalbed of the
Northern Appalachian Coal Basin since 1932 (Lyons, 1997), after the discovery of the Big Run
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-7
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
Field in Wetzel County, West Virginia, in 1905 (Hunt and Steele, 1991). As of 1993, O'Brien
Methane Production, Inc. had at least 20 wells in Pennsylvania's southern Indiana County
(Quarterly Review, 1993). Coalbed methane production development in the Northern
Appalachian Basin has lagged, however, due to insufficient reservoir knowledge, inadequate
well-completion techniques, and coalbed methane ownership issues revolving around whether
the gas is owned by the mineral owner or the oil and gas owner (Zebrowitz et al., 1991).
Discharge of produced waters has also proven to be problematic (Lyons, 1997) for coalbed
methane field operators in the Northern Appalachian Coal Basin. Total coalbed methane
production stood at 1.41 Bcf in 2000 (GTI, 2002). As of October 2002, 185 coalbed methane
wells were producing coalbed methane in Pennsylvania (Pennsylvania Department of
Conservation and Natural Resources, 2002).
The Northern Appalachian Basin is situated in the Appalachian Plateau's physiographic
province. The primary aquifer in this area is a Pennsylvanian sandstone aquifer underlain by
limestone aquifers (USGS, 1984). Water quality data from eight historic Northern Appalachian
Coal Basin projects show that estimated TDS levels ranged from 2,000 to 5,000 mg/L at depths
of 500 to 1,025 feet below ground surface (Zebrowitz et al., 1991), well within EPA's water
quality criterion of 10,000 mg/L TDS for a USDW (40 CFR §144.3). Depths to the bottoms of
the USDWs vary greatly in the basin and are better determined on a site-specific basis.
Hydraulic fracturing fluids used in the Northern Appalachian Basin have included water and
sand, and nitrogen foam and sand (Hunt and Steele, 1991). The Christopher Coal
Company/Spindler Wells Project, which took place from 1952 to 1959, stimulated 1 well with 12
quarts of nitroglycerin (Hunt and Steele, 1991). In the Vesta Mines Project of Washington
County, Pennsylvania, the United States Bureau of Mines used gelled water and sand to
complete 5 wells in the Pittsburgh Seam (Hunt and Steele, 1991).
Because most of the coal strata dip, a well's location in a basin determines whether the well is
coincident with a USDW. For example, in the Pittsburgh Coal Group in Pennsylvania, the depth
to the top of the coal group varies from outcrop to about 1,200 feet in the very southwestern
corner of the state (Kelafant et al., 1988). The approximate depth to the bottom of the USDW is
450 feet. Therefore, production wells operating down to approximately 450 feet could
potentially be hydraulically connected to the USDW.
5.8 The Western Interior Coal Region
The Western Interior Coal Region comprises three coal basins, the Arkoma, the Cherokee, and
the Forest City Basins, and encompasses portions of six states: Arkansas, Oklahoma, Kansas,
Missouri, Nebraska, and Iowa (Figure 5-1). The Arkoma Basin covers about 13,500 square
miles in Arkansas and Oklahoma. The Cherokee Basin is part of the Cherokee Platform
Province, which covers approximately 26,500 square miles (Charpentier, 1995) in Oklahoma,
Kansas, and Missouri. The Forest City Basin covers about 47,000 square miles (Quarterly
Review, 1993) in Iowa, Kansas, Missouri, and Nebraska.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-8
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
In the Arkoma Basin, major middle-Pennsylvanian coalbeds occur within the Hartshorne,
McAlester, Savanna, and Boggy Formations (Quarterly Review, 1993). The Hartshorne coals of
the Hartshorne Formation are the most important for methane production in the Arkoma Basin.
Their depth ranges from 600 to 2,300 feet in two productive areas of southeastern Oklahoma
(Quarterly Review, 1993). In the Cherokee Basin, the primary coal seams targeted by operators
are the Riverton Coal of the Krebs Formation and the Weir-Pittsburg and Mulky coals of the
Cabaniss Formation (Quarterly Review, 1993). The Riverton and Weir-Pittsburg seams are
about 3 to 5 feet thick and range from 800 to 1,200 feet deep, while the Mulky Coal, which
ranges up to 2 feet thick, occurs at depths of 600 to 1,000 feet (Quarterly Review, 1993).
Individual coal seams in the Cherokee Group of the Forest City Basin range from a few inches to
about 4 feet thick, with seams up to 6 feet thick (Brady, 2002; Smith, 2002). Depths to the top of
the Cherokee Group coals range from approximately the surface to 230 feet below ground
surface in the shallower portion of the basin, in southeastern Iowa, to about 1,220 feet in the
deeper part of the basin, in northeastern Kansas (Bostic et al., 1993).
As of March 2000, there were 377 coalbed methane wells in the Arkoma Basin of eastern
Oklahoma, ranging in depth from 589 to 3,726 feet (Oklahoma Geological Survey, 2001). The
Arkoma Basin contains an estimated 1.58 to 3.55 Tcf of gas reserves, primarily in the Hartshorne
coals (Quarterly Review, 1993). In the Cherokee Basin, unknown amounts of coalbed methane
gas have been produced with conventional natural gas for over 50 years (Quarterly Review,
1993). Targeted coalbed methane production increased in the late 1980s, and at least 232
coalbed methane wells had been completed as of January 1993 (Quarterly Review, 1993). The
Cherokee Basin contains an estimated 1.38 Mcf of gas per square mile (Stoeckinger, 1989) in the
targeted Mulky, Weir-Pittsburg, and Riverton coal seams of the Cherokee Group (Quarterly
Review, 1993). In total, the basin contains approximately 36.6 Bcf of gas. However, the
Petroleum Technology Transfer Council (1999) indicates that there are nearly 10 Tcf of gas in
eastern Kansas alone (PTTC, 1999). The Forest City Basin was relatively unexplored in 1993,
with about 10 coalbed wells concentrated in Kansas' Atchison, Jefferson, Miami, Leavenworth,
and Franklin Counties (Quarterly Review, 1993). The Forest City Basin contains an estimated 1
Tcf of gas (Nelson, 1999). For the entire region, coalbed methane production was 6.5 Bcf in
2000 (GTI, 2002).
According to the National Water Summary (1984), there are no principal aquifers in the portions
of Oklahoma and Arkansas in the Arkoma Basin, only small alluvial aquifers bounding rivers.
Water quality test results from the targeted Hartshorne seam in Oklahoma have shown the water
to be highly saline (Quarterly Review, 1993). The base of fresh water in Arkansas is about 500
to 2,000 feet below ground surface (Cordova, 1963). However, Cordova (1963) does not define
"fresh water." While the majority of the Cherokee Basin does not contain a principal aquifer, the
Ozark and Douglas aquifers are contained within the basin (National Water Summary, 1984).
The confined Ozark Aquifer, composed of weathered and sandy dolomites, typically contains
water wells that extend from 500 to 1,800 feet in depth (National Water Summary, 1984). The
usually unconfined Douglas Aquifer is a sandstone channel of the Pennsylvanian Age (National
Water Summary, 1984). Wells are usually 5 to 400 feet deep in this aquifer. In Kansas, depth to
the base of the Ozark Aquifer is roughly 1,750 feet below ground surface (Ozark Aquifer Base
Map, 2001). In Oklahoma, the Cherokee Basin also contains the Garber-Wellington and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-9
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
Vamoosa-Ada aquifers (National Water Summary, 1984). Water well depths in these two
aquifers usually range from 100 to 900 feet (National Water Summary, 1984). The Forest City
Basin contains the Jordan Aquifer, the Dakota Aquifer, and glacial drift, alluvial, and Paleozoic-
aged rock aquifers. Wells in these aquifers commonly range in depth from 300 to 2,000 feet, 100
to 600 feet, 10 to 300 feet, 10 to 150 feet, and 30 to 2,200 feet, respectively (National Water
Summary, 1984). Throughout the Western Interior Coal Region, water quality sampling has
shown TDS levels to range from 500 to 40,000 mg/L (Missouri Division of Geological Survey
and Water Resources, 1967).
Hydraulic fracturing is common in the Western Interior Coal Basin. Fracturing fluids such as
linear gel, acid, and nitrogen foam were used extensively in the Western Interior coal region
before 1992, and slick water treatments became common in 1993. Hydraulic fracturing is still
practiced in the basin.
Based on depths to the Hartshorne Coal (0 to 4,500 feet in Arkansas) and the base of fresh water
(500 to 2,000 feet in Arkansas), it appears that coalbed methane extraction wells in the Arkoma
Basin could be coincident with potential USDWs in Arkansas (Andrews et al., 1998; Cordova,
1963). Based on maps provided by the Oklahoma Corporation Commission (2001) showing the
depths of the 10,000 mg/L TDS groundwater quality boundary in Oklahoma, coalbed methane
wells and USDWs would most likely not coincide in Oklahoma. This is based on depths to coals
typically greater than 1,000 feet (Andrews et al., 1998) and depths to the base of the USDW
typically shallower than 900 feet (OCC Depth to Base of Treatable Water Map Series, 2001).
In the Cherokee Basin, coalbed methane wells targeting the Cherokee Group coals in Kansas
coincide with USDWs. Depths to the top of coalbeds range from 800 to 1,200 feet (Quarterly
Review, 1993) while the depth to the base of fresh water is estimated at 1,750 feet (Mapped
information from the Kansas Data Access and Support Center (DASC), 200la). More
information concerning water quality is required prior to any determination of coalbed methane
well/USDW co-location in Missouri. However, current levels of coalbed methane activity are
minimal in that state. In addition, since only a very small portion of the Cherokee Basin falls
within Missouri, this portion of the basin needs to be delineated more precisely to see which
USDWs are in this small part of the basin. Last, in the Forest City Basin, there appears to be
little relationship between water supplies and coalbeds that may be used for coalbed methane
extraction. However, aquifer and well information from the National Water Summary (1984)
indicates that a co-location of the two could exist in Nebraska. More information is needed to
define the relationship between coalbeds and USDWs in the Forest City Basin.
5.9 The Raton Basin
The Raton Basin covers about 2,200 square miles in southeastern Colorado and northeastern
New Mexico (Figure 5-1). It is the southernmost of several major coal-bearing basins along the
eastern margin of the Rocky Mountains. The basin extends 80 miles north to south and as much
as 50 miles east to west (Stevens et al., 1992). It is an elongate, asymmetric syncline, 20,000 to
25,000 feet thick in the deepest part.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-10
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
There are two major coal formations in the Raton Basin, the Vermejo and the Raton. The
Vermejo coals range from 5 to 35 feet thick, while the Raton coal layers range from 10 to more
than 140 feet thick. Although the Raton Formation is much thicker and contains more coal than
the Vermejo Formation, individual coal seams in the Raton are less continuous and generally
thinner.
Methane resources for the basin have been estimated at approximately 10.2 Tcf in the Vermejo
and Raton Formations (Stevens et al., 1992). As of 1992, about 114 coalbed methane
exploration wells had been drilled in the basin (Quarterly Review, 1993). According to GTI, the
average coalbed methane production rate of wells in the Raton Basin was close to 300 thousand
cubic feet per day, and annual production in 2000 was 30.8 Bcf (GTI, 2002).
The coal seams of the Vermejo and Raton Formations developed for methane production also
contain water that meets the criterion for a USDW. The underlying Trinidad Sandstone and
other sandstone beds in the Vermejo and Raton Formations, as well as intrusive dikes and sills,
also contain water of sufficient quality to be used as drinking water.
Coalbed methane well stimulation using hydraulic fracturing techniques is common in the Raton
Basin. Records show that fracturing fluids used are typically gels and water with sand
proppants. Hemborg (1998) showed that in most cases water yield decreased dramatically as
methane production continued over time. However, some wells exhibited increased water
production as methane production continued or increased. Two causal factors were suggested
(Hemborg, 1998) for the rise in water production:
1. Well stimulation had increased the well's zone of capture to include adjacent water-
bearing sills or sandstones that were hydraulically connected to recharge areas, or;
2. Well stimulation had created a connection between the coal seams and the underlying
water-bearing Trinidad Sandstone.
5.10 The Sand Wash Basin
The Sand Wash Basin is in northwestern Colorado and southwestern Wyoming. It is part of the
Greater Green River Coal Region, which includes the Washakie Basin, the Great Divide (Red
Desert) Basin, and the Green River Basin (Figure 5-1). These sub-basins are separated by uplifts
caused by deformation of the basement rock. For example, the Sand Wash Basin is separated
from the adjacent Washakie Basin by the Cherokee Arch, an anticline ridge that runs east to west
along the Colorado - Wyoming border. The Greater Green River Coal Region, in total, covers
an area of approximately 21,000 square miles. The Sand Wash Basin covers approximately
5,600 square miles, primarily in Moffat and Routt Counties of Colorado.
The coal-bearing formations in the region include the lies, Williams Fork, the Fort Union, and
the Wasatch Formations. The total thickness of the coal seams in these formations can be up to
150 feet (Quarterly Review, 1993). Of all the formations, the Williams Fork is the most
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-11
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
significant coal-bearing unit because it has the thickest and most extensive coalbeds. Coal-
bearing strata are 5,000 feet deep along the basin's western portions and outcrop along its
southern and eastern margins. The coal seams are interbedded with sandstones and shale. The
thickest total coal deposits in the Williams Fork Formation, up to 129 feet, are centered on Craig,
CO. These deposits are composed of several separate seams up to 25 feet thick interspersed
between layers of sedimentary rock.
Coalbed methane resources in the Sand Wash Basin have been estimated at 101 Tcf
Approximately 90 percent of this gas is in the Williams Fork Formation. Approximately 24 Tcf
of coalbed methane are located less than 6,000 feet below ground surface (Kaiser et al., 1994a).
Some investigation and very limited commercial development of this resource have occurred,
mostly in the late 1980s and early 1990s. Records from the Colorado Oil and Gas Commission
indicate that approximately 31 Bcf of coalbed methane was produced in Moffat County during
1995 (Colorado Oil and Gas Conservation Commission, 2001). There appears to be no
commercial production at present (GTI, 2002). Development of coalbed methane resources in
the Sand Wash Basin has been slower than in many other areas due to limited economic
viability. The need for extensive dewatering in most wells has been a limiting factor,
compounded by relatively low coalbed methane recovery. In recent years, permits for new gas
wells have been issued, indicating that there may be some continued interest in this area
(Colorado GIS, 2001).
Kaiser and Scott (1994) summarized their extensive investigation of groundwater movement
within the Fort Union and Mesaverde Group. The Mesaverde Group is a highly transmissive
aquifer. The coal seams within the group may be the most permeable part of the aquifer. Lateral
flow within the Fort Union Formation is slower. Groundwater quality in the basin varies greatly.
Typically, chloride and TDS concentrations within the coal-bearing Mesaverde Group are low
and potentially within potable ranges in the eastern portion of the basin, implying the existence
of a USDW. TDS concentrations increase as the water migrates toward the central and western
margins of the basin. TDS concentrations significantly higher than the 10,000 mg/L USDW
water quality standard have been detected in the western portion of the basin.
The use of fracturing fluids, specifically water and sand proppant, has been reported for this
basin. No record of any other fluid types has been noted. Although variable, the water quality
within the fractured coals indicates the presence of USDWs within the coalbeds.
5.11 The Washington Coal Regions (Pacific and Central)
The Pacific Coal Region (Figure 5-1) is approximately 6,500 square miles and lies along the
western and eastern flanks of the Cascade Range, from Canada into northern Oregon within the
Puget downwarp structure. Bellingham, Seattle, Tacoma, and Olympia in Washington, and
Portland, Oregon, lie in or adjacent to the sub-basins. The Central Coal Region (Figure 5-1)
primarily lies within the Columbia Plateau, between the Cascade Range to the west and the
Rocky Mountains to the east, in Idaho. This region extends from the Okanogan highlands to the
north to the Blue Mountains to the south, and encompasses approximately 63,320 square miles.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-12
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
The coal-bearing deposits of the Pacific and the Central Coal Regions are Cretaceous to Eocene
Age and formed within fluvial and deltaic deposit!onal environments prior to the uplift of the
Cascade Mountain Range. The thick coalbeds of the Pacific and Central Basins are thought to
result from peat accumulations in poorly drained swamps of the lower deltas, while the thinner
coalbeds probably formed in the better drained upper deltas (Buckovic, 1979 as cited by Choate
et al., 1980). The complex stratigraphy and structural deformation of the coals of the Pacific
Coal Region are major obstacles to the exploration and development of gas fields. Although the
coals of the Central Coal Region may not be as greatly deformed and unpredictable as those in
the Pacific Coal Region, they are obscured by the Columbia River Basalt Group, in which
individual basalt flows up to 300 feet thick can cover thousands of square miles.
The occurrence of methane in groundwater is one factor leading to the identification of the gas
potential in Washington. Methane in groundwater occurs in the basalts, but only in confined
aquifers (porous or fractured zones near the top or bottom of a basalt layer) and is thought to
have migrated upward from underlying coalbeds. Choate et al. (1980) estimated coalbed
methane resources for four target sub-basins representing 1,800 square miles of the Pacific Coal
Region to be 0.3 trillion to 24 Tcf. Methane had been encountered in 67 oil and gas exploration
wells drilled in this region by 1984. Gas was found at depths of less than 500 feet in 25 wells,
less than 1,000 feet in 38 wells, and less than 2,000 feet in 50 wells. Pappajohn and Mitchell
(1991) estimated the coalbed methane potential of the Central Coal Region to be more than 18
Bcf per square mile. The operation of the Rattlesnake Hills gas field between 1913 and 1941 in
the western part of the Central Coal region indicates that greater potential for development may
exist. According to the available literature, there were no producing fields in either the Pacific
Coal Region or the Central Coal Region in Washington as of 2000 (GTI, 2001).
Water supply wells and irrigation wells in the Columbia River Basalts and water wells in
numerous different lithologies in the Pacific Coal Region have been recognized as containing
methane. Data demonstrating the co-location of a coal seam and a USDW were found for Pierce
County, where methane gas test well results report TDS levels far less than the 10,000 mg/L
USDW water quality threshold (Dion, 1984). These aquifers can be classified as USDWs. Data
demonstrating the co-location of a coal seam and a USDW was found for Pierce County, where
methane gas test well results report TDS levels of 1,330 to 1,660 mg/L, which is far less than the
USDW classification limit (Dion, 1984). Development of methane in the Central Coal Region
may have some impact on highly productive basalt aquifers already used as large sources of
irrigation water for agriculture (Dion, 1984).
Hydraulic fracturing of coalbed methane wells using sand and nitrogen foam treatments has been
documented (Quarterly Review August, 1993). However, optimal stimulation and completion
methods for use in the structurally difficult Pacific gas region are yet to be applied and proven.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-13
image:
EPA 816-R-04-003 Chapter 5
Summary of Coalbed Methane Descriptions
5.12 Summary
Hydraulic fracturing of coalbed methane production wells has been documented in each basin,
although it is not widely practiced in the Powder River, Sand Wash Basin, or the Washington
Coal Regions. Ten of the eleven major coal basins in the United States are located at least
partially within USDWs. The literature also indicates that hydraulic fracturing may have
increased or have the potential to increase the communication between coal seams and adjacent
aquifers in two of the basins: the Powder River and Raton Basins. This may be the explanation
for higher than expected withdrawal rates for production water in the Raton Basin following
some fracturing treatments. In the Powder River Basin, concerns over the creation of such a
hydraulic connection are cited as one reason why hydraulic fracturing of coalbed methane
reservoirs is not widely practiced in the region.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 5-14
image:
EL O EL
& 3. r-
^ o
^ l~t>
5. C/2
3 O
1
(Jl K<
O
O
•(*.
5-1. Evidence In Support nf f'oaM 'SD\V (.'*>-Lwation fn I '.& < 'oal
Basin
San Juan
Black Warrior
Piceance
Ifinta
Powder Riwr
Central
Appalachian
Are
coalbeds
found
wtthntiw
USDWf
YHH
v,,s
Miihlwfy
LWMlV
Explanation and^or evidence
A- IS'QS- an.-ri of th> F niillanri tysh-ni pntluot-*, wati-' iv>nfainirKi k*s Ihan
••N< IM ii HI IIM H i WAV Asiav^s 1rikf.ii hvn --i snlni'lpd ai^l vvnll '*ca sin;
•;ijolaining kss Ifun in.miii n g>l F I is (Ka.j-L-i i.-t a) ll^b-j
liMMin rr-i.|4 FDK, llm vitt\»r quslilv
at hv-jl i. V: iif ?? ',%•«•' Is) fir-; din ;«? walni
biiWe piit^is '^ "lit- Hi'fb-villi' f i.<iHi1lnii omldin v/dkf s Hint Mi''11* O'H'- '*|Udlirv '3lNnld nf k^ iyi "H'.nilij
rn;t'l rnjn fir a 1 ISPW Aosivrlinij ii *hn Alabama Oil and Gns rJnatrt, sunn; \valfs m ff*j»; PnHsv-BM
f miririh >n rjn rK >1 iifK-^l HIM ij*rliinln--:l nl H SUW dtul 'I-JVH 1 IJ,^ M./H!'.-. r:;r*<.nk>iHlily lu:ilw 1ian Ih.'Min ini|,t
mitilwl IVH^S. Hi IH?MS 141 tn 3d •|,-)lk<is pur riniiiV 1111 l'.4!Jiii H| ,4 , 1'7-*t ;HI--ir:: ^1 »l fQMl^i
i
Ftii- iTi«l;nd m*iih.in»- priHlinvq i^niHi;.C,iral fifv. iinrf Ifu- kr#K ••Kaiiff syslnni n Ilie '.IrHHii U^K
1 lUtfldNil Aft' 'IHCfc- f 'alt (J 5'X>" feu: dfidff I rH UU.-ii ^'ilK. llSS dt 'al^a! ilupOl, 'UJ^Ny 6 ',:f MI f«,'| iKi'ftfl 1f|M
•painrl snrf^n 'fi a lanv f>.f IHMI cif thr* k.v-m ( Fvfur i-l J , 1*WF!'i A ofli^ftit" v.'aifr qualify san'flK lahi-n
bun 4 fiJiT In L: '1^M tel -:!w>p wtlliBi Ihc rl .aiiiui'- UHH! AJIIH r Biw VVillidii^ 1 ink 1 <niialnHi frxliilul^: H 1 1)S
MII'SH arax Hanrn is • •* !i,*"fi * w qunli^ tli^l i' nus-t he 'S.-sfnts-id ul P HV^IIH^IIIHI flunks. i*i Jiff"!^:: "iiki lh»'
1 \K WSU.'I dldll'V Ml Pf» * *:f *iJll HlliJ fJIrliAlldA'k t'HliySs IJl^HUy W tfl (Cr>i1lkl. -J.+Jl ||HV»II!1 I DS f*!'«'«S '.MilifA' HT»"]
ahi.vi' in H!»:'-iTiq,'l (:!f;ih Di-partnnnt nf Natural FipMKiri>vs ?!'!•?) 1
A Mxiri pfsparf'l hv lli«.! Unif*'iJ Sail's (ji.'iiliiq-ral kfvi/v il1 i^OS- shiwi.-d twf srinip»'N ••>! wsl'T ai pn)fliic>"i1
frim ^.? COM wells P Il'ii.' F'l^v^'f flivi.T Dssiii all rtdiJ IPS i»'v»>b of li-ss l^n in.'""11 ci'J* (fcou <*1 al , /MMII)
Fht- w;jlfiir pr.idiii»-d hy Oif'.-l wells -n h« 1 '<wrter R-v^rf-iJal f n-IH ':iM*ira:r<v nif^ls dnnkin;i ft'^fftr sf-Tfrt;irds
h i^d. prQfa\(\\\fn\'tx[tr> sin^i as tti^^n t"3'.'i* hn^n r^npffiftfl ;ss n sftpnrnfp- ric s4pplHn«nta s-ri /;:*• 4nr
I nxrnpril rtirkno v«ahjf *^ sirn« aroas 'iPi'Fufn r-1 ?il , Z^IIIH;
IH=;}||is i/i;!i.^l i|HHipK Hi" ixiii.Jilh-nl'A'itli li
• K'-'laf^l "fa. 1 fi[if;, Wiism. 2' n 1 1 . F i is:cj
HI HijjH-its Ihnf ptix'a^i- rtHfls r. Virinnia ;«H s-'
(ti.li >vrtkr in al (^H^F Iwu uf t>* s=ln"-!> wirtiin II IH ir-.'i!iHll bavn
1 W1. 1 fi ipkj -ts 1 %6, 1 ISGS 1 M f J j ,Ajn M :Ci itel i Jifi mi i it i HI
^i'nm«i1wlhinoiH'sk;*ns A-ViHiin ?>t'il VlMMf,, ?i!ii1|
w
hd
>
oo
ON
O
O
O
O
fa
(Ji
.
O
image:
El O £i
& 2- fi
c/a c/a
§^o
O f"^ • (™^
S'gg
en f>
°'oq
3!
o B
3. GO
B 2
OQ C
O rt
Basin
Northern
Appalachian
Western
Interior:
^ftornu
Cherokee
Fofttf c«iy
Raton
Pacific Coal
Arc
coalbeds
found
within Ine
USOW?
Yes
Yes (in
Arkansas)
Unlikely (lit
Unlikely
Yes
Y«s
Explanation anchor evidence
tn i »f 4-,-id> -'xial nrsHip with-n NIH husin is. niN
qunlily ddlrf lirrn **jgtH Iiiv>ttj«i^f4<ir1l'-»i
ir.y.1 (/•j!if(WLi'»jl rfl 1ll'-H)
;nl v«ilh Hie diifilfis. uf I iSfJvVs. iKXIiiffrt i*t ^1, l^iifj.
ilian 'JaJ tinsin (if njvubi sh'AV I Db levuls liek,w/
!>,
iTs within Arkansas ri
. Llui
til CBf,f wall's ami 1
I Wllhi
ilhu iu.ii
WliJlrt IIU
is In rrnsh 'A'fifnf lAnririr.vs at ,il ,
hkHy i
i\ Hi 1
nn'1 i
in llr^t Sldly 11 is IB lids^jd
HiOis h> I'IH l;a«iH nl ll'i-
i*; nupns ix •••*inlDttfK in r^nits.
,«(:(;«|r,f|i' ?IK"1. DA^I ., ;J'''"H,i}
.'SJi-nf 1h.- rlr^lll
ir<n"» Ihu ;|'*»!(M ftiriiin'd> v/ithn
1 H! , V»'.IH}
quality THSIJIS rurn CEtM wf;lls in HIH '-?,il;in R,*nn M>«iii3(islr,iV
i>*l<"S-s Ihnn 1iJ "MSI rrif|.'l
hiiil Tfiy nl |r«-xx Hvm
Iwogas ftHripanifs.fnx1iinf.d wnfer fr«rri OM!S lha? shmverl fRM lnvHs hf-hiw 1i!,',"»i mrj.'1 Alf.mir]
in MurT^I CiHinlv rxirwrtHJ Ol (jirpufritun ilnlluil 1i« C;fiM wiHls Fhn v,r-tJs yiflrhtrt lrs.i,]t* vuliirri'-.'- ^1 ti:sh
Wa't'i (luiiic^U 'Mtii lift vv^ls i.'iHildHii.'d 1 y[Hi nitj.l ^l| IL1H anil WHS r^cJ'-Hi'.ieiJ K
iluiniHi&lrtif*^ Ifn,1 i:n lm:alii «l i^ a i5si.il Si!,
linn y?is ti?sl wvlls 'nli'^jhfs 1 US k-vuK hi«
;HK! ^ 1 ibflW if- f V1*^-
iien 1 ;ii" dii<j 1 fft-ii my.
V'Vti'nr
rrnnuk
i-'ls in
oo
ON
O
O
in
o
fa
c?
rt>
O
o
image:
EPA816-R-04-003
Chapter 5
Summary of Coalbed Methane Descriptions
Figure 5-1. Locus Map of Major United States Coal Basins
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
5-17
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
Chapter 6
Water Quality Incidents
While Chapters 3 through 5 describe the theoretical and technical background for the
potential contamination of USDWs from hydraulic fracturing fluid injection into coalbed
methane wells, this chapter summarizes citizens' accounts of water quality and quantity
incidents. These reports reflect the opinions of citizens living near coalbed methane
operations who expressed concerns about contaminated drinking water wells and wells
experiencing water quantity impacts such as reduced production. EPA has, through
letters and telephone calls, contacted and been contacted by citizens who believed their
water wells were affected by coalbed methane production in the San Juan, Black Warrior,
Central Appalachian, and Powder River Basins. Stakeholders commenting on the study
methodology (65 FR 45774 (USEPA, 2000)) asked that EPA consider personal
experiences regarding coalbed methane impacts on drinking water wells in addition to
data from formal studies.
As a result of the stakeholder comments, EPA published a request in the Federal Register
(66 FR 39396 (USEPA, 2001)) for information from the public, as well as governmental
and regulatory agencies, regarding incidents of groundwater contamination believed to be
associated with hydraulic fracturing of coalbed methane wells. In addition, the Agency
notified over 500 local and county agencies in areas of potential coalbed methane
production making them aware of the Federal Register notice, but EPA received no
information regarding citizen complaints from these officials. Therefore, EPA believes it
knows the major geographic areas where citizens have reported problems that they
attribute to coalbed methane development. These areas are concentrated in the most
active basins: the San Juan, Black Warrior, Central Appalachian, and Powder River
Basins. The Agency has included relevant information from the water quantity and
quality incident reports that it has received.
Many of the reported incidents (such as impacts to water supply quantities or the effects
of discharged groundwater extracted during the coalbed methane production process) are
outside of the scope of SDWA and beyond the scope of this Phase I of the study.
However, all incidences reported in response to the Federal Register request are included
so that this study can be as inclusive as possible with respect to reported incidences and
not inadvertently exclude a relevant reported incident. This study is specifically focused
on assessing the potential for contamination of USDWs from the injection of hydraulic
fracturing fluids into coalbed methane wells, and determining based on these findings,
whether further study is warranted.
It is important to note that activities or conditions other than hydraulic fracturing fluid
injection may account for some of the reported incidences of the contamination of
drinking water wells. These potential causes include surface discharge of fracturing and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-1
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
production fluids, poorly sealed or poorly installed production wells, and improperly
abandoned production wells.
For this phase of the study (Phase I), EPA consulted with state agencies to determine if
they had received reports of groundwater problems, to learn of any follow-up steps
typically taken by the state, and to determine the states' overall findings regarding any
impacts that hydraulic fracturing of coalbed methane wells may have had on
groundwater.
This chapter summarizes correspondence EPA has had with individual citizens and states,
organized by basin, as follows:
• San Juan Basin (Colorado and New Mexico).
• Powder River Basin (Wyoming and Montana).
• Black Warrior Basin (Alabama).
• Central Appalachian Basin (Virginia and West Virginia).
6.1 The San Juan Basin (Colorado and New Mexico)
For over a decade, citizens in the San Juan Basin region have reported that coalbed
methane development has resulted in increased concentrations of methane and hydrogen
sulfide in their water wells. Other complaints about coalbed methane development
include the loss of water, the appearance of anaerobic bacteria in water wells, and the
transient appearance of particulates in well water. In conversations with EPA, most
citizens and local government officials did not specify hydraulic fracturing as the cause
of well water problems. Summaries of reported incidents and state follow-up are
discussed in sections 6.1.1 and 6.1.2, respectively.
EPA reviewed the BLM study summarizing the history of methane seeps, citizen
complaints, and follow-up investigations related to conventional gas and coalbed
methane development in the San Juan Basin to determine if they contained information
pertaining to coalbed methane hydraulic fracturing and its impact, if any, on the quality
of water in drinking water aquifers in the basin. A summary of pertinent findings is
provided in section 6.1.3.
6.1.1 Summary of Reported Incidents
• EPA spoke with a former county employee who, earlier in his career, had
worked for Exxon performing hydraulic fracturing jobs (Holland, 1999). As a
county employee, he took measurements for methane and hydrogen sulfide
inside homes in response to citizen complaints. He indicated that there were
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-2
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
no significant problems until the shallowest formation of coal (the Fruitland
Formation) began being developed. He believed that the main route of
contamination is from older, poorly cemented wells, and he estimated that
hundreds of wells have been affected. He said the biggest problems
associated with the apparent effects of coalbed methane development are the
explosive levels of methane and the toxic levels of hydrogen sulfide in homes.
In his opinion, this is due to the removal of water, rather than to hydraulic
fracturing.
• The San Juan Citizens Alliance estimated that hundreds of private water wells
have been affected by coalbed methane production in the area of Durango,
CO. These complaints include the following:
A lawyer representing several Durango citizens whose wells were
contaminated, allegedly due to coalbed methane development, said
there have always been methane seeps in the river, which have
manifested as bubbling water (McCord, 1999). In the early 1980s,
however, people began to see increased concentrations of natural gas
in their water wells shortly after companies began producing methane
from the Fruitland Formation.
One individual reported that two of his wells were degraded because
of increased methane levels. According to this individual, his
neighbor's pump house door was blown off, presumably as a result of
explosive levels of methane. Amoco bought three ranches after county
officials tested indoor air and found extremely high levels of methane.
This individual also told EPA staff that an area of the Southern Ute
tribal land has increased levels of hydrogen sulfide at the surface. He
reported he had also heard of black water due to pulverized coal.
Another private well owner claimed that her neighbors' wells are
contaminated by gas infiltration from dewatering. First methane
contaminates the well, then hydrogen sulfide, then anaerobic bacteria.
She claimed that data exists showing that methane concentrations in
water have increased by 1,000 parts per million (ppm).
• EPA Region 8 received letters from citizens concerned that coalbed methane
development had contaminated their water with methane and hydrogen
sulfide.
• During a visit to Durango, CO, EPA met with several citizens who claimed to
have experienced problems with their water due to coalbed methane
development. Most of the citizens experienced water loss, but two well
owners from New Mexico claimed that the quality of their water was affected
by hydraulic fracturing. According to their accounts, the water turned cloudy
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-3
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
with grayish sediment a day or two after nearby fracturing events. Eventually,
the well water returned to its normal appearance.
EPA also toured the area during that visit. EPA staff viewed areas where
patches of grass and trees were turning brown and dying. In some places,
large, old-growth trees located within the patch indicated that the area
previously had prolonged normal soil conditions. Many citizens and some
local officials believed that the areas suffered from increased methane and
decreased air in the soil gas in the shallow root zone.
• A La Plata County official reported that citizens have called to complain that
well water flow decreases when coalbed methane wells are hydraulically
fractured (Keller, 1999). He reported that "a lot" of people are hauling water
due to water loss. The county official said that, in two separate reports, well
owners noticed problems with their well water approximately 2 weeks after
nearby fracturing events. They reportedly believe hydraulic fracturing is
responsible because the timing of the water loss coincides with the fracturing.
Citizens know when gas producers fracture wells because they can see and
hear the operation, which involves several trucks, tanks, manifolds, and
mobile trailers. The county official noted that the formation being developed,
the Fruitland Formation, is located approximately 2,400 feet below ground
surface (bgs), and water wells are generally drilled from 100 feet to 200 feet
bgs. He qualified his statements by indicating that wells do go dry for a
variety of reasons.
• EPA contacted the Colorado Department of Health (CDH), which has primacy
for the UIC Program under SDWA. An official with whom EPA spoke said
CDH believes that water removal associated with coalbed methane
development has caused problems in private water wells (Bodnar, 1999).
• EPA received one complaint from a citizen living in the Raton Basin in
Trinidad, CO. She reported that water wells in her area have begun to decline
in production and quality, often producing more and more gas. She believes
the decline of water wells in her area is due to dewatering associated with
coalbed methane production.
6.1.2 State Agency Follow-Up in the San Juan Basin
Colorado Oil and Gas Conservation
The Colorado Oil and Gas Conservation Commission (COGCC) is responsible for
environmental issues related to oil and gas production in the state. The COGCC responds
to every complaint called in to its office (Baldwin, 2000).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-4
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
The COGCC staff believes that increased methane concentrations in water wells and
buildings in some areas are partially due to old, improperly abandoned gas wells and
older, deeper conventional gas wells in which the Fruitland Formation was not
completely isolated. The state bases its opinions on monitoring and studies conducted in
the San Juan Basin in response to complaints (see section 6.1.3). According to COGCC
officials, the state's mitigation program focused on sealing old, improperly abandoned
gas wells and appears to have reduced methane concentrations in approximately 27
percent of the water wells sampled. They believe that methane concentrations will
decrease over time in other water wells where the source of the methane was gas wells.
There are other areas of the San Juan Basin where the methane in water wells is produced
by methanogenic bacteria in the aquifer. Methane concentrations in water wells in these
areas probably will not decrease.
Officials cite studies that use stable carbon and hydrogen isotopes of methane and gas
composition to differentiate between thermogenic methane from the Fruitland,
Mesaverde, and Dakota Formations, and biogenic methane that is produced in shallower
formations by naturally occurring methanogenic bacteria. By 1998, approximately two-
thirds of the water wells for which gas isotopic analyses had been performed appeared to
contain biogenic gas, while one-third appeared to contain thermogenic gas.
The state also noted that, in the interior basin, 1,100 feet of shale separates the Fruitland
Formation and the shallow formations in which private wells are completed.
New Mexico Oil Conservation Division
EPA spoke with a District Geologist employed by the New Mexico Oil Conservation
Division (NMOCD). He said that several years ago the office received many complaints
that methane had contaminated water wells (Chavez, 2001). The state held water fairs at
which anyone could have his or her water tested. In addition, the state initiated a
program for cemented wells (some active, some abandoned) that prohibited open holes
100 feet above the casing string. The District Geologist indicated that the program
seemed to solve the problem and that NMOCD has not received many subsequent
complaints.
6.1.3 Major Studies That Have Been Conducted in the San Juan Basin
As noted previously, EPA reviewed a BLM study on the San Juan Basin to determine if it
contained information pertaining to coalbed methane hydraulic fracturing and its impact,
if any, on the quality of water in drinking water aquifers in the basin. EPA's review of
this report focused on the two potential mechanisms by which hydraulic fracturing may
affect the quality of USDWs: 1) direct injection of hydraulic fracturing fluids into a
USDW or injection of fracturing fluids into a coal seam already in hydraulic
communication with a USDW (e.g., through a natural fracture system), and 2) creation of
a hydraulic connection between the coalbed formation and an adjacent USDW. The
reports did not specifically address hydraulic fracturing, and only very little information
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-5
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
indirectly addresses the question specific to this study: Does the injection of hydraulic
fracturing fluids into coalbed methane wells contaminate USDWsl
The studies provided information on evidence that a hydraulic connection exists between
coalbeds in the Fruitland Formation and overlying shallow aquifers and on possible
conduits that may be the basis of the hydraulic connection. For example, the presence in
a shallow aquifer of methane documented to be from the underlying Fruitland Formation
is indirect evidence of a hydraulic connection, through some type of conduit, between the
Fruitland Formation and shallower formations.
Evidence that a hydraulic connection exists between coalbeds and the shallow aquifer
The U.S. Department of the Interior's BLM (1999) provides a history of gas seeps and
methane contamination of drinking water wells in the San Juan Basin. This section will
review the evidence that indicates the existence of a hydraulic connection between the
deep coalbeds and shallow USDWs.
Even prior to oil and gas drilling operations, shallow water wells in the San Juan Basin
produced methane gas. Some wells in the Cedar Hill, NM, area of the basin were
reported to have a strong sulfur odor. Some shallow water wells around the basin rim
penetrated the Fruitland and Menefee coalbeds and produced methane (BLM, 1999).
Thus, coalbed methane was the source of at least some of the observed methane
contamination. Water from the Fruitland coalbed discharges in the western part of the
basin and migrates upward across the Kirtland shale into the Animas and San Juan Rivers
(Stone et al., 1983). In areas such as La Plata County, CO, along the northern and
western rims of the basin, the methane presumably moves through natural fractures.
In the interior of the basin, gas seeps were observed in pastures in the Animas River
Valley south of Durango near Bondad, CO, and Cedar Hill, NM, in the early to mid-
1980s. Bubbles were also observed in the Animas River and in the tap water of rural
properties in these areas. Methane was responsible for explosions in several pump
houses. A landowner in New Mexico reported that gas was bubbling out of his alfalfa
field and in the Animas River in 1985. Gas seeps were likely the cause of patches of
dead grass growing in soils overlying the Mesaverde sandstone (BLM, 1999). Thus,
conduits between methane-containing units and the surface were present both at the rim
and in the interior of the basin.
After coalbed methane production began in the basin in the late 1980s, a local citizens'
group voiced concerns that natural gas contamination of drinking water wells had
increased in La Plata County. One study reported that 34 percent of the 205 domestic
water wells tested in the county showed measurable concentrations of methane (BLM,
1999). This appears to indicate that there is a conduit for fluid to flow to the shallower
USDW and its drinking water wells.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-6
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
Shortly after the start of coalbed methane production in the basin, 11 coalbed methane
wells were drilled within 2 miles of the Pine River Ranches Subdivision at the rim of the
San Juan Basin. Nine to 35 feet of alluvium separate the surface from the Fruitland
Formation coals in this area. A number of problems were reported following the onset of
coalbed methane production. A man who complained that his well was contaminated
with methane saw streams of gas bubbles in the nearby Los Pinos River. His report of
methane contamination was confirmed by the San Juan Regional Authority (SJRA),
which investigated reported contamination of this well and nearby wells. The other wells
were also contaminated with methane. Two of the 4 residences near the 11 coalbed
methane wells contained explosive levels of methane in crawl spaces (BLM, 1999). The
methane sampled in the shallow wells and the bubbling river and the high concentrations
of methane detected in residences suggest that coalbed methane was following some
conduit from the Fruitland Formation to the surface or to shallow USDWs.
Evidence that methane in shallow drinking water wells originates in the Fruitland
Formation (location of the coalbeds targeted by hydraulic fracturing)
Several lines of evidence show that methane detected in alluvial wells is not a result of
sewage-derived methane contamination (BLM, 1999). Rather, the methane in the
domestic wells studied originates either in conventional gas reservoirs such as the Dakota
sandstone and the Lewis Shale or in the coals of the Fruitland Formation.
The composition of the gas in samples from shallow, private drinking water wells was
analyzed to confirm the well owners' observations. The data obtained showed that the
methane in approximately half of the samples appeared to have originated in the
Fruitland Formation coalbeds and not from other possible sources such as septic tanks
(BLM, 1999).
Similar sampling and analyses conducted in an additional study cited by BLM (1999)
concluded that gas in a domestic well in alluvium overlying the Fruitland Formation had
the same gas composition and carbon-13 isotope ratio as gas from a nearby gas well also
in the Fruitland Formation. This study found that C13 isotopic signatures of individual
near-surface gas samples correlated with production gas from discrete formations beneath
the study area (BLM, 1999). In addition, an area resident's well contained 680 ppm
TDS, primarily sodium bicarbonate. Fruitland-produced water has the same composition,
although other domestic wells in the area do not. (TDS values tend to be in the 100 to
200 ppm in these other domestic wells.) Both the gas and the water analyses indicate that
the shallow aquifer in the area (from which the methane-contaminated domestic wells
draw drinking water) is in hydraulic communication with the deeper Fruitland Formation
coalbeds.
Possible conduits for fluid movement from the coalbeds to the aquifer
Several studies have assessed possible natural or manmade conduits to account for the
confirmed occurrence of methane in wells tapping the shallow aquifer that overlies the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-7
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
deeper coalbeds in the Fruitland Formation. Possible pathways enabling methane to
move from a deep source to a shallow aquifer include natural fractures, hydraulically
induced fractures, disposed of produced water from coalbed methane wells, and poorly
constructed, sealed, or cemented conventional gas wells, coalbed methane wells, shallow
drinking water wells, and cathodic protection wells installed to protect oil and gas
pipelines from corrosion (BLM, 1999).
The history of documented gas seeps and methane occurrence in water wells indicates
that natural fractures probably serve as conduits in parts of the basin where coal
formations are near or at the surface and in the interior of the basin, where the coal
formations are deeper. These conduits may enable hydraulic fracturing fluids to travel
from targeted coalbeds to shallow aquifers. However, there is no unequivocal evidence
that this fluid movement is occurring and, even given the presence of these possible
conduits, other hydrogeologic conditions (such as certain pressure gradients, etc.) would
be required for fluid movement from targeted coalbeds to shallow aquifers.
A study comparing soil-gas-methane concentrations adjacent to 352 gas-well casings and
192 groundwater wells found that the gas-well annuli (i.e., the spaces between the steel
well casings and the walls of the drilled bore holes) were frequently the reason methane
moved from the coalbeds to the near-surface environment (BLM, 1999). Thus, gas-well
annuli are clearly one type of conduit for movement of methane from deeper sources up
to overlying shallow aquifers.
The possibility of leaking gas wells acting as conduits through which methane flows from
the Fruitland Formation to shallow aquifers was investigated by a joint Colorado Oil and
Gas Conservation Commission/BLM study (BLM 1999). One hundred twenty water
wells were tested for methane before and after nearby gas wells were "remediated"
(better sealed). The study concluded that the relationship between gas well remediation
and lower methane concentrations in drinking water was "complex" and may have been
affected by the lingering presence of methane in drinking water after gas well
remediation. More than half the water wells showed no significant changes in methane
occurrence, a quarter showed lower methane levels, and one-tenth showed increased
methane.
In summary, there appears to be evidence that methane seeps and methane in shallow
geologic strata and water wells may occur because the methane moves through a variety
of conduits. These conduits include natural fractures; poorly constructed, sealed, or
cemented manmade wells used for various purposes. No reports provide direct
information regarding hydraulic fracturing. Methane, fracturing fluid, and water with a
naturally high TDS content could possibly move through any of these conduits. In some
cases, improperly sealed gas wells have been remediated, resulting in decreased
concentrations of methane in drinking water wells.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-8
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
6.2 The Powder River Basin (Wyoming and Montana)
EPA spoke with several individuals familiar with coalbed methane activity in the Powder
River Basin area who believe coalbed methane production is causing water quantity
issues. These individuals have reported that dewatering during coalbed methane
production resulted in loss of water from wells and in flooding problems on the surface.
Many of the drinking water wells in the Powder River Basin are screened and completed
in the same formation being dewatered for methane production. According to a
consulting hydrogeologist, as much as 1 million gallons of water are pumped from each
coalbed methane production well during its lifetime. Consequently, the aquifer has
dropped 200 feet in some areas (Merchat, 1999). EPA has also learned that, as of 1999,
oil and gas companies have drilled 2,000 wells in the Powder River Basin, and they
reportedly plan to drill 15,000 in total (Merchat, 1999). However, deeper aquifers are
available, and the oil and gas companies have drilled new water wells in those aquifers
for private individuals.
Reports of incidents in the Powder River Basin are summarized below. However,
hydraulic fracturing is performed infrequently in the Powder River Basin, and no one
living in that area has reported problems relating to the process. Many of the complaints
relate to water quantity issues, which are beyond the scope of this study.
EPA contacted the state and local offices of the Wyoming Health Department and the
Water Quality Division of the Wyoming Department of Environmental Quality to
determine if these departments had received complaints of water quality degradation due
to coalbed methane production. Local authorities reported one complaint of black
sediments in drinking water, but most concerns centered on water loss and flooding
caused by large quantities of water discharged at the surface (Heath, 1999). There has
been discussion among stakeholders regarding the handling of large volumes of water
brought to the surface during coalbed methane production. Some individuals remain
concerned about the consequences of dewatering aquifers, which include loss of the
resource, effects on soil chemistry, flooding, and the potential for coalbed fires and
subsidence.
EPA spoke with a consultant for the Powder River Basin Resource Council (PRBRC), a
citizen's group formed around environmental issues associated with coalbed methane
production (Merchat, 1999). He stated that the biggest concern among people in the area
is loss of water. However, some have had problems with increased methane content in
their water. He said people reported methane in the water results in frothing and bubbles.
The water is generally used for agricultural purposes and for drinking water. He said that
each methane well produces millions of gallons of water in its lifetime. The discharge of
water has created new ponds and swamps that are not naturally occurring in that region.
The secondary effects from pumping water are subsidence and clinker beds (burning
coal). When underground coal catches fire from lightning, it burns until it reaches
groundwater. However, if there is no groundwater, the fire will continue to burn. The
cost of manually extinguishing those fires is enormous. Furthermore, the burning of the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-9
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
coal can leave behind benzo(a)pyrene and other polycyclic aromatic hydrocarbons that
are toxic and/or carcinogenic and could affect drinking water.
EPA Region 8 is participating in a study that addresses the environmental effects of all
aspects of coalbed methane development and not just hydraulic fracturing.
6.3 The Black Warrior Basin (Alabama)
The LEAF v. EPA case arose from an alleged water quality degradation related to
activities in Alabama. As discussed in Chapter 1, the Eleventh Circuit Court's 1997
decision in LEAF v. EPA, 118F.3d 1467, held that because hydraulic fracturing of
coalbeds to produce methane is a form of underground injection, Alabama's EPA-
approved UIC Program must effectively regulate this practice (11th Cir, 1997). In
response to the Court's decision, Alabama supplemented its rules governing the
fracturing of wells to include additional requirements that govern the protection of
USDWs during the hydraulic fracturing of coalbed methane. Summaries of reported
incidents are presented in section 6.3.1 below.
6.3.1 Summary of Reported Incidents
• In the drinking water well case that precipitated LEAF v. EPA, an individual
complained that drinking water from his well contained a milky white
substance and had strong odors shortly after a fracturing event. He also
reported that six months after the fracturing event his water had increasingly
bad odors and occasionally contained black coal fines. The EPA
Administrative Record regarding the Alabama Class II UIC Program contains
other similar descriptions of well water problems.
• Another Alabama citizen reported to EPA problems with her drinking water
well that began in 1989. In her letter, the citizen reported that her property
was located near a coalbed methane gas well and that there was coal mining in
the area. She wrote that she believes hydraulic fracturing of the coalbed
methane well adversely affected her drinking water well, and coal resource
exploitation in the area caused various, significant environmental damage.
The individual believed that the hydraulic fracturing contributed to well
contamination because, shortly after a fracturing event, her kitchen water
contained globs of black, jelly-like grease and smelled of petroleum. She said
her drinking water turned brown and contained slimy, floating particles. She
reported that her neighbors also said their water smelled like petroleum.
She included, as an attachment, a letter from the Alabama Oil and Gas Board
(OGB) approving the use of proppants tagged with radioactive material. Their
approval was based on the hydrogeology and the absence of water wells in the
immediate area, the depths of the coal intervals to be fractured, well
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-10
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
construction, and adherence to a program designed to monitor and contain
radioactive material at the surface. Also attached was a letter from EPA
Region 4 describing analytical results for samples the Agency collected from
her drinking water well on June 26, 1990. The results indicated no purgeable
and extractable organic compounds were detected. In addition, the letter said
that a water/oil inter-phase detector was used to determine if petroleum
products were floating in the well, and none was detected.
• An Alabama homeowner complained to the Natural Resources Defense
Council that recovered hydraulic fracturing fluid from a nearby coalbed
methane well installation was allowed to drain from the coalbed methane well
site to a location near her home. She claimed that this fluid was initially
obtained from an abandoned strip-mining quarry that had been used as a
landfill for municipal and industrial waste. As this fluid drained from the
fracturing site, the homeowner asserted, it killed all animal and plant life in its
path. She further stated that shortly after this fracturing event and the
associated runoff, her 110-foot deep drinking water well became contaminated
with brown, slimy, petroleum-smelling fluid similar to the discharged
fracturing fluid from the coalbed methane well site.
• In response to EPA's July 2001 call for information on water quality incidents
(found in Water Docket W-01-09), an individual reported that her drinking
water well had become filled with methane gas, causing it to hiss (66 FR
39396 (USEPA, 2001)); the tap water became cloudy, oily, and had a strong,
unpleasant odor. In addition, the tap water left behind an oily film and
contained fine particles. The drinking water well owner had her well tested by
a private consultant, who confirmed the presence of methane.
The Alabama OGB tested this drinking water well, but only looked for
naturally occurring contaminants. EPA also sampled and tested this drinking
water well, but not until 6 months after the event. No mention is made of the
analytical results obtained from the drinking water well by these agencies.
6.3.2 State Agency Follow-Up (Alabama Oil and Gas Board)
LEAF v. EPA originated in Alabama. The water well that was reportedly contaminated as
a result of hydraulic fracturing operations was sampled independently by the Alabama
OGB, the Alabama Department of Environmental Management (ADEM), and EPA
Region 4. Water analyses performed by these agencies indicated that the water well had
not been contaminated as a result of the fracturing operation. The Alabama OGB
reported to EPA that it investigates every complaint it receives, and it does not believe
that hydraulic fracturing has affected water wells. Investigations include research into
historical water quality data, some of which pre-dates coalbed methane activity. Such
historical information is important because the coal-bearing Pottsville Formation often
contains high concentrations of iron. Groundwater from this formation may contain iron-
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-11
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
reducing bacteria, which can sometimes result in such water having an unpleasant taste or
odor, or containing a white or red-brown, stringy, gelatinous material (Valkenburg and
others, 1975, as cited by the Alabama OGB, 2002). In addition, sudden iron staining can
occur in water with a history of good quality. Water well yield can also decline due to
the presence of iron-reducing bacteria in high concentrations.
According to the Alabama OGB, one factor considered in each investigation is whether
historical data are available on water quality in a particular area, including data that pre-
date coalbed methane activity. Published reports and open-file data show that the quality
of water in the coal-bearing Pottsville Formation can vary from good to very poor. Data
collected from the 1950s through 1970s in localities throughout a large area where the
Pottsville Formation has served as a source of water contain reports of water having "bad
taste," "bad odors," "oily films or sheens," and waters causing "red stains" and "black
stains" (Geological Survey of Alabama, 1930s to Present; Johnston, 1933, as cited by the
Alabama OGB, 2002).
The Alabama OGB reported to EPA that it has investigated several complaints of
methane gas in water wells. In each instance, the Alabama OGB determined that the
water well problem was unrelated to coalbed methane extraction operations, which often
were not occurring in the areas of reported water problems. Moreover, in some areas
methane gas was reported in water wells many years before the advent of underground
mining and the commercial development of this resource (Geological Survey of
Alabama, 1930s to Present, as cited by the Alabama OGB, 2002). The problem of
methane gas in water wells has generally occurred where water wells, usually less than
200 feet deep, penetrated gas-bearing coal strata, particularly following low rainfall years
that caused a lowering of water tables. In these areas, there commonly had been a recent
increase in the drilling of water wells and an acceleration in the rates of water withdrawal
from the aquifer. When sufficient amounts of water are removed from these water wells,
methane can begin to desorb from the coal seams and be produced.
Alabama's regulations have been approved by EPA for incorporation into Alabama's
Class IIUIC Program. Operators must provide written certification to the Board that the
proposed fracturing operation will not occur in a USDW or that the fracturing fluids do
not exceed the MCLs in 40 CFR §141 Subparts B and G. Fracturing is prohibited from
ground surface to 299 feet bgs. For all fracture jobs performed between 300 feet and 749
feet bgs, the company must perform a reconnaissance of fresh-water supply wells within
!/4 mile of the well to be fractured, submit a fracturing program to the OGB, and perform
a cement bond log analysis. For fracturing events performed between 750 feet and 1,000
feet bgs, only a cement bond log is required. For fracturing events performed below
1,000 feet bgs, operators must submit to the Alabama OGB the depth to be fractured, well
construction information, cementing specifications, and logs identifying overlying,
impervious strata.
In Alabama, Rule 400-3-8-.03 states that coalbeds shall not be hydraulically fractured
until written approval of the Oil and Gas Supervisor has been obtained. The Supervisor
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-12
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
must be notified when an approved fracturing operation is to occur so that an agent of the
Board may be present. In order to receive approval, operators must submit details of the
proposed fracturing operation. The Board's staff evaluates each proposal for compliance
to ensure USDW protection. Basic information that must be submitted with an operator's
proposal to hydraulically fracture a well includes details on the depths of coalbeds to be
fractured; construction of the well, including casing and cementing specifications; a
geophysical log showing the type and thickness of impervious strata overlying the
uppermost coalbed to be fractured; and, if the operation is to be performed in a USDW-
bearing interval, a statement certifying that fracturing fluids will not exceed the MCLs of
federally mandated primary drinking water regulations (40 CFR §141 Subparts B and G).
In addition to the basic information, a fracturing program, a water well inventory within a
H-mile radius, and a cement bond log must be provided with fracturing proposals in the
depth interval 300 to 749 feet. Since water supply wells are generally shallower than
coalbeds, Alabama's Rule 400-3-8-.03 was designed to increasingly strengthen the
requirements for USDW protection with decreasing depths of proposed fracturing
operations. Furthermore, the fracturing of coalbeds shallower than 300 feet is prohibited.
6.4 The Central Appalachian Basin (Virginia and West Virginia)
EPA became aware of several complaints relating to the effects of coalbed methane
production on sources of drinking water in the southwestern portion of Virginia through
correspondence initiated by citizens. Information about water quality incidents was
gathered through meetings and telephone conversations with members of the Virginia
Division of Oil and Gas within the Department of Mines, Minerals and Energy
(VDMME); local health officials; and representatives of a county citizen's group. In
total, VDMME provided EPA with over 70 "Complaint Detail Reports" (registered
between 1990 and 2001) that related to drinking water source impacts by coalbed
methane development.
Although the majority of the incidents outlined in the complaints pertain to water-loss
issues, approximately one-quarter relate to water quality. Virginians living near coalbed
methane production areas reported private well and spring water contamination
evidenced by oily films, soaps, iron oxide precipitates, black sediments, methane gas, and
bad odor and taste. Reports of water loss in the well ranged from noticeably reduced
supply rates to total loss of water from domestic drinking water wells. Summaries of
reported incidents and state follow-up are discussed in sections 6.4.1 and 6.4.2,
respectively.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-13
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
6.4.1 Summary of Virginia Incidents
• The state received complaints of soap bubbles flowing from residential
household fixtures. VDMME attributes soap coming out of water faucets to
the drilling process associated with both conventional wells and coalbed
methane wells. Soaps are used to extract drilling cuttings from the borehole
because the foam expands, rises, and, as it rises, carries the cuttings to the
surface (Wilson, 2001). These soaps may migrate from the borehole into the
drinking water zone that supplies private wells during drilling of the shallow
portion of the hole and before the required groundwater casing is cemented in
place. In the few occurrences of soap contamination, water was provided until
the soap was completely purged from the contributing area surrounding their
water well.
• In early August 2001, EPA met with approximately 15 to 20 residents of
Buchanan and Dickenson Counties in Virginia. Coalbed methane production
activity is steadily increasing in the area surrounding Buchanan County since
the coal reserves in this area have proven to be extremely profitable sources
for coalbed methane in recent years (Wilson, 2001). The subjects of the
citizen complaints were very similar to those logged in the VDMME
complaint reports. Residents described the presence of black sediments, iron
precipitates, soaps, diesel fuel smells, and increased methane gas in drinking
water from their wells. One resident brought a water sample collected from
her drinking water well. The water was translucent with a dark gray color and
with dark black suspended sediment. Several other citizens reported drinking
water supplies diminishing or drying up entirely. One resident of Buchanan
County said that he had an ample water supply from his drinking well for over
54 years, until shortly after coalbed methane wells were installed on his
property. He reported that within 60 days of the coalbed methane well
installations, his 276-foot deep drinking water supply well, which used to
produce over 20 gallons per minute of potable flow, dried up. The resident
mentioned that over 380 homes in the region do not have potable water as a
result of coalbed methane mining activities.
Most of the residents said that their complaints to the state usually resulted in
investigations without resolution. Some residents mentioned that the gas
companies were providing them with potable water to compensate for the
contamination or loss of their drinking water wells. However, the residents
said that this was not adequate compensation for the impacts to, or loss of,
their private drinking water supplies.
• EPA was able to record numerous complaints through telephone
conversations and e-mails with Virginia residents, who reported that they
believed their drinking water wells had been affected by coalbed methane
industry activities. All the logged complaints were from Buchanan and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-14
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
Dickenson Counties. Complaints include water loss, soapy water, diesel
odors, iron and sulfur in wells, rashes from showering, gassy taste, and murky
water. One report discusses a miner who was burned by a fluid, possibly
hydrochloric acid used in hydraulic fracturing, that infiltrated a mineshaft.
Another report describes the contamination of a stream and the resulting fish
kills caused by the runoff from drilling fluids. One complainant explained
that several thousand wells had "gone dry, overnight." According to the
individuals EPA spoke with, compensation to homeowners for these impacts
is in the form of money, newly drilled wells to replace dry or contaminated
wells or temporary provision of potable water, which is supplied "until things
clear out."
6.4.2 State Agency Follow-Up (VDMME)
VDMME, Division of Gas and Oil, is responsible for responding to environmental issues
associated with oil and gas development; it investigates every water problem reported.
Responses may include an interview with the citizen reporting the problem, a site visit,
water well testing, or a review of the physical aspects of the water well and surrounding
activities. According to Robert Wilson of VDMME, his agency tests for contaminants
that may be introduced by drilling such as chlorides, oil and grease, and volatile organics.
The results of those analyses are compared to baseline values. VDMME witnesses
surface casing and plugging jobs as part of its oversight duties. VDMME reviews
information from drilling and completion reports to assist with investigations into
complaints.
Based on investigations of the more than 70 complaints received, VDMME believes that
coalbed methane production has not affected private drinking water wells. VDMME
recognizes soap migrating into drinking water wells, but considers this only a transient
problem. While a number of complaints report a noticeable reduction in or a total loss of
drinking water supply, in almost all cases, the state investigator determined that the water
loss was not likely to be caused by local hydraulic fracturing events or coalbed methane
production activity because:
• The distance from the private well to the nearest coalbed methane well is too
far (1,500 feet or more) to have any impact.
• There is no hydrologic connection between the water contribution zones of the
private and coalbed methane wells; therefore, it is physically impossible for
coalbed methane wells to affect private drinking water wells.
• The well was constructed according to VDMME regulatory guidelines;
therefore, a sufficient buffer exists between the private well and the coalbed
methane well.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-15
image:
EPA 816-R-04-003 Chapter 6
Water Quality Incidents
• The existing supply was reduced because of recent drought conditions in the
region.
• The complainant experienced mechanical difficulty with his or her pumping
system, which led to a reduction in pumped water; however, the supply was
not affected.
According to VDMME, these citizen complaints refer to incidents that can occur during
the drilling of any type of well, not just coalbed methane. The few incidents of this kind
were equally divided between conventional wells and coalbed wells (VDMME, 2002).
6.5 Summary
In this chapter, EPA has presented information (in addition to technical, conceptual, or
theoretical information presented previously) on personal experiences with regard to
coalbed methane activities and their potential (or perceived potential) to impact drinking
water wells. These personal accounts of potential incidences in four producing coal
basins across the United States do not present scientific findings. However, the body of
reported problems considered collectively suggest that water quality (and quantity)
problems might be associated with some of the production activities common to coalbed
methane extraction. These activities include surface discharge of fracturing and
production fluids, aquifer/formation dewatering, water withdrawal from production wells,
methane migration through conduits created by drilling and fracturing practices, or any
combination of these. Other potential sources of drinking water problems include various
aspects of resource development, naturally occurring conditions, population growth and
historical practices.
In several of the coalbed methane investigation areas, local agencies concluded that
hydraulic fracturing could not affect drinking water wells. Generally, these conclusions
were based on there being a significant horizontal and/or vertical distance between the
coalbed methane production wells and the drinking water wells.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 6-16
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
Chapter 7
Conclusions and Recommendations
Under SDWA, EPA's UIC Program is responsible for ensuring that fluids injected into
the ground do not endanger USDWs. The goal of the Phase I study was to assess the
potential for contamination of USDWs due to the injection of hydraulic fracturing fluids
into coalbed methane wells, and to determine, based on these findings, whether further
study is warranted.
EPA's approach for evaluating the potential for contamination of USDWs was an
extensive information collection and review of empirical and theoretical data. EPA
reviewed water quality incidents potentially associated with hydraulic fracturing and
evaluated the theoretical potential for hydraulic fracturing to affect the quality of USDWs
through one of two mechanisms:
1. Direct injection of fracturing fluids into a USDW in which the coal is located,
or injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
7.1 Reported Water Quality Incidents
Citizens from Wyoming, Montana, Alabama, Virginia, Colorado, and New Mexico
contacted EPA because they were concerned that their water wells were affected by
coalbed methane production. The major geographic areas where citizens reported
experiencing problems due to coalbed methane development are concentrated in the coal
basins with the most coalbed methane activities - the San Juan, Black Warrior, Central
Appalachian, and Powder River Basins. This study was initiated, partly, in response to
those citizens' concerns. EPA followed-up on letters and telephone calls from citizens
and resulting leads to understand specific complaints and citizens' concerns.
EPA published a Federal Register notice (66 FR 39396 (USEPA, 2001) requesting
information on water quality incidents believed to be associated with hydraulic fracturing
of coalbed methane wells. EPA notified over 500 local and/or county agencies in areas
with potential coalbed methane production activity to make them aware of the Federal
Register notice requesting information on coalbed methane-related complaints. The
Agency received no information on complaints from these officials.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-1
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
EPA reviewed responses and follow-up actions conducted by state agencies to address
groundwater complaints involving coalbed methane. Hydraulic fracturing is not widely
practiced in the Powder River Basin (which includes Wyoming and Montana) and
concerned citizens from that area reported surface water and groundwater quantity
problems rather than specifying hydraulic fracturing as a problem. Studies of
groundwater quality in the San Juan Basin (which includes parts of Colorado and New
Mexico) do not address hydraulic fracturing directly. However, problems with
groundwater quantity and quality in Colorado may have plausible explanations other than
hydraulic fracturing activities. For example, natural fractures, and poorly constructed,
sealed, or cemented wells used for various purposes, may provide conduits for methane
to move into shallow geologic strata and water wells, or even to surface water (BLM,
1999). The New Mexico Oil Conservation Division reported that citizens began
reporting increased levels of methane in their water wells after coalbed methane
development began in the San Juan Basin. New Mexico initiated a plugging and
abandonment program to seal old, improperly abandoned production wells, which
appears to have mitigated the problem (Chavez, 2001).
EPA also obtained individual incident reports from Virginia. None of Virginia's follow-
up investigations provided evidence that hydraulic fracturing of coalbed methane wells
had caused drinking water well problems. Incidents in Alabama were investigated by the
Alabama Oil and Gas Board, the Alabama Department of Environmental Management,
and EPA Region IV. Samples from drinking water wells did not test positive for
constituents found in fracturing fluids. After reviewing all the available data and incident
reports, EPA sees no conclusive evidence that water quality degradation in USDWs is a
direct result of injection of hydraulic fracturing fluids into coalbed methane wells and
subsequent underground movement of these fluids.
7.2 Fluid Injection Directly into USDWs or into Coal Seams Already In
Hydraulic Communication with USDWs
To determine if USDWs are threatened by the direct injection of fracturing fluids into a
USDW, EPA: 1) reviewed information on 11 major U.S. coal basins mined for coalbed
methane to determine if coal seams lie within USDWs, and 2) identified components of
fracturing fluids. EPA also used the information on the 11 major U.S. coal basins as well
as information collected on water quality incidents potentially associated with hydraulic
fracturing to determine if coal seams are already in hydraulic communication with
USDWs. Hydraulic fracturing has been, or is being, performed in every basin reviewed.
As summarized in Table 5-1 in Chapter 5, evidence suggests that coalbeds in 10 of the 11
major coal basins in the United States are located at least partially within USDWs. The
coalbeds in the Piceance Basin in Colorado, however, are several thousand feet below
USDWs, and are unlikely to be in hydraulic communication with USDWs.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-2
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
Hydraulic fracturing fluids injected into coalbed methane wells consist primarily of
water, or inert nontoxic gases, and/or nitrogen foam and guar (a naturally occurring
substance derived from plants). According to information gathered from MSDSs, on-site
reconnaissance of fracturing jobs, and interviews with service company employees, some
hydraulic fracturing fluids may contain constituents of potential concern. Table 4.1 in
Chapter 4 lists examples of chemicals found in hydraulic fracturing fluids according to
the MSDSs. Constituents of potential concern include the following substances either
alone or in combination: bactericides, acids, diesel fuel, solvents, and/or alcohols.
Although the largest portion of fracturing fluid constituents is nontoxic (>95% by
volume), direct fluid injection into USDWs of some potentially toxic chemicals does take
place.
For example, potentially hazardous chemicals are introduced into USDWs when diesel
fuel is used in fracturing fluids in operations targeting coal seams that lie within USDWs.
Diesel fuel contains constituents of potential concern regulated under SDWA - benzene,
toluene, ethylbenzene, and xylenes (i.e., BTEX compounds). However, the threat posed
to USDWs by introduction of these chemicals is reduced significantly by coalbed
methane production's dependence on the removal of large quantities of groundwater (and
injected fracturing fluids) soon after a well has been hydraulically fractured. EPA
believes that this groundwater production, combined with the mitigating effects of
dilution and dispersion, adsorption, and potentially biodegradation, minimize the
possibility that chemicals included in the fracturing fluids would adversely affect
USDWs.
Because of the potential for diesel fuel to be introduced into USDWs, EPA requested, and
the three major service companies agreed, to eliminate diesel fuel from hydraulic
fracturing fluids that are injected directly into USDWs for coalbed methane production.
Industry representatives estimate that these three companies perform approximately 95
percent of the hydraulic fracturing projects in the United States. These companies signed
an MOA on December 15, 2003 and have indicated to EPA that they no longer use diesel
fuel as a hydraulic fracturing fluid additive when injecting into USDWs for coalbed
methane production (USEPA, 2003).
7.3 Breach of Confining Layer
The second mechanism by which hydraulic fracturing may affect the quality of USDWs
is fracturing through a hydrologic confining layer, and creation of a hydraulic
communication between a coal seam and an overlying USDW. If sufficiently thick and
relatively unfractured shale is present, however, it may act as a barrier not only to
fracture height growth, but also to fluid movement.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-3
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
A hydraulic fracture will propagate perpendicularly to the minimum principal stress. In
some shallow formations, the least principal stress is the overburden stress; thus, the
hydraulic fracture will be horizontal. In deeper reservoirs, the least principal stress will
likely be horizontal; thus, the hydraulic fracture will be vertical. In general, horizontal
fractures are most likely to exist at shallow depths (less than 1,000 feet) (Nielsen and
Hansen, 1987 as cited in Appendix A: DOE, Hydraulic Fracturing). Most coal seams
currently used for methane production are relatively shallow compared to conventional
oil production wells, but still lie deeper than 1,000 feet.
Hydraulic fracturing may have increased or have the potential to increase the
communication between coal seams and adjacent formations in some instances. For
example, in the Raton Basin, some fracturing treatments resulted in higher than expected
withdrawal rates for production water. Those increases, according to literature published
by the Colorado Geologic Survey, may be due to well stimulations creating a connection
between targeted coal seams and an adjacent sandstone aquifer (Hemborg, 1998). In the
Powder River Basin, concerns over the creation of such a hydraulic connection are cited
as one reason why hydraulic fracturing of coalbed methane reservoirs is not widely
practiced in the region. Some studies that allow direct observation of fractures (i.e.,
mined-through studies) also provided evidence that fractures move through interbedded
layers, sometimes taking a stair-step pathway through complex fracture systems, and
sometimes enter or propagate through geologic strata above the coal (i.e., roof rock)
(Diamond, 1987a and b; Diamond and Oyler, 1987; Jeffrey et al., 1993).
Fracture height is important to the issue of whether or not hydraulic fracturing fluids can
affect USDWs because shorter fractures are less likely to extend into a USDW or connect
with natural fracture systems that may transport fluids to a USDW. The extent of a
fracture is controlled by the characteristics of the geologic formation (including the
presence of natural fractures), the volume and types of fracturing fluid used, the pumping
pressure, and the depth at which the fracturing is being performed. Deep vertical
fractures can propagate vertically to shallower depths and develop a horizontal
component (Nielsen and Hansen, 1987, as cited in Appendix A: DOE, Hydraulic
Fracturing). In these "T-fractures," the presence of coal fines or a zone of stress contrast
may cause the fracture to "turn" and develop horizontally, sometimes at the contact of the
coalbed and an overlying formation (Jones et al., 1987b; Morales et al., 1990).
The low permeability of relatively unfractured shale may help to protect USDWs from
being affected by hydraulic fracturing fluids in some basins. At some sites, shale may act
not only as a hydraulic barrier, but also as a barrier to fracture height growth. Shale's
ability to act as a barrier to fracture height growth is due primarily to the stress contrast
between the coalbed and the higher-stress shale (see Appendix A)
Another factor controlling fracture height can be the highly cleated nature of some
coalbeds. In some cases, highly cleated coal seams will prevent fractures from growing
vertically. When the fracture fluid enters the coal seam, it is contained within the coal
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-4
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
seam's dense system of cleats and the growth of the hydraulic fracture will be limited to
the coal seam (see Appendix A).
Mined-through studies indicate many hydraulic fractures that penetrate into, or
sometimes through, formations overlying coalbeds can be attributed to the existence of
pre-existing natural fractures. However, given the concentrations and flowback of
injected fluids, and the mitigating effects of fate and transport processes, EPA does not
believe that possible hydraulic connections under these circumstances represent a
significant potential threat to USDWs.
7.4 Conclusions
Based on the information collected and reviewed, EPA has concluded that the injection of
hydraulic fracturing fluids into coalbed methane wells poses little or no threat to USDWs
and does not justify additional study at this time. This decision is consistent with the
process outlined in the April, 2001 Final Study Design, in which EPA indicated that it
would determine whether further investigation was needed after analyzing the Phase I
information. Specifically, EPA determined that it would not continue into Phase II of the
study if the investigation found that no hazardous constituents were used in fracturing
fluids, hydraulic fracturing did not increase the hydraulic connection between previously
isolated formations, and reported incidents of water quality degradation were attributed to
other, more plausible causes.
Although potentially hazardous chemicals may be introduced into USDWs when
fracturing fluids are injected into coal seams that lie within USDWs, the risk posed to
USDWs by introduction of these chemicals is reduced significantly by groundwater
production and injected fluid recovery, combined with the mitigating effects of dilution
and dispersion, adsorption, and potentially biodegradation. Additionally, EPA has
reached an agreement with the major service companies to voluntarily eliminate diesel
fuel from hydraulic fracturing fluids that are injected directly into USDWs for coalbed
methane production.
Often, a high stress contrast between adjacent geologic strata results in a barrier to
fracture propagation. This may occur in those coal zones where there is a geologic
contact between a coalbed and a thick, higher-stress shale that is not highly fractured.
Some studies that allow direct observation of fractures (i.e., mined-through studies)
indicate many fractures that penetrate into, or sometimes through, formations overlying
coalbeds can be attributed to the existence of pre-existing natural fractures. However,
and as noted above, given the concentrations and flowback of injected fluids, and the
mitigating effects of dilution and dispersion, fluid entrapment, and potentially
biodegradation, EPA does not believe that possible hydraulic connections under these
circumstances represent a significant potential threat to USDWs.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-5
image:
EPA 816-R-04-003 Chapter 7
Conclusions and Recommendations
EPA also reviewed incidents of drinking water well contamination believed to be
associated with hydraulic fracturing and found no confirmed cases that are linked to
fracturing fluid injection into coalbed methane wells or subsequent underground
movement of fracturing fluids. Although thousands of coalbed methane wells are
fractured annually, EPA did not find confirmed evidence that drinking water wells have
been contaminated by hydraulic fracturing fluid injection into coalbed methane wells.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs 7-6
image:
Public Comment and
Response Summary
for the Study on the Potential
Impacts of Hydraulic Fracturing
of Coalbed Methane Wells on
Underground Sources of
Drinking Water
FINAL
image:
Office of Water
Office of Ground Water and Drinking Water (4606M)
EPA816-R-04-004
www.epa.gov/safewater
June 2004
Printed on Recycled Paper
image:
EPA816-R-04-004
Public Comment and Response Summary
for the Study on the Potential Impacts of
Hydraulic Fracturing of Coalbed Methane Wells on
Underground Sources of Drinking Water
FINAL
June 2004
United States Environmental Protection Agency
Office of Water
Office of Ground Water and Drinking Water
Drinking Water Protection Division
Prevention Branch
1200 Pennsylvania Avenue, NW (4606M)
Washington, DC 20460
image:
TABLE OF CONTENTS
LIST OF ACRONYMS AND ABBREVIATIONS iv
I. INTRODUCTION 1
II. SCOPE OF THE STUDY 5
A. Areas Not Included in the Review 5
1. Focus of the Report 5
2. Monitoring 6
3. Use of Modeling Results 8
B. Literature Used for the Study 8
C. Basins Included in the Study 9
D. Citizen Complaints/Instances of Water Well Contamination 9
E. Peer Review Panel 11
III. FRACTURE FLUIDS 12
A. Components of Fracturing Fluids 12
1. Health Effects 12
2. Diesel Fuel 13
3. MTBE 14
B. Comparison of Concentrations of Hydraulic Fracturing Fluid Components to MCLs ..15
C. Concentrations of Constituents in Fracturing Fluids/Fluid Recovery Rates 15
1. Estimates of Concentrations of Constituents in Fracturing Fluids 15
2. Fluid Recovery Rates 18
3. Amount of Fracturing Fluids 19
4. Movement of Fracturing Fluids 19
IV. FRACTURE BEHAVIOR AND PRACTICES 20
A. Fracture Growth 20
B. Multiple Fractures 21
C. Relationship of Drinking Water Wells to Hydraulic Fracturing Activities 22
D. Differences in State Geology 22
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study i FINAL
image:
V. REGULATION OF HYDRAULIC FRACTURING PRACTICES 23
A. States' Authority 23
B. Regulation of Hydraulic Fracturing under SDWA 23
VI. LANGUAGE USED IN THE REPORT 24
A. Use of the Term "USDW" 24
B. Use of Scientific Terms 25
C. Use of Qualifying Language 25
VII. CHAPTER-SPECIFIC COMMENTS 26
A. Glossary 26
B. Other Executive Summary Comments 26
C. Other Chapter 1 Comments (Introduction) 27
D. Other Chapter 2 Comments (Methodology) 28
E. Other Chapter 3 Comments (Characteristics of CBM Production and HF Practices) ... 28
F. Other Chapter 4 Comments (HF Fluids) 28
G. Other Chapter 5 Comments (Basin Descriptions) 28
H. Other Chapter 6 Comments (Water Quality Incidents) 29
I. Other Chapter 7 Comments (Conclusions and Recommendations) 30
VIII. BASIN DESCRIPTIONS 30
A. San Juan Basin 30
B. Black Warrior Basin 31
C. Piceance Basin 31
D. Uinta Basin 31
E. Powder River Basin 31
F. Central Appalachian Basin 31
G. Northern Appalachian Basin 32
H. Western Interior Basin 32
I. Raton Basin 32
J. Sand Wash Basin 32
K. Washington Coal Regions (Pacific and Central) 32
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study ii FINAL
image:
LIST OF TABLES
TABLE 1: LIST OF PUBLIC COMMENTERS
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study iii FINAL
image:
LIST OF ACRONYMS AND ABBREVIATIONS
BLM
BTEX
CBM
CCL
CFR
COGCC
EIS
EPA
FR
GWPC
MCL
MOA
MSDS
MTBE
NAS
PWS
RfD
SDWA
UCMR
UIC
USDW
Bureau of Land Management
Benzene, Toluene, Ethylbenzene, and Xylenes
Coalbed Methane
Contaminant Candidate List
Code of Federal Regulations
Colorado Oil and Gas Conservation Commission
Environmental Impact Statement
United States Environmental Protection Agency or Agency
Federal Register
Ground Water Protection Council
Maximum Contaminant Level
Memorandum of Agreement
Material Safety Data Sheet
Methyl Tert Butyl Ether
National Academy of Science
Public Water System
Reference Dose
Safe Drinking Water Act
Unregulated Contaminant Monitoring Regulation
Underground Injection Control Program
Underground Source of Drinking Water
Public Comment and Response Summary for
Hydraulic Fracturing CBM Study
IV
June 2004
FINAL
image:
Public Comment and Response Summary
for the Study on the Potential Impacts of Hydraulic Fracturing of
Coalbed Methane Wells on Underground Sources of Water
I. INTRODUCTION
The United States Environmental Protection Agency's (EPA's) Office of Ground Water and
Drinking Water completed its Phase I study, which assesses the potential for contamination of
underground sources of drinking water (USDWs) from the injection of hydraulic fracturing
fluids into coalbed methane (CBM) wells. EPA (or the Agency) began collecting information on
hydraulic fracturing in the fall of 2000. Based on the information collected and reviewed, EPA
has concluded that the injection of hydraulic fracturing fluids into CBM wells poses little or no
threat to USDWs and does not justify additional study at this time.
The draft report, titled, "Draft Evaluation of Impacts to Underground Sources of Drinking Water
by Hydraulic Fracturing of Coalbed Methane Reservoirs" (hereafter referred to as the draft
report), was made available for public comment by an announcement in the Federal Register on
August 28, 2002.J The 60-day public comment period officially ended on October 28, 2002.
The Agency received and reviewed comments from 105 commenters. Several of these were
signed by multiple parties (which were counted as one commenter), including a few coalitions of
environmental organizations. The commenters include private citizens; environmental and
citizen groups; government agencies at the local, state, and national levels; oil and gas
companies; trade associations; and four other commenters that do not fit these categories. Table
1 below provides a listing of these commenters.
1 US Environmental Protection Agency. 2002. Underground Injection Control (UIC) Program; Hydraulic
Fracturing of Coalbed Methane (CBM) Wells Report—Notice. Federal Register. Vol. 67, No. 167. p. 55249, August
28, 2002.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 1 FINAL
image:
TABLE 1 : LIST OF PUBLIC COMMENTERS
Docket ID1
Edocket ID
(OW-2001-0002)2
Organization (State)
Environmental/Citizens Groups
II-D1.014
II-D1.025
II-D1.040
II-D1.046
II-D1.055
II-D1.060
II-D1.072
II-D1.101
II-D1.076; II-D2.001; II-D2.002
045
055
068
074
043
085
100
139
106-109
Bull Mountain Landowners Association (MT)
Land and Water Fund of the Rockies (CO)
Dickenson County Citizens Committee (VA)
Western Organization of Resource Councils and Coalition of
11 Other Environmental/Citizens Groups (DC)
Coalition of 28 Environmental/Citizens Groups (varies)
Oil & Gas Accountability Project and Coalition of 34 Other
Environmental/Citizens Groups (CO)
National Resources Defense Council (DC)
San Juan Citizen's Alliance (CO)
Kentucky Resources Council, Inc. (KY)
Private Citizens
II-D1.004
II-D1.050
II-D1.012; II-D1.017
II-D1.002; II-D1.003; II-D1.006;
II-D1.008; II-D1.009; II-D1.011;
II-D1.016; II-D1.018; II-D1.022;
II-D1.023; II-D1.024; II-D1.026;
II-D1.030; II-D1.031; II-D1.032;
II-D1.034; II-D1.037; II-D1.038;
II-D1.043; II-D1.044; II-D1.049;
II-D1.058; II-D1.065; II-D1.067;
II-D1.081; II-D1.083; II-D1.084;
II-D1.085; II-D1.086; II-D1.087;
II-D1.088; II-D1.089; II-D1.093;
II-D1.095; II-D1.097; II-D1.099;
II-D1.100; II-D1.102; II-D2.008
II-D1.015; II-D1.027; II-D1.029;
II-D1.041; II-D1.098
II-D1.007
II-D1.039; II-D1.048; II-D2.007
II-D1.005; II-D1.033; II-D1.051
II-D1.013; II-D1.019
II-D1.042
II-D1.028; II-D1.094
033
031
041; 048
110; 032; 035;
037; 038; 040;
047; 049; 052;
053; 054; 056;
060; 061; 112;
128; 065; 066;
071; 072; 075;
083; 092; 094;
118; 120; 121;
122; 123; 124;
125; 126; 131;
133; 135; 137;
138; 140; 148
046; 057; 059;
069; 136
036
067; 030; 142
034; 062; 076
044; 050
070
058; 132
Citizen (AK)
Citizen (AL)
Citizen (CA - 2)
Citizen (CO - 39)
Citizen (FL - 5)
Citizen (KS)
Citizen (MT - 3)
Citizen (NM - 3)
Citizen (NY - 2)
Citizen (UT)
Citizen (state unknown - 2)
State/Local/Federal Agencies
II-D1.010
II-D1.045
II-D1.047
II-D1.057
II-D1.059
II-D1.061
II-D1.062
II-D1.063
II-D1.064
II-D1.066
039
073
029
082
084
086
087; 088
089
090
093
Sandia National Laboratories (NM)
San Miguel County Board of Commissioners (CA)
Alabama Oil and Gas Board (AL)
State of New Mexico Energy, Minerals and Natural
Resources Department (NM)
Virginia Division of Gas and Oil (VA)
Colorado Geological Survey (CO)
Michigan Department of Environmental Quality (Ml)
Pennsylvania Department of Conservation and Natural
Resources (PA)
State of Utah Department of Natural Resources, Division of
Oil, Gas and Mining (UT)
Alaska Oil and Gas Conservation Commission (AK)
Public Comment and Response Summary for
Hydraulic Fracturing CBM Study
June 2004
FINAL
image:
TABLE 1 : LIST OF PUBLIC COMMENTERS
Docket ID1
II-D1.068
II-D1.069
II-D1.073
II-D1.079
II-D1.080
II-D1.082
II-D1.092
II-D1.096
II-D1.103
II-D2.006
II-D2.009
Edocket ID
(OW-2001-0002)2
095
096
101; 102
116
117
119
130
134
147
141
149
Organization (State)
State of South Dakota (SD)
Ohio Department of Natural Resources (OH)
Conservation Division of the Kansas Corporation
Commission (KS)
State of Louisiana, Department of Natural Resources (LA)
Colorado Oil & Gas Conservation Commission (CO)
State of Missouri Department of Natural Resources,
Geological Survey & Resource Assessment Division (MO)
Indiana Department of Natural Resources, Division of Oil and
Gas (IN)
State of Oklahoma, Office of the Secretary of Energy (OK)
Delta County Commissioners (CO)
Office of Fossil Energy, Department of Energy (DC)
Ohio Department of Natural Resources, Division of Mineral
Resources Management (OH)
Oil and Gas Companies
II-D1.070
II-D1.075
II-D1.090
097
105
127
Halliburton Energy Services (TX)
Chevron Texaco North American Upstream (TX)
Shell Exploration & Production Company (TX)
Trade Associations
II-D1.035
II-D1.036
II-D1.052
II-D1.053
II-D1.054
II-D1.056
II-D1.071
II-D1.074
113
064
077
080
042
081
099
104
Domestic Petroleum Council (DC)
Independent Petroleum Association (DC)
Interstate Oil and Gas Compact Commission (OK)
Independent Oil & Gas Association of West Virginia (WV)
Coalbed Methane Association of Alabama (AL)
Oklahoma Independent Petroleum Association (OK)
Ground Water Protection Council (OK)
American Petroleum Institute (DC)
Other
II-D1.020
II-D1.021
II-D1.077
II-D1.078
051
111
114
129
Pace Law School (NY)
University of Montana, Montana Bureau of Mines and
Geology, Montana Tech (MT)
Steven Harper, Attorney at Law (CO)
Hansen Environmental Consultants (WA)
1 Docket Identification numbers are assigned by the Water Docket in order to track each public comment with a
unique identification number. Note that if a comment has a prefix of "II-D2," it indicates that the comment was
received after the October 28, 2002 comment deadline. Comments with the following docket logs were updates,
repeats, or clarifications of other comments: II-D1.91; II-D2.03; II-D2.04; and II-D2.05.
2 An electronic version of each public comment is available through EPA's electronic public docket and comment
system, EPA Dockets at http://www.epa.gov/edocket/. Each comment begins with the prefix "OW-2001-0002-".
Edocket numbers were assigned to comment materials, as well as other relevant background documents in the
order they were posted to the edocket Web site.
Public Comment and Response Summary for
Hydraulic Fracturing CBM Study
June 2004
FINAL
image:
The remainder of this document contains summaries of the major public comments and EPA's
responses related to the Agency's August 2002 report. The document is divided into seven other
major sections as follows:
Section II: Scope of the Study discusses public comments and EPA's responses on
areas not included in the study, the literature used for the review, the number of coal
basins included in the study, citizen complaints regarding water well contamination,
and the peer review panel who reviewed the initial draft of the report.
Section III: Fracturing Fluids describes public comments and EPA's responses
related to the components of fracturing fluids, EPA's comparison of the concentration
of fracturing fluid constituents to maximum contaminant levels (MCLs), EPA's
estimates for the concentrations of fracturing fluid chemicals at the point-of-injection
and the edge of the fracture zone, the amount of fracturing fluids that is recovered
from CBM reservoirs, the amount of fracturing fluids used in hydraulic fracturing
procedures, and the movement of "stranded" fluids in the coalbed formations.
• Section IV: Fracture Behavior and Practices discusses comments raised and
EPA's responses to these comments regarding fracture growth, multiple fracturing of
the same well, the relationship of drinking water wells to hydraulic fracturing
activities, and differences in state geology.
• Section V: Regulation of Hydraulic Fracturing Practices describes comments and
the Agency's responses regarding the states' authority over hydraulic fracturing
practices, and the regulation of hydraulic fracturing under the Safe Drinking Water
Act (SOWA).
• Section VI: Language Used in the Report summarizes specific comments and the
Agency's responses related to the use of the term "USDW" in the report, use of
scientific terms, and the tone of the language in the report.
Section VII: Chapter-Specific Comments describes comments and the Agency's
response regarding the glossary, executive summary, and Chapters 1 through 7 that
were not already covered under Sections II through VI of this document.
Section VIII: Basin Descriptions describes comments that pertain to the basin-
specific descriptions in Attachments 1 through 11 of the report and EPA's response to
these comments. The comments and responses in Section VIII do not include
comments that were already discussed in Sections II through VII of this document.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 4 FINAL
image:
II. SCOPE OF THE STUDY
A. Areas Not Included in the Review
1. Focus of the Report
Summary of Comments: One commenter indicated that the report should have focused on the
possible impacts to human health instead of the hydraulic fracturing process. This commenter
added that Chapter 4 of the report should have focused on dose-response curves and not on the
properties of hydraulic fracturing fluids. The commenter also stated that EPA should have been
able to conduct this analysis because the Agency should have access to research conducted on
the toxicity of all constituents used in CBM production.
Another commenter stated that the study did not address the uncertainty in the risk assessment
due to omissions and errors in the data used for the study. This commenter indicated that some
of the reasons for these omissions and errors could be inadequate reporting by private well
owners and counties, inadequate testing, and inadequate enforcement which would result in an
underassessment of risk. This commenter also indicated that the report does not address risk
resulting from deviations and failures in drilling, fracturing, and monitoring practices, especially
for newer wells, or sufficiently address the testing error for volatile chemicals used in hydraulic
fracturing.
EPA Response: The Phase I study was not intended to be a risk assessment, but rather, to be a
fact-finding effort based primarily on existing literature to assess the potential threat to USDWs
from the injection of hydraulic fracturing fluids into CBM wells and to determine based on these
findings, whether additional study is warranted. The study is tightly focused on hydraulic
fracturing of CBM wells and does not include other aspects of drilling or CBM production. EPA
reviewed water quality incidents potentially associated with hydraulic fracturing, as well as
evaluated the theoretical potential for hydraulic fracturing to affect USDWs. EPA researched
over 200 peer-reviewed publications, interviewed approximately 50 employees from industry
and state or local government agencies, and communicated with approximately 40 citizens and
groups who are concerned that CBM production affected their drinking water wells.
For the purposes of this study, EPA assessed USDWs impacts by the presence or absence of
documented drinking water well contamination cases caused by CBM hydraulic fracturing, clear
and immediate contamination threats to drinking water wells from CBM hydraulic fracturing,
and the potential for CBM hydraulic fracturing to result in USDW contamination based on two
possible mechanisms described below.
1. Direct injection of fracturing fluids into a USDW in which the coal is located, or
injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an adjacent
USDW.
EPA's report includes a discussion of the types of fracturing fluids and additives, and fluid
volumes that may be used in hydraulic fracturing operations. This discussion is intended to
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 5 FINAL
image:
provide further background on the hydraulic fracturing process. In addition, the study provides a
review of the fate and transport of injected fluids in the subsurface in order to determine whether
a detailed risk assessment is warranted.
2. Monitoring
Summary of Comments: Several commenters questioned how EPA could decide whether
hydraulic fracturing poses a risk to USDWs without collecting or reviewing monitoring data.
Several commenters wanted EPA to proceed to Phase II of the study and to install monitoring
wells in areas where hydraulic fracturing of CBM wells was occurring. One commenter
recommended that, at a minimum, EPA identify whether any type of monitoring has been
conducted by consulting firms, local or state agencies, or members of the academic community,
and if this monitoring exists, to include the results in the report.
Another commenter recommended that EPA, in cooperation with the National Academy of
Science (NAS), conduct unannounced inspections of hydraulic fracturing projects in order to
collect samples of hydraulic fracturing fluids, and observe and measure the total volume of
injected hydraulic fracturing fluid. This commenter also recommended that EPA establish
reference doses (RfDs) and MCLs for all chemicals currently used in hydraulic fracturing fluids
in significant volumes.
EPA Response: EPA has researched and reviewed a variety of monitoring information that may
be related to the issue of possible conduits for fracturing fluid transport into USDWs. These data
are discussed in Chapter 6 of the report. For example, EPA reviewed a 1999 Bureau of Land
Management (BLM) report which focused on monitoring and data interpretation of methane
concentrations in groundwater in the San Juan Basin area. EPA reviewed this report to
determine if it contained information pertaining to hydraulic fracturing of CBM and its impacts,
if any, to the quality of water in drinking water aquifers in this basin.
Chapter 6 of the report provides a detailed discussion of citizen complaints and state responses to
their concerns. Complaints were responded to by various state agencies, and many of those
responses included testing of water for contaminants. For example, the Virginia Department of
Mines, Minerals and Energy is responsible for: responding to environmental issues associated
with oil and gas development (including CBM); investigating all reported water problems; and
testing water samples for contaminants that may be introduced by drilling (such as chlorides, oil
and grease, and volatile organics).
EPA disagrees that monitoring data is needed to determine whether a Phase II study is
warranted. As discussed in the previous response, EPA conducted an extensive literature review,
conducted numerous interviews, reviewed water quality incidents potentially associated with
hydraulic fracturing, and evaluated the theoretical potential for hydraulic fracturing to affect
USDWs. EPA's decision that the injection of hydraulic fracturing fluids into CBM wells poses
little or no threat to USDWs and does not justify additional study at this time is consistent with
the process outlined in the April, 2001 Final Study Design. In its final study design, EPA
indicated that the Agency would make a determination regarding whether further investigation
was needed after analyzing the Phase I information.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 6 FINAL
image:
EPA has recently taken a specific and important measure to address one of the primary concerns
regarding hydraulic fracturing fluid - the use of diesel fuel. During EPA's research, the Agency
realized that diesel is sometimes used a component of fracturing fluids and is of specific concern
because it contains BTEX compounds (benzene, toluene, ethylbenzene, and xylenes) for which
MCLs have been established under SDWA. Because of the potential problem diesel can cause,
EPA requested its removal from hydraulic fracturing fluids. On December 15, 2003, EPA
entered into a Memorandum of Agreement (MOA) with three major service companies - BJ
Services Company, Halliburton Energy Services, Inc., and Schlumberger Technology
Corporation - to voluntarily eliminate diesel fuel from hydraulic fracturing fluids that are
injected directly into USDWs for CBM production. If necessary, these companies will select
replacements that will not cause hydraulic fracturing fluids to endanger USDWs. Industry
representatives estimate that these three companies conduct an estimated 95 percent of the
hydraulic fracturing projects in the United States. These three have indicated to EPA that they
no longer use diesel fuel as a hydraulic fracturing fluid additive when injecting into USDWs.
EPA, through its Underground Injection Control (UIC) Program, as authorized under SDWA
Part C, Sections 1421-1426), is responsible for ensuring that fluids injected into the ground do
not endanger USDWs or cause a public water system (PWS) to violate its drinking water
standards due to the contamination of a USDW by these injected fluids. Most states have
primary enforcement authority (primacy) for implementation of the UIC Program, and thus have
the authority under SDWA to place controls on any injection activities that may threaten
USDWs. 40 CFR 145.12, Requirements for Compliance Evaluation Programs, requires that
authorized states have programs for periodic inspections of injection operations. States may also
have additional authorities by which they can regulate hydraulic fracturing. While surprise
inspections are not specifically mandated, state programs have a responsibility to conduct
inspections, as necessary, to determine compliance with permit conditions, and to verify the
accuracy of monitoring data and other information. EPA requires that all UIC inspectors be
certified in, and that inspectors be knowledgeable about, proper operation of injection facilities,
protection of USDWs, and SDWA requirements.
Regarding the establishment of RfDs and MCLs for all hydraulic fracturing fluid chemicals used
in significant volumes, EPA follows an established procedure for identifying the contaminants
for which these standards will be set. The Contaminant Candidate List (CCL) and the
Unregulated Contaminant Monitoring Regulation (UCMR) are the primary review mechanisms
by which EPA identifies drinking water contaminants which pose the most urgent threat to
public health. The CCL process uses the best available information on contaminants of concern
and emerging contaminants to prioritize according to potential public health threat, and identify
candidates for possible regulation. The UCMR provides occurrence information for determining
human exposure, establishing the baseline for health effects and economic analyses, contaminant
co-occurrence analyses, and treatment technology evaluation (related to the CCL contaminants).
After identifying the top priorities for regulatory determination, EPA begins the process of
determining RfDs and associated enforceable standards for protection of public health.
3. Use of Modeling Results
Summary of Comments: One commenter recommended that EPA compare the results of
hydraulic fracturing after the process to "modeling" conducted before the process to "provide
some degree of predictability of the impact of the fracturing before the actual work is done."
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 7 FINAL
image:
This commenter also recommended that any modeling should consider the effect of other
existing activities and conditions that could affect the outcome of the model (e.g., existing oil
and gas wells, water wells, location and type of surface structures). This commenter also stated
that consideration of the impact of these "man induced activities and conditions" should be an
integral part of any fracture program and of any analysis of CBM fracturing impact. This
commenter stated that the fracturing process and fluids alone may not cause "harm" within the
study's parameters, but when coupled with the existing "man induced conditions" could cause
"considerable damage and risk."
EPA Response: As discussed in Chapter 3 of the report, operators use a number of techniques to
estimate fracture dimensions to design fracture stimulation treatments. Operators have a
financial incentive to keep the hydraulically induced fracture generally within the target coal
zone, so that expenditures on hydraulic horsepower, fracturing fluids, and proppants are
minimized. For precise and statistically reliable measurements, however, fracture height and
length can be measured (as opposed to modeled) accurately by microseismic monitoring.
Tiltmeter measurements can also provide fracture height and length measurements somewhat
accurately. The results of hydraulic fracturing "after the process" have also been investigated in
the mined-through studies by the U.S. Bureau of Mines and others. These studies provide
important, directly-measured characteristics of hydraulic fracturing in coal seams and
surrounding strata. In addition, paint tracer studies conducted as part of mined-through studies
can provide lower bound estimates on the extent of fluid movement.
During its analysis of the threat of CBM fracturing practices on USDWs, EPA considered the
impact of human activities (such as improperly sealed or abandoned wells). Chapter 6 of the
report summarizes citizen complaints and resulting investigations by state agencies into possible
impacts of hydraulic fracturing on drinking water wells and surface waters. In some cases,
improperly sealed gas wells have been remediated, resulting in decreased concentrations of
methane in drinking water wells.
B. Literature Used for the Study
Summary of Comments: Some commenters indicated that the literature used for the study was
outdated. Another commenter questioned whether the search terms that the Agency used to find
references for the report would locate "health-related" literature. This commenter also
questioned whether the acronym "USDW" and/or "underground sources of drinking water" was
used as a search term. Another commenter stated that the report was "simply a compilation of
existing data, with no new information, references, or conclusions."
EPA Response: The search terms used by the Agency did not include health-related terms
because the study's goals did not include conducting a human-health risk assessment or
conducting a new investigation into the toxicity of any of the components of hydraulic fracturing
fluids.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 8 FINAL
image:
As stated in the study design (66 FR 39396)2, EPA focused the study on a review of existing
data. EPA's literature search included publications and documents that were publically available
as of December 2000/January 2001. EPA reviewed over 200 peer-reviewed publications. Much
of the appropriate literature comes from the mid-1990s when funding was available for this kind
of research. EPA also reviewed additional studies recommended by commenters and the peer
review panelists, and incorporated information from these documents into the study, when
appropriate. Further, EPA obtained information for the study through interviews with
approximately 50 employees from industry and state or local government agencies, and
communication with approximately 40 citizens and groups who are concerned that CBM
production affected their drinking water wells.
C. Basins Included in the Study
Summary of Comments: One commenter questioned why EPA's report only included 11 basins.
This commenter indicated that there are 16 separate basins considered to have CBM resources in
the lower 48 states. Further, the commenter stated that the Illinois Basin, which was not
discussed in the study, is a major coal-bearing region in the central Midwest.
EPA Response: EPA's literature search did not find any CBM activity or hydraulic fracturing in
the Illinois Basin. Other basins which have little or no current CBM production activity (e.g.,
Alaska) were also omitted from the study.
D. Citizen Complaints/Instances of Water Well Contamination
Summary of Comments: Many commenters stated that EPA and state agencies have not done an
adequate job of investigating citizen complaints related to contamination of water wells near
hydraulically fractured CBM wells. Some commenters also stated that the Agency disregarded
these complaints by concluding in its draft report that hydraulic fracturing of CBM wells poses a
low risk. Some commenters also believed that the volume of complaints was enough to warrant
the need for the Agency to continue its study. One commenter criticized the Agency for only
having a 30-day collection period associated with the July 30, 2001 Federal Register notice in
which the Agency requested information on groundwater contamination incidents that could be
due to hydraulic fracturing of CBM wells. This commenter added that EPA's outreach efforts
were unlikely to have reached the general public, and also recommended that EPA set up
hotlines and make resources available to "allow immediate, comprehensive investigations of
citizen complaints related to hydraulic fracturing impacts on USDWs."
Conversely, others commenters indicated that based on the volume of hydraulic fracturing
activities, that if the threat to public health from hydraulic fracturing of CBM wells were
significant, confirmed instances of water well contamination would exist. Some of these
US Environmental Protection Agency. 2001. Underground Injection Control; Request for Information of
Ground Water Contamination Incidents Believed To Be Due to Hydraulic Fracturing of Coalbed Methane
Wells. Federal Register. Vol. 66, No. 146. p. 39396, July 30, 2001.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 9 FINAL
image:
commenters indicated that EPA's report should acknowledge the 1998 study conducted by the
Ground Water Protection Council (GWPC), "Survey Results On Inventory and Extent of
Hydraulic Fracturing in Coalbed Methane Wells in the Producing States," GWPC (December 15,
1998) because this survey of state oil and gas regulators provides further support for EPA's study
conclusions.
EPA Response: The response of state agencies and EPA to citizen complaints are documented in
Chapter 6. EPA has responded to complaints, particularly at the Regional level. For instance, in
the Powder River Basin, located in Wyoming and Montana, citizen complaints dealt primarily
with water quantity issues, which were beyond the scope of this study. EPA Region 8 is
participating in a study that addresses the environmental effects of all aspects of CBM
development and not just hydraulic fracturing. In response to citizen complaints, the Alabama
Department of Environmental Management and EPA Region 4 also conducted independent
sampling on wells in the Black Warrior Basin. Water analyses indicated that the wells had not
been contaminated as a result of the hydraulic fracturing activities.
In some regions responses to citizen complaints are made primarily at the state level. For
example, the Colorado Department of Health and the Colorado Oil and Gas Conservation
Commission (COGCC) responds to many complaints. In Colorado, the primary response of the
COGCC to citizen complaints has been the remediation of old, improperly sealed gas wells. The
remediation of such wells has reduced methane concentrations in approximately 27 percent of
the water wells sampled. Reduction of methane concentrations in many of the additional wells is
expected over time due to the COGCC's efforts.
Regarding public outreach efforts need improvement, EPA has made considerable efforts to
ensure its outreach and communications reach the general public. In addition to making the
August 2002 draft available for public comments, EPA's outreach steps included:
• Publishing Federal Register notices (EPA's primary mechanism for
communicating with the public):
requesting comment on how an EPA study should be structured (65 FR
4S774)3;
requesting information on any impacts to groundwater believed to be
associated with hydraulic fracturing (66 FR 39396) (see footnote 2)
including a mailing to over 200 county agencies making them aware of the
Federal Register notice; and
requesting comments on the August 2002 draft of the study (67 FR 55249)
(see footnote 1).
• Holding a public meeting on August 24, 2000, to obtain additional stakeholder
input on the study. Several of these commenters recommended that EPA's study
include accounts of personal experiences with regard to CBM impacts on
drinking water wells. These experiences are discussed in Chapter 6.
3 US Environmental Protection Agency. 2000. Underground Injection Control (UIC) Program; Proposed
Coal Bed Methane (CBM) Study Design. Federal Register. Vol. 65, No. 143. p. 45774, July 25, 2000.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 10 FINAL
image:
• Providing periodic updates for stakeholders, including citizens groups, in the form
of written communication; and
Maintaining a Web site where stakeholders can view the project documents; get
updates on the progress of the project (including announcements of the release of
Federal Register notices); and provide information to EPA.
Regarding the comment that EPA only provided 30 days for the public to provide information on
CBM-related groundwater contamination incidents following the July 30, 2001 Federal Register
notice, note that the Agency has considered all complaints received from the public, regardless
of the time at which EPA received them. In addition, EPA's Web site
www.epa.gov/safewater/uic/cbmstudy.html has a link to a form that allows people to submit
information on the potential effects of hydraulic fracturing.
In response to the commenter's suggestion regarding hotlines, EPA has its Safe Drinking Water
Hotline, which callers within the United States may reach at (800) 426-4791. Citizens are
welcome to contact EPA or the states regarding these issues.
Regarding the comment about the volume of CBM activities and lack of confirmed instances of
water well contamination, during its review, EPA found no confirmed cases that are linked to
fracturing fluid injection into CBM wells or subsequent underground movement of fracturing
fluids. Although thousands of CBM wells are fractured annually, EPA did not find confirmed
evidence that drinking water wells have been contaminated by hydraulic fracturing fluid
injection into CBM wells. EPA has included language to that effect in its final report,
"Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs", June 2004, EPA document number: EPA 816-R-04-003
(hereafter referred to as final report).
E. Peer Review Panel
Summary of Comments: Many commenters questioned the composition of EPA's peer review
panel, who reviewed the initial draft report. These commenters stated that this panel was heavily
biased toward industry that has a stake in the outcome of the study. These commenters
recommended that EPA convene a panel that is free of conflict of interest. Some recommended
using members of the NAS as panelists.
One commenter indicated that he could not ascertain the composition of the panel although
Appendix B of the report is supposed to contain a table with the list of the peer review panel.
Another commenter stated that EPA made it very difficult for the public to obtain a copy of the
peer review report, and that these comments were not attached in an appendix as originally
promised.
EPA Response: EPA has a formal Agency Peer Review Policy that establishes the criteria and
requirements for independent evaluation of scientific and technical studies and documents.
Consistent with that policy, the Agency established a seven-member technical expert peer review
panel, who performed a technical review of the study. Panel members were selected by
identifying individuals with scientific or technical expertise in hydraulic fracturing through
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 11 FINAL
image:
reviewing peer-reviewed publications in scientific journals and through communications with
professional societies, trade and business associations, state organizations, and other federal
agencies. EPA considered over 20 candidates before selecting 7 individuals based on their
experience in the fields of hydraulic fracturing, rock mechanics, and/or natural gas production,
and for their varying perspectives (industry, state government, academia, and a national
laboratory). The charge to this committee was to review the report to determine if: 1) the report
is complete, thorough, and accurate; and 2) the scientific/technical studies reviewed are applied
in a sound, unbiased manner.
EPA posted the list of these reviewers and their qualifications on its Web site at
www.epa.gov/safewater/uic/cbmstudy.html. EPA inadvertently omitted the table that identifies
the peer reviewers in Appendix B of the draft report. This table is included in the final report.
III. FRACTURE FLUIDS
A. Components of Fracturing Fluids
1. Health Effects
Summary of Comments: Many commenters were concerned about the amount and health effects
of certain chemicals used in hydraulic fracturing fluids and cited these concerns as reasons to
continue the study. Some argued that very small quantities of toxic chemicals, such as benzene
or methyl tert butyl ether (MTBE), could contaminate millions of gallons of groundwater.
Other commenters were concerned about the way in which the constituents of fracturing fluids
and their potential health effects were presented in the draft report. For example, one commenter
wanted the report to clearly convey the following: a wide variety of fracturing fluids exist, the
health effects identified in the report apply to only some of the constituents that may or may not
be present in the fracturing fluid, the health effects are associated with the product in its "pure
form," and all the fluids additives are greatly diluted during fracturing operations.
EPA Response: As discussed in section II. A.2, EPA has recently entered into agreements with
three major service companies to voluntarily eliminate diesel fuel from hydraulic fracturing
fluids injected directly into USDWs for CBM production. Compounds such as benzene are
components of diesel. These agreements will significantly reduce the use of diesel fuel in
hydraulic fracturing fluids that are injected directly into USDWs for CBM production.
Chapter 4 of the final report provides a general description of the fate and transport processes
which would minimize potential exposure to chemicals used in hydraulic fracturing fluids.
Based on a 1991 fracturing fluid recovery study conducted in coal by Palmer et al., as much as
68 to 82 percent of the fracturing fluids may be removed when the methane is extracted.4 This
study is discussed in Chapter 3 of the report. As detailed in Chapter 4 of the report, the
4 Palmer, ID., Fryar, R.T., Tumino, K.A., and Puri, R. 1991. Comparison between gel-fracture and
water-fracture stimulations in the Black Warrior basin; Proceedings 1991 Coalbed Methane Symposium, University
of Alabama (Tuscaloosa), pp. 233-242).
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 12 FINAL
image:
unrecovered fluids will undergo processes that may limit their availability, concentration, and
movement. These fluids may be significantly diluted and dispersed as they are transported
through the subsurface. They may also interact chemically or physically with geologic material
which may retard their movement and further disperse their concentrations.
EPA identified fluids and fluid additives commonly used in hydraulic fracturing through
literature searches, reviews of relevant material safety data sheets (MSDSs) provided by service
companies, and discussions with field engineers, service company chemists, and state and federal
employees. The draft and final reports provide a discussion of the wide variety of hydraulic
fracturing fluids that may be used. Table 4-1 of the report lists components that may be
contained in fracturing fluids based on MSDSs. The final report emphasizes that not all
fracturing fluid constituents, identified in Table 4-1 of this report, may be present in fracturing
fluids, that the potential human health effects presented in the table apply to these compounds in
their pure form, and that these compounds are significantly diluted prior to use.
An environmental impact statement (EIS) prepared by the BLM also identified MTBE as a
compound that may be found in fracturing fluid (U.S. Department of the Interior, CO State
BLM, 1998).5 However, EPA was unable to find any indications in the literature, on MSDSs, or
in interviews with service companies that MTBE is used in fracturing fluids to stimulate CBM
wells.
2. Diesel Fuel
Summary of Comments: Several commenters supported EPA's recommendation that the
industry use "water-based" alternatives in lieu of hazardous constituents such as diesel fuel.
Some argued that EPA should make this a requirement and not a recommendation. Some of
these commenters pointed to EPA's recommendation to "remove any threat whatsoever" from
hydraulic fracturing fluid as a contradiction to the study's conclusions and as a reason to continue
the study.
Conversely, several commenters indicated that there are valid reasons for using certain
chemicals to enhance CBM production and that in choosing alternatives, the CBM well operators
must take into account the specific geologic conditions of the site. These commenters
recommended that EPA "encourage flexibility" with respect to the production of methane. One
of these commenters noted that the draft report suggests that water-based alternatives are:
currently available, feasible, and acceptable substitutes for diesel-based gels. This commenter
indicated that the report findings should recognize that more research is needed on these
potential alternatives. This commenter added that not all of the potential alternatives to the use
of diesel may be water-based, citing polymer-based alternatives as one possibility. This
commenter recommended that the term "water-based alternatives" be changed to read "non-
diesel-based alternatives."
5 U.S. Department of the Interior, Bureau of Land Management, Colorado State Office. 1998. Glenwood
Spring Resource Area: Oil & Gas Leasing Development, Draft Supplemental Environmental Impact Statement, June
1998.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 13 FINAL
image:
One commenter indicated that in the State of Alabama, diesel is not used nor is it approved for
hydraulic fracturing. The commenter added that service companies in his state primarily use a
linear gel composed of guar gum, a surfactant, and silica.
EPA Response: The discussion of potential alternatives to the use of diesel is not included in the
final report because it is outside the scope of the study. Instead, the report highlights the MOA
with three major service companies to voluntarily eliminate the use of diesel fuel in hydraulic
fracturing fluids injected directly into USDWs for the purpose of CBM production and if
necessary, select replacements that will not cause hydraulic fracturing fluids to endanger
USDWs (see the response to comment in section II. A.2).
Regarding the comment on the use of diesel in the State of Alabama, Table A2-1 in Attachment
2 of the draft and final report indicates that diesel is not used in that state.
3. MTBE
Summary of Comments: Several commenters were concerned about the use of MTBE in
fracturing fluids. Many of them included the following statement in their comments: "only 28
tablespoons of MTBE could contaminate millions of gallons of groundwater."
One commenter indicated that the report contains several inconsistent statements regarding
MTBE as a component of fracturing fluids. This commenter noted that in Chapter 4 of the draft
report, EPA states that, based on its literature reviews and interviews with service companies, the
Agency did not find any evidence that MTBE is used in fracturing fluids. This commenter also
indicated that later in the same chapter, EPA states that "some gelling agents can contain
hazardous substances including . . . [MTBE.]," and cites as its source a Supplemental EIS
issued by BLM. This commenter provided arguments why he believed that the supplemental
EIS was in error in listing MTBE as a potential component in fracturing fluids. This commenter
further recommended that EPA should not have used this EIS as a source for identifying
constituents in fracturing fluids or at a minimum, should have indicated the shortcomings
associated with using this type of document to determine the components of fracturing fluids.
This commenter provided a detailed discussion of some of the problems with using this
particular EIS.
EPA Response: As stated in the response to comment in section III. A. 1, an EIS prepared by the
Colorado State BLM (1998) identified MTBE as a compound that may be found in fracturing
fluid. EPA found no information in the literature, MSDSs, or through interviews with service
companies indicating that MTBE is used in fracturing fluids to stimulate CBM wells. MTBE is
not used during the manufacture of diesel fuel. It is generally only added to gasoline. However,
in an effort to be fully inclusive of all the Agency's literature search findings, EPA included the
information found in the EIS and noted that EPA was not able to confirm MTBE use in
fracturing fluids.
B. Comparison of Concentrations of Hydraulic Fracturing Fluid Components to MCLs
Summary of Comments: A few commenters questioned the appropriateness of EPA's use of
MCLs to compare the projected concentrations of fracturing fluids that may be injected into
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 14 FINAL
image:
USDWs. The commenters argued that MCLs apply to "treated water" and that the water
associated with the formations in which hydraulic fracturing occurs would not be suitable for
drinking water without first being treated.
EPA Response: Under the mandate of SDWA, EPA establishes MCLs as enforceable maximum
permissible levels for contaminants in drinking water, to ensure the safety of public drinking
water supplies. Because the concern about contamination relates to USDWs, which are actual or
future supplies of drinking water for human consumption, MCLs are used in this study as
standard reference points to compare calculated or anticipated levels of contaminants in
hydraulic fracturing fluids and in the subsurface. MCLs provide a context for discussions
regarding the concentrations of individual contaminants.
C. Concentrations of Constituents in Fracturing Fluids/Fluid Recovery Rates
1. Estimates of Concentrations of Constituents in Fracturing Fluids
Summary of Comments: EPA received several comments on its estimates of the concentrations
of the constituents of concern in fracturing fluids that may be present at the point-of-injection
and at the edge of the fracture zone. Many commenters were alarmed about the estimated
concentrations of some of these constituents such as benzene because they were above the MCL.
Further, some were concerned that EPA had revised its estimates since publication of the report.
Conversely, other commenters indicated that EPA had overstated these concentrations. Each of
these comments is discussed in more detail below.
One commenter indicated that EPA's estimates for the constituents of concern at the edge of the
fracture zone, which assume a dilution factor of 30, still exceed drinking water standards for
benzene, aromatics, 1-methylnapthalene, and methanol. This commenter added that EPA
estimated high concentrations for the estimated point-of-injection for some chemicals for which
drinking water standards have not yet been developed. This commenter acknowledged that these
concentrations will be reduced as they mix with groundwater; however, he stated that very small
amounts of some chemicals like benzene and MTBE can contaminate millions of liters of
groundwater. Further, this commenter noted that most CBM wells are hydraulically fractured
more than once, and therefore, "the groundwater in which it resides," will receive multiple doses
of the fracturing fluids chemicals. The commenter stated a figure from the report that between
50,000 and 350,000 gallons of fracturing fluids are typically used in coalbed fracture treatments.
Another commenter indicated that the report does not recognize that some of the constituents in
fracturing fluids may affect human health at very low concentrations. This commenter added
that with the potentially thousands of CBM wells being developed, the problem is magnified.
Several commenters claimed that EPA revised its calculations after the draft report was released.
Some of these commenters indicated that EPA changed its scientific and policy conclusions
under pressure from industry. One commenter provided detailed comments on the revised
calculations. This commenter argued that EPA changed some of the parameters that were used
in the draft report (such as length and height of a fracture, volume of injected hydraulic
fracturing fluids, percentage of unrecovered hydraulic fracturing fluids) and they resulted in
smaller estimated concentrations, including a revised estimate for benzene that does not exceed
the MCL. This commenter questioned the basis for EPA's revising its estimates.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 15 FINAL
image:
Other commenters were concerned that EPA did not adequately explain the assumptions used to
generate its calculations. For example, one commenter indicated that it was unclear whether
EPA based its estimates at the edge of the fracture zone on a specific fracture length or fracture
radius. Some commenters also stated that EPA did not consider factors that would influence the
availability and decrease the concentrations of the constituents at the edge of the fracture zone.
These factors included: the recovery of the majority of the fracturing fluid, the relatively low
permeability of coalbed formations will limit the movement of groundwater away from the
wellbore, the coal will adsorb some of the constituents onto its surfaces, acids react with certain
rock constituents and become spent, and some fracturing fluid constituents such as benzene will
biodegrade. Some commenters also recommended that EPA's report should further emphasize
that any constituents of concern in fracturing fluids are present only in very minimal amounts.
One commenter indicated that EPA had "significantly mischaracterized the nature of its
estimates at both the point-of-injection and the edge of the fracture zone" because EPA had used
a "worst case" scenario for estimating these concentrations. The commenter stated that, although
the report indicates that EPA used mid-range values, the Agency used the maximum amount of
diesel fuel that service companies reported to EPA instead of an average value. This commenter
also explained why he believed that some of the point-of-injection concentrations that were
presented in Table 4-2 of the draft report, such as that estimated for methanol, appeared to be
inconsistent with the discussion in the text. Further, this commenter also recommended that
EPA include its newer calculations in the report.
EPA Response: The values presented in the draft report are oversimplified estimates based on
dilution alone and are not accurate enough to predict that a 30 times decrease is above or below
the MCL. In the final report, EPA has revised its procedure for assessing the potential effect of
fracturing fluid constituents on USDWs from that presented in the August 2002 draft as follows:
The draft report included point-of-injection calculations for all constituents that may
be contained in fracturing fluids. The final report focuses only on those constituents
for which MCLs are established (i.e., BTEX compounds).
• EPA has revised the fraction of BTEX compounds in diesel used to estimate the
point-of-injection concentrations from a single value to a documented broader range
of values for the fraction of BTEX in diesel fuel. For example, the fraction of
benzene in diesel was revised from 0.00006gbenzene/gdiesel to a range with a minimum
value of 0.000026 gbenzene/gdiesel and a maximum value of 0.001 gbenzene/gdiesel. If the
maximum value for benzene in diesel is used to estimate the concentration of benzene
at the point-of-injection, the resulting estimate is 17 times higher than that presented
in the draft report.
• In the final report, EPA used more current values for two of the parameters used to
estimate the point-of-injection concentrations of BTEX compounds. Specifically, the
estimates in this report use a density of the diesel fuel-gel mixture of 0.87 g/mL
compared to 0.84 g/mL in the draft report, and a fraction of diesel fuel in gel of 0.60
gdiese/ggei compared to 0.52 g^^ggd in the draft report. The use of these more current
values does not affect the order of magnitude of the revised point-of-injection
calculations.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 16 FINAL
image:
The August 2002 draft report included estimates of the concentration of benzene at an
idealized, hypothetical edge of the fracture zone located 100 feet from the point-of-
injection. Based on new information and stakeholder input, EPA concluded that the
edge of fracture zone calculation is not an appropriate model for reasons including:
- Mined-through studies reviewed by EPA indicated that hydraulic fracturing
injection fluids had traveled several hundred feet beyond the point-of-
injection.
The assumption of well-mixed concentrations within the idealized fracture
zone is insufficient. One mined-through study indicated an observed
concentration of gel in a fracture that was 15 times the injected concentration,
with gel found to be hanging in stringy clumps in many fractures. The
variability in gel distribution in hydraulic fractures indicates that the gel
constituents are unlikely to be well mixed in groundwater.
Based on more extensive review of the literature, the width of a typical
fracture was estimated to be much thinner than that used in the draft report
(0.1 inch versus 2 inches). The impact of the reduced width of a typical
fracture is that the calculated volume of fluid that can fit within a fracture is
less. After an initial volume calculation using the new width, EPA found that
the volume of the space within the fracture area may not hold the volume of
fluid pumped into the ground during a typical fracturing event. Therefore,
EPA assumes that a greater volume of fracturing fluid must "leakoff' to
intersecting smaller fractures than what was assumed in the draft report, or
that fluid may move beyond the idealized, hypothetical "edge of fracture
zone." This assumption is supported by field observations in mined-through
studies, which indicate that fracturing fluids often take a stair-step transport
path through the natural fracture system.
• In the draft report, EPA approximated the edge of fracture zone concentrations
considering only dilution. Based on new information and stakeholder input on
the draft report, EPA does not provide estimates of concentrations beyond the
point-of-injection in the final report. Developing such concentration values with
the precision required to compare them to MCLs would require the collection of
significant amounts of site-specific data. This data in turn would be used to
perform a formal risk assessment, considering numerous fate and transport
scenarios. These activities are beyond the scope of Phase I of this study.
• In Chapter 4 of the final report, EPA provides a qualitative evaluation of the fate
and transport of unrecovered fracturing fluids on residual concentrations of BTEX
in groundwater. EPA describes in Chapter 4 how subsurface flow would
significantly disperse and dilute BTEX compounds in groundwater, minimizing
potential exposure to these constituents. BTEX compounds may also interact
chemically or physically with geologic material which may retard their movement
and further disperse their concentrations.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 17 FINAL
image:
See also EPA's response to comment in section III. A.I of this document.
No data or conclusions in the final report or in any previous draft were altered to accommodate
any industry parties, states, environmental groups, or others. This study was a thorough and
transparent data collection and technical evaluation exercise. The report and its conclusions
were prepared by career technical staff at EPA.
The study was designed based upon a transparent process including public comment on the
conceptual study design which included comments from state drinking water and oil and gas
agencies, industry, environmental groups, and private citizens. EPA consulted with experts in
the United States Geological Survey and the Department of Energy. Consistent with principles
of good science, a draft of the study was subjected to a technical peer review from hydraulic
fracturing experts. The conclusions of the study were not submitted for review to any private
sector parties.
2. Fluid Recovery Rates
Summary of Comments: Many commenters were concerned that a large percentage of fracturing
fluid remains behind and is available to potentially migrate into USDWs, citing these concerns
as a reason to continue EPA's study. Some commenters indicated that EPA was inconsistent in
the recovery percentages that the Agency cited in the report. Two commenters noted that the
recovery experiment that is referenced in the report only ran for 19 days and that additional
fracturing fluids may be recovered after that time. Another commenter stated that one fluid
recovery rate (i.e., 61 percent) should not be "indiscriminately applied to over 14,000 CBM
wells."
Some commenters cited a study by three Amoco scientists in which the study found "that a
significant volume of fracturing fluids is not withdrawn." These commenters explained that the
scientists found that the gelling agents used in the fracturing fluids remained in the coal samples
although they had been flushed with water and strong acids. The commenters argued that, since
these chemicals are not fully recovered, they could "serve as continuous sources of groundwater
contamination."
EPA Response: Section III. A. 1 provides a discussion of processes that can limit the availability,
concentration, and movement through groundwater of unrecovered fracturing fluids. EPA has
ensured that the recovery percentages cited in the report are both internally consistent and
consistent with the literature reviewed. Three studies on recovery rates of hydraulic fracturing
fluids were reviewed in Chapter 3 of the report. Only one of these studies, Palmer et al., 1991,
involved hydraulic fracturing of coalbeds (refer to footnote 1 for the study reference). Thus, the
Palmer study was considered the most relevant of the three studies for the purposes of this
report. The final report clarifies that the recovery rate of 61 percent was based on a 19-day
flowback period. Palmer et al., 1991, predicted recovery rates as high as 82 percent over a
longer recovery period.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 18 FINAL
image:
Regarding the study by three Amoco scientists, EPA contacted one of the commenters to obtain
a copy of the study to review.6 The commenter was unable to provide the study and EPA's
additional library research efforts were also unsuccessful at obtaining this study.
3. Amount of Fracturing Fluids
Summary of Comments: Some commenters were concerned about the volume of fracturing
fluids used in a "typical fracturing job" and cited the following statement from the report,
"Coalbed fracture treatments typically use 50,000 to 350,000 gallons of various fracturing fluids,
and from 75,000 to 320,000 pounds of sand as proppant... ." Others questioned the accuracy of
the quantities of fracturing fluid and proppant cited in the report, stating that these figures were
more consistent with a massive hydraulic fracture. Another commenter stated that the unique
properties that make many coal formations effective receptacles for methane also allow them to
hold large quantities of water. This commenter stated that injection of hydraulic fracturing fluids
into USDWs risks permanent contamination of these USDWs because fracturing fluids often
contain large amounts of toxic chemicals.
EPA Response: EPA has clarified in the final report that more typical injection volume may be
closer to a maximum of 150,000 gal/well, and a median value of 57,500 gal/well. These values
are based on average injection volume data provided by Halliburton for six CBM locations.
Refer to section III.A. 1 regarding factors that would influence the availability, concentration, and
movement of fracturing fluids and their constituents.
4. Movement of Fracturing Fluids
Summary of Comments: Some commenters stated that unrecovered fracturing fluids will flow
toward the well because of the pressure gradients. Others noted that this was only true while the
well was in production. These commenters argued that once pumping stops, the aquifer will
attempt to resume a normal flow pattern and the remaining hydraulic fracturing fluids will move
freely within the coalbed formation.
EPA Response: Chapter 4 of the final report has been expanded to more clearly explain:
• hydraulic gradients that occur during injection versus those during fluid recovery;
• the significance of the capture zone of the production well on fracturing fluid
recovery (i.e., the portion of the aquifer that contributes water to the well); and
• the movement of fracturing fluids (and what influences their movement) both inside
and outside the capture zone.
6 Puri, R., G.E. King, and ID. Palmer, 1991, "Damage to Coal Permeability During Hydraulic Fracturing,"
Society of Petroleum Engineers Proceedings from Rocky Mountain Regional Meeting and Low-Permeability
Reservoirs Symposium, Denver, CO, p. 109-115, (SPE #21813).
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 19 FINAL
image:
IV. FRACTURE BEHAVIOR AND PRACTICES
A. Fracture Growth
Summary of Comments: EPA received many comments on the statements in its report that,
"Vertical fracture heights in coalbeds have been measured in excess of 500 feet and lengths can
reportedly reach up to 1,500 feet." Some of these comm enters stated that these distances
indicate the potential for communication with and contamination of USDWs. Other commenters
believed that these measurements were incorrect. Some commenters also discussed whether
confining layers act as barriers to vertical fracture growth.
One commenter described in detail why he believed that confining layers above and below the
hydraulically fractured coal formations would also be fractured and permeated by fluids. This
commenter noted that the fracture heights cited in the report exceed the thickness of the thickest
coal formations identified in the report. In addition, this commenter noted that the report
indicates that some of the coal seams are bounded by sandstone and conglomerate (which have
different lithological properties, and therefore different fracturing properties, than shale).
Further, he indicated that the report supports his position that the risk for migration of fracturing
fluids into adjacent USDWs is significant because it indicates that "Stimulation fluids in coal
penetrate from 50 to 100 feet away from the fracture and into the surrounding formation. In
these and other cases, when stimulation ceases and production resumes, these chemicals may not
be completely recovered and pumped back to the CBM well, and, if mobile, may be available to
migrate through an aquifer." This commenter also noted that the report shows that many of the
coal formations are located in mountainous regions such as the Rocky Mountains and
Appalachian Mountains. The commenter stated that the rock formations in these regions,
including the coal formations, have been subjected to intense orogenic and tectonic stress
resulting in regional, systematic fractures and faults. The commenter argued that it is likely that
coal formations, and other rocks above and below them, are characterized by cracks and
fractures, and that because of these deformation features, rates of groundwater transport tend to
be higher.
One commenter indicated that the report's description of how fractures travel is incorrect (i.e.,
they travel horizontally vs. vertically). This commenter added that there is some vertical
expansion as the fracture moves horizontally but that this is not the primary direction of
fracturing. This commenter stated that their state geologists estimate vertical fracture heights at
50 to 60 feet. Another commenter provided detailed comments on the studies that were
conducted on fracture height growths. This commenter indicated that he had been involved in
numerous fracture experiments (in all types of reservoirs) where the fracture height has actually
been measured (using microseismic or downhole tiltmeter), as well as in mineback tests where
hydraulic fractures have been excavated. Based on his experience, the fracture height has always
been less than or equal to the height that would be predicted by just using stresses in the various
layers (which the commenter indicated was the only factor considered in all the references used
in the draft report). The commenter reported that in some cases, the differences were factors of
two or three. This commenter also provided detail on factors that influence fracture height
growth, such as horizontal stress in the coal, the horizontal stress in the surrounding layers, the
characteristics of the layering, and the type of hydraulic fracturing fluids being pumped.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 20 FINAL
image:
Another commenter noted that the discussion on fracture dimensions in the report was based on
literature from 1993 and earlier, but acknowledged that there were "virtually no post-1993
published reports on hydraulic fracturing." The commenter recommended that EPA contact
operators, service companies, and state regulatory agencies for current practices and models.
Further, this commenter noted that newer data based on more sophisticated FracPro models are
available for many basins. He added that, in his state, model results indicate that fracture height
is "generally less than 100 feet, whereas fracture half length is typically between 150 and 700
feet." This commenter also noted that the report should state that the fracture heights have been
"modeled" not "measured" because vertical fracture heights have never been fully measured in
the field.
EPA Response: EPA has revised Chapter 3 to provide clarification on the characterization of
fracturing behavior during hydraulic stimulations. The statement that fractures have been
"measured in excess of 500 feet and lengths can reach up to 1,500 feet" has been removed
because it refers to modeled estimates, rather than direct measurements. Instead, the results of
22 mined-through studies have been summarized, because they provide direct measurements of
the dimensions of hydraulic fractures, as well as lower bounds on the extent of fracturing fluid
movement. Chapter 3 has also been revised to better distinguish between fracture
characterizations based on modeling vs. those that are directly measured.
In addition, EPA has revised Chapter 3 to clarify the issue of hydraulic barriers and barriers to
fracture growth above coalbeds. EPA agrees with the commenter that when shales overlying
targeted coals are extensively fractured, they may not act as barriers to hydraulic fracture growth
or as hydraulic barriers. On the other hand, thick, relatively unfractured shale may present a
barrier to upward fracture growth because of the stress contrast between the coalbed and the
overlying shale. Deep vertical fractures can propagate vertically to shallower depths and
develop a horizontal component. In the formation of these "T-fractures," the fracture tip may fill
with coal fines or intercept a zone of stress contrast, causing the fracture to turn and develop
horizontally, sometimes at the contact of the coalbed and an overlying formation.
B. Multiple Fractures
Summary of Comments: Some commenters raised concern over the statement in the draft report
that "each well, over its lifetime is fractured several times" and urged EPA to continue to Phase
II of the study. Others questioned the accuracy of EPA's statement that wells are fractured
multiple times. One commenter indicated that in their state, most wells have not been re-
fractured multiple times but that instead, two to four coal groups were generally fractured in each
well.
EPA Response: EPA has revised the statements regarding multiple stimulations in Chapter 3. In
the draft report EPA stated that "many coalbeds are refractured at sometime after the initial
treatment." The text has been revised to indicate that the literature on refracturing that was
reviewed pertains only to the Black Warrior Basin. EPA's extensive literature review did not
find any information indicating that wells are fractured multiple times in any basin other than the
Black Warrior Basin.
C. Relationship of Drinking Water Wells to Hydraulic Fracturing Activities
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 21 FINAL
image:
Summary of Comments: Some commenters were concerned about the potential for fracturing
fluids to contaminate USDWs due to the high occurrence of coal reservoirs within USDWs. One
commenter cited a statement from the report "if coalbeds are located within USDWs, then any
fracturing fluids injected into coalbeds have the potential to contaminate the USDW." The
commenter added that the report indicates that as much as 91 percent of U.S. coal reservoirs may
be located within USDWs.
Two commenters indicated that hydraulic fracturing activities take place at depths far below
groundwater sources used as drinking water sources. One of these commenters added that his
company's records show that it conducts hydraulic fracturing at shallow depths, (i.e., less than
300 feet below ground surface), in less than one percent of all hydraulic fracturing jobs. This
commenter provided this as one reason that he believed that hydraulic fracturing is unlikely to
pose a threat to drinking water.
EPA Response: EPA found that 10 of the 11 coal basins, included in the study, may lie, at least
in part, within USDWs. Given the concerns associated with the use of diesel fuel and the
introduction of BTEX constituents into USDWs, EPA negotiated an MOA with three major
hydraulic fracturing service companies for the voluntary elimination of diesel fuel in hydraulic
fracturing fluids injected directly into USDWs for the purpose of CBM production.
Nevertheless, even when fracturing fluids are injected directly into coalbeds located in USDWs,
fracturing fluid components are likely to be significantly diluted and dispersed, as well as subject
to other fate and transport processes (discussed in Chapter 4) which are likely to lower their
concentrations or prevent their mobility underground. Also see the response to comment in
section III.A.I.
D. Differences in State Geology
Summary of Comments: Several commenters indicated that the report did not adequately
address the variability present in the different geologic formations that are subject to hydraulic
fracturing, and therefore, did not address the possible impacts associated with that variability
regarding regional groundwater flow and/or the occurrence and distribution of CBM resources,
on assessing the potential threat of hydraulic fracturing on USDWs. One commenter indicated
that to accurately represent the threats to USDWs, risk levels should be "differentiated based on
modeling and actual data on similar geologic conditions."
EPA Response: EPA agrees that variability of geologic formations and regional groundwater
flow are key to the assessment and understanding of the potential threat to USDWs posed by
hydraulic fracturing. The study findings and conclusions are based on literature from each of the
11 major coal basins in the United States. In addition, the draft and final report contains separate
attachments which discuss basin-specific geologic and hydrogeologic investigations related to
each of the 11 basins. The discussions provided were intended to characterize regional coal
basin methane production with respect to its effect on USDWs and to supplement the generalized
information provided within the body of the report. EPA also agrees that if modeling risk levels,
the variability of geologic conditions should be considered. However, such a modeling exercise
is beyond the scope of the current study.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 22 FINAL
image:
V. REGULATION OF HYDRAULIC FRACTURING PRACTICES
A. States' Authority
Summary of Comments: Several commenters recommended that EPA expand its discussion in
the final report of the states' role in regulating hydraulic fracturing. Others suggested clarifying
the language from the draft report regarding states' authority to regulate hydraulic fracturing.
For example, one commenter indicated that EPA's statement, "States with primacy for their UIC
program enforce and have the authority to place controls on any injection activities that may
threaten USDW's" implies that state UIC Programs can or would regulate hydraulic fracturing.
The commenter recommended that EPA add clarifying language that removes the implication
that hydraulic fracturing is commonly regulated under UIC Programs.
One commenter stated that the report was inaccurate in its description of Virginia's authority to
place restrictions on the depth at which hydraulic fracturing can occur. The commenter
indicated that the "restrictions" are instead voluntary procedures. The commenter also clarified
the purpose of these procedures.
EPA Response: EPA did not conduct a systematic review of state regulations of hydraulic
fracturing and, therefore, has no basis for expanding its discussion of the state's role in the
regulation of hydraulic fracturing. However, the Agency added clarifying language regarding
the state's ability to regulate hydraulic fracturing. EPA also added clarifying wording to the
report regarding Virginia's voluntary program.
B. Regulation of Hydraulic Fracturing under SDWA
Summary of Comments: Several commenters wanted EPA to regulate hydraulic fracturing of
CBM wells under SDWA and did not believe that recommended measures such as using
"water-based alternatives" instead of diesel were sufficient. One commenter stated that based on
Legal Environmental Assistance Foundation, Inc. v. U.S. E.P.A., 118 F.3d 1467, 1470 (llth Cir.
1997), EPA is to decide how to regulate hydraulic fracturing under SDWA, and not to determine
whether "further investigation was necessary to evaluate any potential threats" before EPA acts.
Another commenter was concerned whether EPA was using the presence of documented cases of
"health harm from non-regulation" as the criterion for determining whether to regulate hydraulic
fracturing injection activities under SDWA. This commenter argued that the purpose of the UIC
Program is "to forestall and prevent such harm by isolating the injected fluids from aquifers that
are or could be developed as USDWs"; and therefore, using proven harm as a regulatory
threshold goes against the purpose and intent of the law.
Conversely, other commenters indicated that EPA should "recognize the need for industry to be
allowed reasonable flexibility in the means that its uses to produce CBM." These commenters
also indicated that under 42 U.S.C. § 300h(b)(2), Congress intended that EPA not impose
restrictions through the UIC Program that interfere with or impede activities related to oil and
gas development unless such restrictions are essential for preventing endangerment of drinking
water sources. Another commenter specifically recommended that UIC permits not be required
for hydraulic fracturing practices.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 23 FINAL
image:
EPA Response: Based on the information collected and reviewed, EPA has determined that the
injection of hydraulic fracturing fluids into CBM wells poses little or no threat to USDWs.
Continued investigation under a Phase II study is not warranted at this time. The lack of
confirmed incidents of drinking water well contamination due to hydraulic fracturing fluid
injection from past hydraulic fracturing activities was one among many factors EPA considered.
If threats to USDWs from hydraulic fracturing of CBM wells were significant, EPA would
expect to have found confirmed instances of drinking water well contamination from the
practice. Although thousands of CBM wells are fractured annually, EPA did not find confirmed
evidence that drinking water wells have been contaminated by the injection of hydraulic
fracturing fluids into CBM wells.
EPA's recent agreements with three major service companies, discussed in section II.A.2, will
significantly reduce the use of diesel fuel in hydraulic fracturing fluids that are injected directly
into USDWs for CBM production.
It is important to note that states with primary enforcement authority (primacy) for their UIC
Programs implement and enforce their regulations, and have the authority under SDWA to place
additional controls on any injection activities that may threaten USDWs. States may also have
additional authorities by which they can regulate hydraulic fracturing. With the expected
increase in CBM production, the Agency is committed to working with states to monitor this
issue.
VI. LANGUAGE USED IN THE REPORT
A. Use of the Term "USDW"
Summary of Comments: Some commenters indicated that EPA used the term "USDW" too
broadly. In particular, one commenter indicated that the report "carelessly utilizes the USDW
term in the context of hydrocarbon bearing formations." This commenter added that these
hydrocarbon-bearing aquifers subjected to hydraulic fracturing are unlikely to be used for
drinking water, especially without treatment for two reasons: 1) the high total dissolved solids
level of the waters in these formations; 2) the waters in these formations may be considered an
"exempted aquifer" under SDWA because the aquifer is mineral, hydrocarbon, or geothermal
energy producing, or can be demonstrated to be commercially producible. This commenter also
stated that the inferences in the report, that some risks may be attributed to hydraulic fracturing,
conflict with "the reality that such a formation would not be used for water supply without
treatment, if it were ever to be used."
EPA Response: EPA disagrees that it has applied the term "USDW" too broadly in the report.
SDWA mandates the protection of USDWs from injection activities - "if such injection may
result in the presence in underground water which supplies or can reasonably be expected to
supply any PWS of any contaminant, and if the presence of such contaminant may result in such
system's not complying with any national primary drinking water regulation or may otherwise
adversely affect the health of persons." The broad definition of a USDW by Congress was to
ensure that future USDWs would be protected, even where those aquifers were not currently
used as a drinking water source or could not be used without some form of water treatment such
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 24 FINAL
image:
as desalination. It is also important to note that an exempted aquifer is a USDW, but is exempt
from regulation.
B. Use of Scientific Terms
Summary of Comments: A few commenters provided corrections to some of the terminology
used in the report. One commenter felt that there was a general misuse of geologic terminology
in the report, and specifically indicated that the geologic terms "system," "formation," and
"seams" should not have been used interchangeably. This commenter provided other specific
clarifications or corrections to some of the discussions in the report (e.g., Section 3.1 regarding
the depositional history of coal-bearing rocks in the United States).
EPA Response: EPA appreciates the careful review of the report by many of the commenters.
EPA has revised some of the terminology used in the report and incorporated some of the
clarifications suggested by the commenters.
C. Use of Qualifying Language
Summary of Comments: Both the commenters that supported EPA's conclusions and those who
opposed it indicated that the tone of the language used throughout the report conflicted with
EPA's conclusions. Commenters cited examples of this language that included the following:
• "Based on the information collected, the ootential threats to USDWs posed by hydraulic
fracturing appear to be low and do not justify additional study.";
• ..."the avvarent risk to oublic health from hvdraulic fracturing is not compelling enough
to warrant expending resources on a phase II effort"; and
• "the apparent threat to public health from hydraulic fracturing."
One of the commenters indicated that this language showed "a weak articulation of EPA's
confidence in its own report." Many of the commenters who were opposed to EPA's findings,
pointed to EPA's qualified statements as a reason to continue the study.
Another commenter, who supported EPA's findings, stated that the primary definition of the
word, "apparent," is, "something that is clearly seen or understood, obvious, self-evident,
glaring." This commenter, among others who supported the Agency's findings, recommended
that EPA replace all uses of the word "apparent" when describing the threat posed to USDWs by
hydraulic fracturing with words that more accurately describe the low likelihood of this threat.
EPA Response: In the final report, EPA has eliminated the use of the word "apparent" and
"appears" to describe its study conclusions and has made the language more consistent with the
report's results.
VII. CHAPTER-SPECIFIC COMMENTS
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 25 FINAL
image:
This summary of chapter-specific comments focuses mainly on those comments that have not
been summarized within the issue-specific Sections II through VI of this document. Comments
were received on almost every chapter of the document, ranging from minor editorial
suggestions, to factual corrections. EPA appreciates the thorough comments that were submitted
regarding the contents of the hydraulic fracturing report. The Agency has considered all
comments, researched the accuracy of some comments (where necessary), and incorporated
comments where appropriate.
A. Glossary
Summary of Comments: One commenter submitted recommended changes to the list of
acronyms and abbreviations, and the glossary pertaining to "M"; "KC1"; "pad"; and the phrase,
"wells that have been 'screened-ouf cannot be used for gas production."
EPA Response: After reviewing and checking on the accuracy of the above comments, EPA
incorporated changes to the glossary and list of acronyms, where appropriate.
B. Other Executive Summary Comments
Summary of Comments: EPA received many comments that were specific to the executive
summary of the report, including recommendations for revising the text, tables, and figures. A
few commenters suggested that the language regarding the findings and conclusions of the study
needs to be clearer and stronger (e.g., qualifiers such as "appear to be low" and "persuasive
evidence" weakens the conclusions). Another suggested that, in general, the executive summary
and the main document need to point out that not all USDWs are currently being used nor will
they ever be used as sources of drinking water. Some commenters felt that the executive
summary was inappropriately long and provided suggestions for making the section shorter,
including eliminating all tables from this section. Many commenters provided specific editorial
comments.
A few commenters expressed concern regarding the "graphic language" in Table ES-2 (Summary
ofMSDSsfor Hydraulic Fracturing Fluid Additives) used to describe the health effects of
fracturing fluids, and noted that they felt it may be unnecessarily alarming, and potentially
misleading to readers (i.e., it does not clarify that the health effects only pertain to some
constituents that may or may not be present in the fracturing fluids). Commenters added that
Table ES-2 suggests that linear gel delivery systems always contain diesel and does not indicate
that fluid additives are greatly diluted. One commenter felt that the information provided in
Table ES-4 (Evidence in Support of Coal-USDW Co-Location in U.S. Coal Basins) was too
general, and believed that the information should just be presented in the more detailed sections
from which it was summarized. Other commenters were concerned that the information
provided in Table ES-5 (Summary of Reported Incidents that Associate Water Quality/Quantity
with CoalbedMethane (CBM) Activity) could be misleading to the public.
One commenter felt that the executive summary figures in general were "confusing and
misleading." Other commenters questioned the accuracy and clarity of Figure ES-2 (Graphical
Representation of the Hydraulic Fracturing Process in Coalbed Methane Wells), which depicts
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 26 FINAL
image:
drinking water wells drawing down into coal seams. One commenter questioned the accuracy of
the illustrations in Figures ES-3 (Direct Fluid Injection into a USDW (Coal within USDW)) and
ES-4 (Fracture Creates Connection to USDW) regarding the depth of the water wells and the
direction of fluid migration (i.e., fracturing fluids are shown to be flowing away from the well
bore toward the drinking water wells). The commenter pointed out that the descriptive text on
page ES-10 conflicts with the depiction of fluid migration in Figure ES-4.
EPA Response: EPA has reviewed and considered all comments regarding the executive
summary of the document. The Agency originally designed the executive summary to be a
stand-alone document. Because many readers of such a document (such as Congress or the
leaders of various stakeholder organizations) may have limited time to dedicate to the review of
a large technical document, EPA included essential summary information, including tables and
figures, in the executive summary. However, based on the comments received, EPA has pared
down the executive summary by taking out most of the tables and summarizing key information
from these tables in narrative form. EPA incorporated many of the specific suggestions related
to the figures (e.g., decreasing the depth of drinking water wells), and in some instances,
provided clarifying language to explain the figures.
C. Other Chapter 1 Comments (Introduction)
Summary of Comments: A few commenters provided comments regarding the Introduction to
the hydraulic fracturing report. Comments included questions about the accuracy of the figures,
and how they were depicted: groundwater flow; the relation between well depths and coal
seams; and the point-of-injection for the fracturing fluids. One commenter objected to the
statement that the study was "based on a high level of interest of stakeholders..." when it was the
commenters' understanding that it was based only on a "handful" of complaints.
EPA Response: The statement that the study was "based on a high level of interest of
stakeholders..." is an accurate statement but the term "stakeholders" was vague. To be more
descriptive, Chapter 1 of the final report indicates that a reason for conducting the study was
"concerns voiced by individuals who may be affected by coalbed methane development. . ." The
Agency addressed each of the other comments by either incorporating suggested language or
making relevant clarifications in the document language and figures.
D. Other Chapter 2 Comments (Methodology)
No substantive comments received on this chapter.
E. Other Chapter 3 Comments (Characteristics of CBM Production and HF Practices)
Summary of Comments: EPA received several comments regarding the information in Chapter
3. In particular, several commenters questioned the study's assumptions regarding recovery rates
and fracture heights. A more detailed summary of the comments received on these topics can be
found in sections III.C.2 and IV.A, respectively. One commenter had several specific questions
regarding statements made in this chapter, including: the meaning of the term "conventional
coal mines"; statements regarding the number of CBM wells in Alabama; the discussion of the
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 27 FINAL
image:
origin of CBM; the statement that "coal has very little natural permeability"; contradictions
between the discussion of fluids migration in this chapter compared to that shown in Figures
ES-4 and 1-3; accuracy and clarity of statements regarding the rate of fluid recovery; and the
statement that many CBM wells are re-fractured.
EPA Response: EPA appreciates the detailed comments that were submitted regarding Chapter
3 of the hydraulic fracturing report. The Agency made several editorial corrections and
clarifications to this chapter based on these comments. A more detailed response regarding
recovery rates, fracture heights, and re-fracturing of the same wells can be found in sections
III.C.2, IV.A, and IV.B, respectively.
F. Other Chapter 4 Comments (HF Fluids)
Summary of Comments: Comments specific to Chapter 4 of the report included questions about
the calculation of the constituents of concern at the point-of-injection, and other editorial
comments and suggestions.
EPA Response: In response to comments received on Chapter 4, EPA has incorporated
clarifying language regarding its calculations of BTEX compounds at the point-of-injection.
Other editorial corrections and clarifications have also been incorporated. For a discussion of
how EPA revised its procedure for assessing the potential effect of fracturing fluid constituents
on USDWs from that presented in the draft report, refer to section III.C. 1.
G. Other Chapter 5 Comments (Basin Descriptions)
Summary of Comments: Several comments were received regarding the basin descriptions,
including updates from a few states on the numbers of wells in the applicable basins. One
commenter suggested additional references that should be used to correct some of the statements
regarding the Pottsville Formation. The other four commenters each provided specific editorial
suggestions on one of the following four basins: the Central Appalachian Basin, the Northern
Appalachian Basin, the Uinta Basin, and the Powder River Basin.
EPA Response: EPA has incorporated the updated well information provided by states. All
other editorial comments were considered, and most were incorporated. Other basin-specific
issues are discussed in section VIII of this document.
H. Other Chapter 6 Comments (Water Quality Incidents)
Summary of Comments: Several comments were received regarding the water quality incidents
chapter of the report. Commenters made specific editorial suggestions, and provided
clarifications about specific complaints, additional information about how their state investigates
complaints, and information about state-specific hydraulic fracturing regulations. One
commenter stated that the discussion of the Pottsville, Allegheny, Conemaugh, and
Monongahela Groups were "oversimplified" and questioned the conflicting use of the terms
"cyclothem" and "complex" when describing the depositional environments of the Allegheny
Group.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 28 FINAL
image:
A few commenters expressed concern that the descriptions of public complaints (including the
information summarized in Table 6-2) are presented in the report as if the information was
factual, without linking the complaints to actual findings following the state and EPA
investigations. One commenter indicated that EPA does not present any data from state
agencies, which suggests to the commenter that no real scientific studies were conducted.
Commenters recommended that the complaints be immediately followed by a summary of the
evaluation and resolution of the complaint. One recommendation was that, if kept in the report,
the information be moved to an appendix.
Finally, some commenters felt that EPA was contradictory regarding the question of whether
hydraulic fracturing of CBM wells threaten USDWs. For example, one commenter indicated
that EPA had concluded in Chapter 6 that there is insufficient evidence to determine if there is a
link between fracturing and USDW contamination. However, elsewhere in Chapter 6, EPA
states that "water quality problems might be associated with some of the variety of production
activities common to CBM extraction. These production activities include... methane migration
through conduits created by drilling and fracturing practices..."
EPA Response: In response to stakeholder's comments on EPA's original study methodology,
EPA compiled citizen complaints and reported incidences of CBM impacts on drinking water
wells and included these accounts in Chapter 6 of the report. In the final report, EPA has
clarified the rationale for including citizen complaints in its report.
The final report also clarifies that many of the reported impacts (such as impacts to water supply
quantities and effects of discharge of groundwater extracted in the CBM production process)
included in Chapter 6 are outside of the scope of SDWA and beyond the scope of the Phase I
study. The goal of the Phase I study was to assess the potential for contamination of USDWs
due to the injection of hydraulic fracturing fluids into CBM wells, and to determine based on
these findings if further study is warranted. EPA also incorporated information that was
provided by states regarding incident reports, and state-specific regulations. Finally, the Agency
took Table 6-2 out of the document because, as suggested by some commenters, summarizing
citizen complaints in a tabular format oversimplified this information, and created a potential for
misinterpretation. The information in Table 6-2 is presented in detail in the main body of
Chapter 6.
See also EPA's response to comment in section II.D of this document regarding other issues
pertaining to water contamination incidents and citizen complaints.
I. Other Chapter 7 Comments (Conclusions and Recommendations)
Summary of Comments: Most comments received regarding Chapter 7 of the report also relate
back to prior report chapters. Several commenters had specific suggestions or questions
regarding the conclusions and recommendations section of the report. Some of these
commenters agreed with the conclusions of the study, but recommended that EPA put more
emphasis on the conclusions, and include information about the findings of the study earlier in
the document. Specifically, commenters suggested that, at the beginning of the document, EPA
include a statement clarifying that: "EPA finds no evidence of harm from hydraulic fracturing
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 29 FINAL
image:
while investigating the reported incidents that spurred the study." These commenters felt that
EPA's findings that Phase II of the study is unnecessary, and that little or no public health threat
is posed by hydraulic fracturing should be more strongly stated in the conclusions of the report.
Note that commenter opinions regarding Chapter 7 of the report do not reflect the overall
commenter perspectives regarding the outcome and conclusions of the study. Most of the
commenters expressed opinions regarding the study's conclusions, but did not state them within
the context of Chapter 7.
EPA Response: EPA has reviewed all commenter suggestions regarding Chapter 7, and
incorporated the majority of these comments where appropriate. Other revisions to Chapter 7,
which relate back to changes in previous chapters, have been made in order to ensure internal
consistency within the document.
VIII. BASIN DESCRIPTIONS
This summary of basin-specific comments focuses mainly on those comments that have not been
summarized within the issue-specific Sections II through VI of this document. Many comments
were received that provided minor editorial suggestions and factual corrections regarding basin
descriptions. The Agency has considered all comments, researched the accuracy of some
comments (where necessary), and incorporated public comments where appropriate.
A. San Juan Basin
Summary of Comments: One commenter provided suggested edits and corrections pertaining to
the San Juan Basin geology, hydrology and USDW identification, and CBM production activity.
This commenter also provided additional references.
EPA Response: EPA reviewed and considered all suggested edits and corrections and has
incorporated revisions to the San Juan Basin descriptions. EPA also reviewed the additional
references provided by the commenter, and incorporated additional pertinent information.
B. Black Warrior Basin
Summary of Comments: One commenter provided a variety of editorial comments and factual
clarifications regarding the Black Warrior Basin. Examples of information the commenter
questioned include: coal thickness; total dissolved solids levels; number of active Class II wells
in this area; fracture height vs. length; and chemical components of fracturing fluids.
EPA Response: EPA has incorporated into the final report the majority of the commenter's
suggestions regarding the description of the Black Warrior Basin.
C. Piceance Basin
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 30 FINAL
image:
Summary of Comments: One commenter provided a brief description of the activities and
progress of the pilot program in the White River Dome field.
EPA Response: The final report contains the information provided by the commenter.
D. Uinta Basin
Summary of Comments: One commenter indicated that the information on the Castlegate Field
is out of date. The commenter clarified that the field is currently in production, and explained
why he believes that cross-contamination from the Blackhawk to the Castlegate Sandstone and
Star Point Sandstone (as indicated in the report) is unlikely.
EPA Response: EPA has made revisions to the basin description based on this information.
E. Powder River Basin
Summary of Comments: No substantive comments were submitted on this section.
F. Central Appalachian Basin
Summary of Comments: One commenter provided clarifications and corrections regarding CBM
activity, regulations, and drinking water sources in Virginia.
EPA Response: EPA has incorporated many of the commenter's clarifications into the basin
description.
G. Northern Appalachian Basin
Summary of Comments: One commenter provided information on the square mileage and
number of CBM wells in this basin, with associated references. This commenter, who is the
individual that was interviewed for some of the information provided in this attachment,
provided edits to the interview summary. Another commenter suggested several editorial
corrections pertaining to the location of specific coal groups, the use of the term "group," and the
use of the term "separated laterally" vs. "vertical separation."
EPA Response: EPA has incorporated all appropriate information into the basin description.
H. Western Interior Basin
Summary of Comments: This commenter questioned the accuracy of the statement that "coal
seams could be coincident with a USDW" within the Cherokee Basin. The commenter discussed
the aerial extent to which various coal seams in the Cherokee Basin coincide with USDWs, and
recommended that EPA also review a 1997 paper entitled "Kansas coal resources and their
potential for coalbed methane."
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBM Study 31 FINAL
image:
EPA Response: EPA has modified the report to indicate that "all or part of targeted coal seams
could be coincident with a USDW," thereby clarifying the summary of the data provided in
Table A8-2, which presents the relative depths of coal seams and USDWs.
I. Raton Basin
Summary of Comments: No comments were submitted on this section.
J. Sand Wash Basin
Summary of Comments: One commenter pointed out that in the Sand Wash Basin, the pilot at
Craig Dome was abandoned "due to excessive water production." This commenter also believed
that EPA's findings that hydraulic fracturing poses very little potential threat to USDWs does not
account for proximity or overlap with natural fault lines. The commenter stated that: "if a
fracture propagates into and along a fault plane, it may contaminate a USDW."
EPA Response: EPA has incorporated the commenter's information into Attachment 10 of the
final report.
K. Washington Coal Regions (Pacific and Central)
Summary of Comments: No comments were submitted on this section.
Public Comment and Response Summary for June 2004
Hydraulic Fracturing CBMStudy 32 FINAL
image:
EPA 816-R-04-003 Master References
REFERENCES
This master reference list pertains only to Chapters 1 through 7 of this
document. Separate reference lists are provided for each appendix and
attachment, and are provided at the end of each of these sections.
Alabama Oil and Gas Board, Administrative Code, Oil and Gas Report 1, 400-3.
Alabama Oil and Gas Board. 2002. Public Comment OW-2002-0002-0029 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Andrews, Richard D., Cardott, Brian J., and Storm, Taylor. 1998. The Hartshorne Play
in Southeastern Oklahoma: regional and detailed sandstone reservoir analysis and
coalbed-methane resources. Oklahoma Geological Survey, Special Publication
98-7.
Baldwin. 2000. Colorado Oil and Gas Conservation Commission, personal
communication.
Bodnar, G. 1999. Colorado Department of Health, personal communication.
Bostic, Joy L., Brady, L. L., Howes, M. R., Burchett, R. R., and Pierce, B. S. 1993.
Investigation of the coal properties and the potential for coal-bed methane in the
Forest City Basin. U. S. Geological Survey, Open File Report 93-576.
Brady, L. L. 2002. Kansas Geological Survey, personal communication.
Buckovic. 1979. The Eocene Deltaic System of West-central Washington: in
Armentrout, J. M., Cole, M. R., and TerbestH., eds., Cenozoic paleogeography
of the western United States: Los Angeles, Soc. Econ. Paleont. and Min., Pacific
Section, Pacific Coast Paleogeog. Symposium 3, p. 147-163, as cited by Choate et
al., 1980.
Carrol, R.E., Pashin, J.C., and Kugler, R.L. 1993. Burial history and source rock
characteristics of Mississippian and Pennsylvanian strata, Black Warrior basin,
Alabama, Alabama Geological Society Guidebook, pp. 79-88.
Charpentier, Ronald R. 1995. Cherokee Platform Province. U. S. Geological Survey,
National Assessment of United States Oil and Gas Resources.
Chavez, F. 2001. New Mexico Oil Conservation Division, personal communication.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-1
image:
EPA816-R-04-003 Master References
Choate, R, Johnson, D.A., and McCord, J.P. 1980. Geologic overview, coal, and
coalbed methane resources of the Western Washington coal region, Lakewood,
Colorado. TRW Energy Systems Group Report for U.S. Department of Energy,
Morgantown Energy Technology Center, Contract DE-AC21-78MC08089, pp. 353-
372.
Choate, R., Lent, T., and Rightmire, C.T. 1993. Upper Cretaceous geology, coal, and the
potential for methane recovery from coalbeds in the San Juan Basin - Colorado and
New Mexico. AAPG Studies in Geology, 38:185-222.
Clark, R.C. and Brown, D. W. 1977. Petroleum's Properties and Analyses in Biotic and
Abiotic Systems. In: Effects of Petroleum on Arctic and Subarctic Marine
Environments and Organisms, (ed. Malins, D.C.) Vol. 1).
Close, Jay. C. 1993. Natural Fractures in Coal; Chapter 5 of AAPG Studies in Geology
38, "Hydrocarbons from Coal", pp. 119-133.
Colorado GIS. 2001. Approved Drilling Permits,
http://cogccweb.state.co.us/cogis/DrillingPermitsList.asp
Colorado Oil and Gas Conservation Commission and New Mexico Oil Conservation
Division, personal communication, 2001.
Colorado Oil and Gas Conservation Commission. 2001.
http://www.oil-gas.state.co.us/
Condra, G. E. and Reed, E. C. 1959. The geological section of Nebraska. Nebraska
Geological Survey Bulletin 14A, 1959.
Consolidated Industrial Services, Inc. 2001. Hydraulic fracturing site visit notes,
Western Interior Coal Region, State of Kansas.
Cordova, Robert M. 1963. Reconnaissance of the ground-water resources of the
Arkansas Valley Region, Arkansas. Contributions to the Hydrology of the United
States, Geological Survey Water-Supply Paper 1669-BB, 1963.
Cramer, D.G. 1992. The unique aspects of fracturing western U.S. coalbeds; Journal of
Petroleum Technology, October 1992, pp. 1126-1133.
DASC Web site. 2001a. Kansas elevation map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html.
DASC Web site. 2001b. Ozark Aquifer base map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-2
image:
EPA816-R-04-003 Master References
DeBruin, Rodney H., Lyman, Robert M., Jones, Richard W., and Cook, Lance W. 2000.
Information Pamphlet 7. Wyoming State Geological Survey.
Diamond, W.P. 1987a. Underground observations of mined-through stimulation
treatments of coalbeds. Quarterly Review of Methane from Coal Seams
Technology, v. 4, n. 4 (June 1987), pp. 19-29.
Diamond, W.P. 1987b. Characterization of Fracture Geometry and Roof Penetration
Associated with Stimulation Treatments in Coalbeds. Proceedings of the 1987
Coalbed Methane Symposium, University of Alabama (Tuscaloosa), pp. 243.
Diamond, W.P. and D.C. Oyler. 1987. Effects of stimulation treatments on coalbeds and
surrounding strata, evidence from underground observations. US Department of
Interior, RI9083, USBM, pp. 1-47.
Dion, N. P. 1984. Washington Ground-Water Resources. In National Water Summary,
U.S. Geological Survey Water-Supply Paper No. 2275, pp. 433-438.
Duigon, Mark T. and Smigaj, Michael J. 1985. First report on the hydrologic effects of
underground coal mining in Southern Garrett County, Maryland, U.S. Geological
Survey Report of Investigations No. 41.
Energy Information Agency. 2001. "U.S. Crude Oil, Natural Gas, and Natural Gas
Liquids Reserves 2000 Annual Report," DOE/EIA-0216 (2000) Advance Summary,
Nov. 2001, p. 17.
Elbel, J.L. 1994. Field evaluation of acid-fracturing treatments with geometry
simulation, buildup, and production data; SPE Production and Facilities, v.9 n.l
(February 1994), pp. 43-46.
Elder, C.H. and Deul, M. 1974. Degasification of the Mary Lee coalbed near Oak
Grove, Jefferson county, Alabama, by vertical borehole in advance of mining; U.S.
Bureau of Mines Report 7968.
Eleventh Circuit Court of Appeals, 1997. LEAF v. EPA, 118F.3d 1467.
Ellard, J.S., Roark, R.P., and Ayers, W.B. 1992. Geologic controls on coalbed methane
production: an example from the Pottsville formation, Black Warrior Basin,
Alabama USA. Symposium on Coalbed Methane Research and Development in
Australia. James Cook University, p. 45-61.
Ely, J.W., Zbitowski, R.I., and Zuber, M.D. 1990. How to develop a coalbed methane
prospect: a case study of an exploratory five-spot well pattern in the Warrior basin,
Alabama; SPE Paper 20666, Proceedings of the 65th Annual Technical Conference
(New Orleans), pp. 487-496.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-3
image:
EPA816-R-04-003 Master References
Ely, John W. 1985. Secondary recovery of oil, oil wells, hydraulic fracturing.
Stimulation Engineering Handbook, ix, 357 p.
Fetter, C.W. 1994. Applied Hydrogeology, 3rd ed. New York: Macmillan College
Publishing Company.
Flores, R.M. and Bader, L.R., 1999. Fort Union Coal in the Powder River Basin,
Wyoming and Montana: A synthesis. U.S. Geological Survey Professional Paper
1625-A, Chapter PS, 49 pp., on CD-ROM.
Flowerday, C. F., Kuzelka, R. D., and Pederson, D. T., compilers. 1998. The Ground
Water Atlas of Nebraska.
Foster, James B. 1980. Fresh and saline ground-water map of West Virginia. U. S.
Geological Survey, West Virginia Geological and Economic Survey, Map WV-
12.
Friedman, Samuel A. 1982. Determination of reserves of methane from coalbeds for use
in rural communities in eastern Oklahoma. Oklahoma Geological Survey, Special
Publication 82-3, 1982.
Garrison, James R., Jr. (non-survey author), van den Bergh, T.C.V. (non-survey author),
Barker, Charles E., Tabet, David E. (non-survey author). 1997. Deposit!onal
sequence stratigraphy and architecture of the Cretaceous Ferron Sandstone;
implications for coal and coalbed methane resources; a field excursion. Link, Paul
Karl (non-survey editor), Kowallis, Bart J. (non-survey editor), Mesozoic to Recent
geology of Utah, Geology Studies, 42(2): 155-202.
Geological Survey of Alabama, 1930's to Present, Hydrogeology Well Records, open-file
data for Tuscaloosa, Jefferson and Walker Counties, Alabama, as cited by the
Alabama State Oil and Gas Board, 2002.
Gloyn, Robert W. and Sommer, Steven N. 1993. Exploration for coalbed methane gains
momentum in Uinta Basin. Utah Geological Survey, Oil & Gas Journal.
Exploration, pp. 73-76, May 31, 1993.
Graham, G. 2001. Colorado Division of Water Resources, personal communication.
Gray, Ian. 1987. Reservoir engineering in coal seams: the physical process of gas
storage and movement in coal seams; SPE Reservoir Engineering, v. 2, no. 1, pp.
7-14.
GRI (Gas Research Institute). 1995. Fracturing Experience at the Rock Creek Multiple
Coal Seams Project; Topical Report, prepared by S.W. Lambert, J.L. Saulsberry,
P.F. Steidl, M.W. Conway, and S.D. Spafford, July 1995.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-4
image:
EPA816-R-04-003 Master References
GRI (Gas Research Institute). 1996. Coordinated Studies in Support of Hydraulic
Fracturing of Coalbed Methane, Final Report # GRI-95/0283, prepared by Stim-
Lab, Inc. February 1996, pp. 2.4-1-2.4-3.
GTI (Gas Technology Institute) Web site. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Halliburton, Inc., Virginia Site Visit. 2001. EPA observed hydraulic fracturing of gas
wells performed by Halliburton, Inc. for Consol Energy (VA) in the Central
Appalachian Basin. August 9, 2001.
Halliburton, Inc. 2002. Personal communication with Halliburton staff, fracturing fluid
experts: Joe Sandy, PatFinley, and Steve Almond. March 1, 2002.
Halliburton, Inc. 2003. Personal communication with Halliburton staff, fracturing fluid
expert, Steve Almond. April 2003.
Hanson, M.E., Neilsen, P.E., Sorrels, G.G., Boyer, C.M., and Schraufnagel, R.A. 1987.
Design, execution, and analysis of a stimulation to produce gas from thin multiple
coal seams; SPE Paper 16860, Proceedings, 1987 Society of Petroleum Engineers
annual technical conference and exhibition.
Harper, T.R, Hagans, J.T., and Martins, J.P. 1985. Fracturing without proppant. SPE
13858, Proceedings SPE Low Permeability Reservoirs Symposium (Denver), p.
83.
Heath, Brian. 1999. Environmental Analyst, Wyoming Department of Environmental
Quality, Water Quality Division. Telephone Interview with EPA conducted
August 19, 1999.
Hemborg, H. T. 1998. Spanish Peak Field, Las Animas County, Colorado: Geologic
setting and early development of a coalbed methane reservoir in the Central Raton
Basin. Colorado Geological Survey, Dept. of Natural Resources, Denver, CO,
Resource Series 33, 34 pp.
Hill, David. 2001. Gas Technology Institute. Personal communication, peer-review
panelist.
Hinkel, J.J., Nimerick, K.H., England, K., Norton, J.C., and Roy, M. 1991. Design and
evaluation of stimulation and workover treatments in coal seam reservoirs;
Proceedings 1991 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), Tuscaloosa, pp. 453-458.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-5
image:
EPA816-R-04-003 Master References
Holditch, S.A 1990. Completion methods in coal seam reservoirs. SPE Paper No.
20670. Proceedings 1990 SPE Annual Technical Conference and Exhibition
(Production Operations and Engineering), pp. 533-542.
Holditch, S.A. 1993 Completion methods in coal-seam reservoirs; Journal of Petroleum
Technology, v.45 n.3 (March 1993), pp. 270-276.
Holditch, S.A., Ely, J.W., Semmelbeck, M.E., Carter, R.H., Hinkle, 1, and Jeffrey, R.G.
1988. Enhanced recovery of coalbed methane through hydraulic fracturing; SPE
Paper 18250, Proceedings 1988 SPE Annual Technical Conference and
Exhibition (Production Operations and Engineering), p. 689.
Holditch, S.A., Ely, J.W., and Carter, R.H. 1989. Development of a coal seam fracture
design manual; Proceedings, 1989 Coalbed Methane Symposium, Tuscaloosa,
Alabama, pp. 299-320.
Holland, W. 1999. La Plata County, personal communication.
Hopkins, Herbert T. 1966. Fresh-saline water interface map of Kentucky. U. S.
Geological Survey, Kentucky Geological Survey, Series X.
Hudson, Harold & Staff, BJ Services, 2002. Personal communication, conference call
with EPA on April 19, 2002.
Hunt, A. M., and Steele, D. J. 1991. Coalbed methane development in the Northern and
Central Appalachian Basins - past, present and future. The 1991 Coalbed
Methane Symposium, The University of Alabama/Tuscaloosa, May 13-16, 1991.
Jeffrey, R.G., J.R. Enever, T. Ferguson, J. Bride, R. Phillips, and S. Davidson. 1993.
Small-scale hydraulic fracturing and mineback experiments in coal seams.
Proceedings of the 1993 International Coalbed Methane Symposium. University
of Alabama (Tuscaloosa). May 17-21, 1993. pp. 79-88.
Jeu, S.J., Logan, T.L., and McBane, R.A. 1988. Exploitation of deeply buried coalbed
methane using different hydraulic fracturing techniques; SPE paper 18253,
Proceedings 63rd Annual Technical Conference (Houston).
Johnson, R. C. 1989. Geologic history and hydrocarbon potential of Late Cretaceous-
age, low-permeability reservoirs, Piceance Basin western Colorado: U. S.
Geological Survey Bulletin 1787-E, 51 p.
Johnston, William D., Jr., 1933. Ground water in the Paleozoic rocks of northern
Alabama: Geological Survey of Alabama, Special Report No. 16, as cited by the
Alabama State Oil and Gas Board, 2002.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-6
image:
EPA816-R-04-003 Master References
Jones, A.H., and Schraufnagel, R.A. 1991. In-situ stress variations in the Black Warrior
basin; Proceedings of 1991 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), Tuscaloosa, p. 75.
Jones, A.H., Bell, G.J., and Morales, R.H. 1987a. Examination of potential mechanisms
responsible for the high treatment pressures observed during stimulation of
coalbed reservoirs; SPE Paper 16421, Proceedings, Department of Energy/SPE
Symposium: Gas from Low Permeability Reservoirs, p. 317.
Jones, A.H., Bell, G.J., and Morales, R.H. 1987b. The influence of coal fines/chips on
the behavior of hydraulic fracture stimulation treatments; Proceedings of 1987
Coalbed Methane Symposium, University of Alabama (Tuscaloosa), Tuscaloosa,
pp. 93-102.
Kaiser, W.R. and Ayers, W.B.Jr. 1994. Coalbed methane production, Fruitland
formation, San Juan Basin: geologic and hydrologic controls. New Mexico
Bureau of Mines and Minerals Bulletin 146: Coalbed methane in the upper
Cretaceous Fruitland formation, San Juan Basin, New Mexico and Colorado, pp.
187-207.
Kaiser, W.R. and Scott, A.R. 1994. Hydrologic setting of the Fort Union Formation,
Sand Wash Basin. Report of Investigations - Geologic and Hydrologic Controls
on Coalbed Methane, Texas, University, Bureau of Economic Geology, 220, pp.
115-125.
Kaiser, W.R., Scott, A., Zhou, N., Hamilton, D.S., and Tyler, R. 1994a. Resources and
Producibility of Coalbed Methane in the Sand Wash Basin. Report of
Investigations - Geologic and Hydrologic Controls on Coalbed Methane, Texas
University, Bureau of Economic Geology, 220, pp. 129-145.
Kaiser, W.R., Swartz, T.E., and Hawkins, G.J. 1994b. Hydrologic framework of the
Fruitland formation, San Juan Basin. New Mexico Bureau of Mines and Minerals
Bulletin 146: Coalbed methane in the upper Cretaceous Fruitland formation, San
Juan Basin, New Mexico and Colorado, pp. 133-164.
Kelafant, J. R., Wicks, D. E., Kuuskraa, V. A. March 1988. A geologic assessment of
natural gas from coal seams in the Northern Appalachian Coal Basin. Topical
Report - Final Geologic Report (September 1986 - September 1987).
Keller, Adam. 1999. Planner with La Plata County, personal communication.
Koenig, R.A. 1989. Hydrologic characterization of coal seams for optimum dewatering
and methane drainage; Quarterly Review of Methane from Coal Seams
Technology, v.7, pp. 30-33.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-7
image:
EPA816-R-04-003 Master References
Larsen, Very E., 1989. Preliminary evaluation of coalbed methane geology and activity
in the Recluse Area, Powder River Basin, Wyoming, Quarterly Review of
Methane from Coal Seams Technology, June.
Laubach, S. E., Marrett, R. A., Olson, J. E., and Scott, A. R., 1998, Characteristics and
origins of coal cleat: a review: International Journal of Coal Geology, v. 35, p.
175-207, as cited by Olson, Jon, 2002.
Laubach, S.E. and C.M. Tremain, 1991. Regional coal fracture patterns and coalbed
methane development, in J.C. Roegiers, ed., Rock Mechanics as a
Multidisciplinary Science, Proceedings of the 32nd U.S. Symposium, University
of Oklahoma atNorman. Rotterdam, A.A. Balkema, p. 851-859.
Law, Ben E., Rice, Dudley D., Flores, Romeo M. 1991. Coalbed gas accumulations in
the Paleocene Fort Union Formation, Powder River Basin, Wyoming. Rocky
Mountain Association of Geologists, Coalbed Methane.
Levine, J.R. 1993. Coalification: the evolution of coal as a source rock and reservoir
rock for oil and gas; AAPG Studies in Geology 38, pp. 39-77. [Note: This
source was also cited by the Alabama Oil and Gas Board. 2002. Public Comment
OW-2002-0002-0029 to "Draft Evaluation of Impacts to Underground Sources of
Drinking Water by Hydraulic Fracturing of Coalbed Methane Reservoirs."
Federal Register. Vol. 63, No. 185. p. 33992, September 24, 2002.]
Lyons, Paul C. 1997. Central-Northern Appalachian Coalbed Methane Flow Grows. Oil
& Gas July 7, 1997, pp. 76-79.
Mavor, M.J., Dhir, R., McLennan, J.D., and Close, J.C. 1991. Evaluation of the
hydraulic fracture stimulation of the Colorado 32-7 No. 9 well, San Juan basin;
Rocky Mountain Association of Geologists Guidebook, "Coalbed methane of
Western North America", Fall Conference and Field Trip, pp. 241-249.
McCord, E. 1999. Attorney at Law, personal communication.
Macfarlane, A. 2001. Kansas Geological Survey, personal communication.
McKee, C.R., Bumb, A.C., and Koenig, R.A. 1989. Stress-dependent permeability and
porosity of coal; Proceedings 1987 Coalbed Methane Symposium, University of
Alabama (Tuscaloosa), pp. 16-19.
Merchat, Walt. 1999. Consultant for the Powder River Basin Resource Council,
Telephone Interview with EPA conducted May 17, 1999.
Messina, Inc. Web site. 2001. Fracturing fluid manufacturer for the oil and gas industry.
www.messinaoil-chem.com.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-8
image:
EPA816-R-04-003 Master References
Missouri Division of Geological Survey and Water Resources. 1967. Mineral & Water
Resources of Missouri, 43(2).
Montgomery, Scott L. 1999. Powder River Basin, Wyoming: An expanding coalbed
methane (CBM) play. American Association of Petroleum Geologists Bulletin,
(August).
Morales, R,H, McLennan, J.D., Jones, A.H., and Schraufnagel, R.A. 1990.
Classification of treating pressures in coal fracturing; Proceedings of the 31st U.S.
Symposium on Rock Mechanics, 31, pp. 687-694.
Mukherjee, Hemanta and Cudney, Greg. 1993. Extension of acid fracture penetration by
drastic fluid-loss control. JPT Journal of Petroleum Technology, 45(2): 102-105.
Naceur, K.B. and Touboul, E. 1990. Mechanisms controlling fracture height growth in
layered media; SPE Production Engineering, v.5 n.2 (May 1990), pp. 142-150.
Nielsen, P. E. and Hanson, M. E. 1987. Analysis and Implications of Three Fracture
Treatments in Coals at the USX Rock Creek Site Near Birmingham, Alabama,
1987 Coalbed Methane Symposium, Tuscaloosa, AL (Nov. 16-19, 1987).
Nolte, K.G. and Economides, M.J. 1991. Fracture design and validation with uncertainty
and model limitations; Journal of Petroleum Technology, v.43 n.9 (September
1991), pp. 1147-1155.
Nolte, K.G. and Smith, M.B. 1981. Interpretation of fracturing pressures; Journal of
Petroleum Technology, September 1981, pp. 1767-1775.
Oklahoma Corporation Commission (OCC), Depth to Base of Treatable Water Map
Series, 2001.
Oklahoma Geological Survey Web site. 2001. http://www.ou.edu/special/ogs-pttc.
Olson, Jon. 2001. University of Texas at Austin. Personal communication, peer-review
panelist.
Osborne, Paul. 2002. USEPA Region VIIIUIC Program, personal communication.
Palmer, ID., Davids, M.W., and Jeu, S.J. 1989. Analysis of unconventional behavior
observed during coalbed fracturing treatments, Proceedings of the 1989 Coalbed
Methane Symposium, University of Alabama/Tuscaloosa, April 17-20, 1989, pp.
395-411.
Palmer, ID. and Sparks, D.P. 1990. Measurement of induced fractures by downhole TV
camera in coalbeds of the Black Warrior Basin; Society of Petroleum Engineers
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-9
image:
EPA816-R-04-003 Master References
paper number 20660, Proceedings, 1990 Society of Petroleum Engineers annual
technical conference and exhibition, pp. 445-458.
Palmer, ID., Fryar, R.T., Tumino, K.A., and Puri, R. 1991a. Comparison between gel-
fracture and water-fracture stimulations in the Black Warrior basin; Proceedings
1991 Coalbed Methane Symposium, University of Alabama (Tuscaloosa), pp.
233-242.
Palmer, ID., Fryar, R.T., Tumino, K.A., and Puri, R. 1991b. Water fracs outperform gel
fracs in coalbed pilot; Oil and Gas Journal, August 12, 1991, pp.71 - 76.
Palmer, ID., King, N.S., and Sparks, D.P. 1991c. The character of coal fracture
treatments in Oak Grove field, Black Warrior basin, SPE paper no. 22914,
Proceedings, 1991 Society of Petroleum Engineers annual technical conference
and exhibition, pp. 277-286.
Palmer, ID., King, N.S., and Sparks, D.P. 1993a. The character of coal fracture
treatments in the Oak Grove field, Black Warrior Basin. In Situ, Journal of Coal
Research, 17(3):273-309.
Palmer, ID., Lambert, S.W., and Spitler, J.L. 1993b. Coalbed methane well completions
and stimulations. Chapter 14 of AAPG Studies in Geology 38, pp. 303-341.
Pappajohn, S. P., and Mitchell, T. E. 1991. Delineation of prospective coalbed methane
trends in western and central Washington State. In Schwochow, S. D. Coalbed
methane of western North America guidebook for the Rocky Mountain
Association of Geologists Fall Conference and Field Trip Sept. 17-20, 1991,
Glenwood Springs, CO, pp. 163-178.
Pashin, J.C. and Hinkle, F. 1997. Coalbed Methane in Alabama. Geological Survey of
Alabama Circular 192, 71 pp.
Pashin, J.C. 1994a. Coal body geometry and syn-sedimentary detachment folding in the
Oak Grove coalbed methane field, Black Warrior basin, Alabama; AAPG
Bulletin, v. 78, pp. 960-980.
Pashin, J.C. 1994b. Cycles and stacking patterns in Carboniferous rocks of the Black
Warrior foreland basin; Transactions of the Gulf Coast Association of Geological
Societies (AAPG), 44, pp.555-563.
Pashin, J.C., Ward, W.E., Winston, R.B., Chandler, R.V., Bolin, D.E., Richter, K.E.,
Osborne, W.E, and Sarnecki, J.C. 1991. Regional analysis of the Black Creek-
Cobb coalbed methane target interval, Black Warrior Basin, Alabama. Alabama
Geological Survey Bulletin 145, 127 pp.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-10
image:
EPA816-R-04-003 Master References
Pennsylvania Department of Conservation and Natural Resources, 2002.
Petroleum Technology Transfer Council Web site. 1999.
http://www.pttc.org
Petzet, G. Alan. 1996. Utah coalbed gas exploration poised for growth. Oil & Gas
Journal, Exploration, p. 54, August 5, 1996.
Platt, Steve. January, 2001. U.S. EPA Region 3, personal communication.
Potter, T.L. and K.E. Simmons. 1998. Total Petroleum Hydrocarbon Criteria Working
Group Series, Volume 2. Composition of Petroleum Mixtures. The Association
for Environmental Health and Science. Available on the Internet at:
http://www.aehs.com/publications/catalog/contents/tph.htm.
Powder River Coalbed Methane Information Council (PRCMIC). 2000. Coalbed
Methane Development Information, Sheridan, Wyoming.
Powell, R. J., McCabe, M. A., Slabaugh, B. F., Terracina, J. M., Yaritz, J. G., and Ferrer,
D. 1999. Applications of a new, efficient hydraulic fracturing fluid system. SPE
Production and Facilities, 14(2): 139-143, 1064-668X.
Quarterly Review. 1993. Coalbed methane - state of the industry. Methane From Coal
Seams Technology, August, 1993.
R. Morrison & Associates, Inc. 2001. Environmental Tool Box (newsletter). "Does
diesel #2 fuel oil contain benzene?" Fall 2001.
Rahim, Z. and Holditch, S.A. 1992. The effects of mechanical properties and selection
of completion interval upon the created and propped fracture dimensions in
layered reservoirs; SPE Paper 24349, Proceedings of the 1992 Rocky Mountain
Regional Meeting (Denver).
Rahim, Z., Holditch, S.A, Zuber, M.D., and Buehring, D.R. 1998. Evaluation of fracture
treatment using a layered-reservoir description: Field examples; SPE Production
and Facilities, v.13 nl (February 1998), pp. 21-27.
Rice, C.A., Ellis, M.S., and Bullock, J.H., Jr. 2000. Water co-produced with coalbed
methane in the Powder River Basin, Wyoming: preliminary compositional data.
U.S. Geological Survey Open-File Report 00-372.
Rice, D.D. 1993. Composition and origins of coalbed gas; AAPG Studies in Geology
38, pp. 159-184.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-11
image:
EPA816-R-04-003 Master References
Saulsberry, J.L., Schraufnagel, R.A., and Jones, A.H. 1990. Fracture height growth and
production from multiple reservoirs; SPE paper no. 20659, Proceedings, 1990
Society of Petroleum Engineers annual technical conference and exhibition, pp.
433-443.
Schlumberger, Ltd. 2001. Material Safety Data Sheets for Hydraulic Fracturing Fluids,
Memo to EPA from P. A. VanAllan, Health, Safety and Environmental Councel,
Western Hemisphere. September 18, 2001.
Schraufnagel, R.A. 1993. Coalbed methane production. Chapter 15 of AAPG Studies in
Geology 38, pp. 341-361.
SDWA (Safe Drinking Water Act) Sections 1422 and 1425.
Sedam, A. C., and Stein, R. B. 1970. Saline ground-water resources of Ohio. Hydrologic
Investigations Atlas HA-366, Department of the Interior, U. S. Geological
Survey.
Smith, D. 2002. Missouri Geological Survey, personal communication.
Soot, P. 1991. Tax incentives spur development of coalbed methane; Oil and Gas
Journal, June 1991, pp. 40-42.
Spafford, S.D. 1991. Stimulating multiple coal seams at Rock Creek with access
restricted to a single seam; Proceedings 1991 Coalbed Methane Symposium,
University of Alabama (Tuscaloosa), p. 243.
Steidl, P.F. 1991. Inspection of induced fractures intercepted by mining in the Warrior
basin, Alabama; Proceedings 1991 Coalbed Methane Symposium, University of
Alabama (Tuscaloosa), pp. 181-191.
Stevens, S., Lombardi, T. E., Kelso B. S., and Coates, J. M. 1992. A geologic
assessment of natural gas from coal seams in the Raton and Vermejo Formations,
Raton Basin. GRI Topical Report 92/0345, 84 pp.
Stevens, S.H., Kuuskraa, J.A., and Schraufnagel, R.A. January, 1996. Technology spurs
growth of U.S. coalbed methane. Oil and Gas Journal, pp. 56-63.
Stone, W.J., Lyford, F.P., Frenzel, P.F., Mizell, N.H. and Padgett, E.T. 1983.
Hydrogeology and water resources of San Juan Basin, New Mexico. New Mexico
Bureau of Mines and Mineral Resources, Hydrologic Report 6, 70 p.
TBEG (Texas Bureau of Economic Geology). 2000. Personal communication with
TBEG Staff.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-12
image:
EPA816-R-04-003 Master References
Tessin, Robert. 2001. Colorado Oil and Gas Conservation Commission, personal
communication.
Thompson, J. E., Mchain, C., Gregory, G., and Gerbrandt, D. 1991. New continuous-
mix process for gelling anhydrous methanol minimizes hazards. Proceedings-
SPE Annual Technical Conference and Exhibition, October 6-9, 1991.
Publication by Society of Petroleum Engineers of AIME, pp. 425-436.
Tyler, R., Scott, A.R. and Kaiser, W.R. 1998. Defining coalbed methane exploration
fairways: An example from the Piceance Basin, Rocky Mountain Foreland.
Western United States, Conference Document, March 23-25.
http://georefcos.com/cgi-bin/getRec?un=2001-012340
U.S. Department of the Interior, Bureau of Land Management, Colorado State Office.
1998. Glenwood Spring Resource Area: Oil & Gas Leasing Development, Draft
Supplemental Environmental Impact Statement, June 1998.
U. S. Department of the Interior, Bureau of Land Management, San Juan Field Office.
1999. Coalbed Methane Development in the Northern San Juan Basin of
Colorado, A Brief History and Observations, Working Document, December
1999.
US Environmental Protection Agency. 1987. Guidelines for Delineation of Wellhead
Protection Areas. Office of Ground-Water Protection. June 22, 1987.
US Environmental Protection Agency. 1993. Memorandum: Assistance on compliance
of 40 CFR Part 191 with ground water protection standards. From James R.
Elder, Director, Office of Ground Water and Drinking Water, to Margo T. Oge,
Director, Office of Radiation and Indoor Air. June 4, 1993.
US Environmental Protection Agency. 1999. USEPA's Program to Regulate the
Placement of Waste Water and other Fluids Underground. EPA 810-F-99-019,
December 1999.
US Environmental Protection Agency. 2000. Underground Injection Control (UIC)
Program; Proposed Coal Bed Methane (CBM) Study Design. Federal Register.
Vol. 65, No. 143. p. 45774, July 25, 2000.
US Environmental Protection Agency. 2001. Underground Injection Control; Request
for Information of Ground Water Contamination Incidents Believed To Be Due to
Hydraulic Fracturing of Coalbed Methane Wells. Federal Register. Vol. 66, No.
146. p. 39396, July 30, 2001.
US Environmental Protection Agency. 2002. Underground Injection Control (UIC)
Program; Hydraulic Fracturing of Coalbed Methane (CBM) Wells Report--
Notice. Federal Register. Vol. 67, No. 167. p. 55249, August 28, 2002.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-13
image:
EPA816-R-04-003 Master References
United States Geological Survey. 1971. State of Ohio, 1:500,000 topographic map.
United States Geological Survey. 1973. State of Kentucky, 1:500,000 topographic map.
United States Geological Survey. 1984. National Water Summary. Hydrologic events,
selected water-quality trends, and ground-water resources. United States
Geological Survey Water-Supply Paper No. 2275.
Utah Department of Natural Resources. 2002. Public Comment OW-2001-0002-0090
to "Draft Evaluation of Impacts to Underground Sources of Drinking Water by
Hydraulic Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol.
63, No. 185. p. 33992, September 24, 2002.
VA Division of Gas and Oil, 2002. Public Comment OW-2001-0002-0084 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Virginia Department of Mines, Minerals, and Energy (VDMME). 2001. Personal
communication with VDMME staff.
Valkenburg, Nicholas, Christian, Robert, and Green, Margaret, 1975, Occurrence of iron
bacteria in ground-water supplies of Alabama: Alabama Geological Survey
Circular 96, 45p, as cited by the Alabama State Oil and Gas Board, 2002.
Warpinski, Norm, 2001. Sandia Laboratories. Personal communication, peer-review
panelist.
Warpinski, N.R., Branagan, P.T., Satler, A.R., Cippolla, C.L., Lorenz, J.G., and Thorne,
BJ. 1988. A case study of a stimulation experiment in a fluvial, tight, sandstone
gas reservoir. Society of Petroleum Engineers Paper No. 18258, Proceedings 63rd
Annual Technology Conference, October 1988 (Houston), pp. 616-632.
Warpinski, N.R., Schmidt, R.A., and Northrop, D.A. 1982. In-situ stresses: the
predominant influence on hydraulic fracture containment; Journal of Petroleum
Technology, March 1982, pp. 653-664.
Warpinski, Norman R. 1996. Hydraulic Fracture Diagnostics, Journal of Petroleum
Technology, (Oct. 1996).
Willberg, D.M., N. Steinsberger, R. Hoover, R.J. Card. 1998. Optimization of Fracture
Cleanup Using Flowback Analysis. SPE #39920. Proceedings-SPE Rocky
Mountain Regional/Low Permeability Reservoirs Symposium and Exhibition,
April 5-8, 1998. Publication by Society of Petroleum Engineers, pp. 147-159.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-14
image:
EPA816-R-04-003 Master References
Willberg, D.M., RJ. Card, L.K. Britt, M. Samuel, K.W. England, K.E. Cawiezel, H.
Krus. 1997. Determination of the Effect of Formation Water of Fracture Fluid
Cleanup Through Field Testing in the East Texas Cotton Valley. SPE #38620.
Proceedings-SPE Annual Technical Conference and Exhibition, October 5-8,
1997. Publication by Society of Petroleum Engineers, pp. 531-543.
Wilson, Robert. February, 2001. Director, Virginia Division of Gas & Oil, Department
of Mines, Minerals, and Energy, personal communication.
Winston, R.B. 1990. Vitrinite reflectance of Alabama's bituminous coal; Alabama
Geological Survey Circular 139, 54 pp.
Wright, C.A. 1992. Effective design, real-data analysis, and post-job evaluation of
hydraulic fracturing treatments. Methane from Coal Seams Technology Journal,
pp. 29-32 (July).
Zebrowitz, M. J., Kelafant, J. R., and Boyer, C. M. 1991. Reservoir characterization and
production potential of the coal seams in Northern and Central Appalachian
Basins. Proceedings of the 1991 Coalbed Methane Symposium, The University of
Alabama/Tuscaloosa, May 13-16, 1991.
Zuber, M.D., Kuuskraa, V.A., and Sawyer, W.K. 1990. Optimizing well spacing and
hydraulic fracture design for economic recovery of coalbed methane. SPE
Formation Evaluation, 5(1):98-102.
Zuber, M.D., Reeves, S.R., Jones, A.H., and Schraufnagel, R.A. 1991. Variability in
coalbed-methane well performance: a case study; Journal of Petroleum
Technology, v.43 n.4 (April 1991), pp. 68-475.
Zuber, Michael D. 1998. Production characteristics and reservoir analysis of coalbed
methane reservoirs. Lyons, Paul C. (editor). Appalachian coalbed methane.
International Journal of Coal Geology, 38 (l-2):27-45. Meeting: Appalachian
coalbed methane, Lexington, KY, United States, Sept. 27-30, 1997.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs MR-15
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Appendix A
Department of Energy - Hydraulic Fracturing White Paper
1.0 Introduction
The first hydraulic fracturing treatment was
pumped in 1947 on a gas well operated by Pan
American Petroleum Corporation in the Hugoton
field.1 The Kelpper Well No. 1, located in Grant
County, Kansas was a low productivity well, even
though it had been acidized. The well was
chosen for the first hydraulic fracture stimulation
treatment so that hydraulic fracturing could be
compared directly to acidizing. Since that first
treatment in 1947, hydraulic fracturing has
become a standard treatment for stimulating the
productivity of oil and gas wells.
Hydraulic fracturing is the process of pumping a
fluid into a wellbore at an injection rate that is too
high for the formation to accept in a radial flow
pattern. As the resistance to flow in the
formation increases, the pressure in the wellbore
increases to a value that exceeds the breakdown
pressure of the formation that is open to the
wellbore. Once the formation "breaks-down", a
crack or fracture is formed, and the injected fluid
begins moving down the fracture. In most
formations, a single, vertical fracture is created
that propagates in two directions from the
wellbore. These fracture "wings" are 180° apart,
and are normally assumed to be identical in shape
and size at any point in time. In naturally
fractured or cleated formations, such as gas shales
or coal seams, it is possible that multiple fractures
can be created and propagated during a hydraulic
fracture treatment.
Fluid that does not contain any propping agent,
often called "pad", is injected to create a fracture
that grows up, out and down, and creates a
fracture that is wide enough to accept a propping
agent. The purpose of the propping agent is to
"prop open" the fracture once the pumping
operation ceases, the pressure in the fracture
decreases, and the fracture closes. In deep
reservoirs, we use man-made ceramic beads to
prop open the fracture. In shallow reservoirs,
sand is normally used as the propping agent. The
sand used as a propping agent in shallow
reservoirs, such as coal seams, is mined from
certain quarries in the United States. The silica
sand is a natural product and will not lead to any
environmental concerns that would affect the
United States Drinking Water (USDW).
The purposes of this paper are (1) to discuss the
processes an engineer uses to design and pump a
hydraulic fracture treatment, and (2) to provide an
overview of the theories, design methods and
materials used in a hydraulic fracture treatment.
Currently, a discussion is taking place on the
effects of hydraulic fracturing in coal seams on
the USDW. Gas production from coal seams is
increasing in importance in the United States. In
2000, over 6% of the natural gas production in
the US was produced from coal seams, and that
percentage will increase in the future. Because of
the ever-increasing importance of natural gas
production from coal seams, coal seam examples
have been included in this technical paper.
Objectives of Hydraulic Fracturing
In general, hydraulic fracture treatments are used
to increase the productivity index of a producing
well, or the injectivity index of an injection well.
The productivity index defines the volumes of oil
or gas that can be produced at a given pressure
differential between the reservoir and the well
bore. The injectivity index refers to how much
fluid can be injected into an injection well at a
given pressure differential.
There are many different applications for
hydraulic fracturing, such as:
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-l
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
• Increase the flow rate of oil and/or gas from
low permeability reservoirs,
• Increase the flow rate of oil and/or gas from
wells that have been damaged,
• Connect the natural fractures and/or cleats in
a formation to the wellbore,
• Decrease the pressure drop around the well to
minimize sand production,
• Decrease the pressure drop around the well to
minimize problems with asphaltine and/or
paraffin deposition,
• Increase the area of drainage or the amount of
formation in contact with the wellbore, and
Connect the full vertical extent of a reservoir
to a slanted or horizontal well.
Obviously, there could be other uses of hydraulic
fracturing, but the majority of the treatments are
pumped for these seven reasons.
A low permeability reservoir is one that has a
high resistance to fluid flow. In many
formations, chemical and/or physical processes
alter a reservoir rock over geologic time.
Sometimes, these diagenetic processes restrict the
openings in the rock and reduce the ability of
fluids to flow through the rock. Low
permeability rocks are normally excellent
candidates for stimulation by hydraulic
fracturing.
Regardless of the permeability, a reservoir rock
can be damaged when a well is drilled through
the reservoir and when casing is set and cemented
in place. Damage occurs because drilling and/or
completion fluids leak into the reservoir and plug
up the pores and pore throats. When the pores are
plugged, the permeability is reduced, and the
fluid flow in this damaged portion of the reservoir
may be substantially reduced. Damage can be
severe in naturally fractured reservoirs, like coal
seams. To stimulate damaged reservoirs, a short,
conductive hydraulic fracture is often the desired
solution. As such, hydraulic fracturing works
very well in many damaged, coal seam reservoirs.
In many cases, especially for low permeability
formations, damaged reservoirs and horizontal
wells in a layered reservoir, the well would be
"uneconomic" unless a successful hydraulic
fracture treatment is designed and pumped. Thus,
the engineer in charge of the economic success of
such a well, must (1) design the optimal fracture
treatment, and then (2) go to the field to be
certain the optimal treatment is pumped
successfully.
Candidate Selection
The success or failure of a hydraulic fracture
treatment often depends on the quality of the
candidate well selected for the treatment.
Choosing an excellent candidate for stimulation
often ensures success, while choosing a poor
candidate will normally result in economic
failure. To select the best candidate for
stimulation, the design engineer must consider
many variables. The most critical parameters for
hydraulic fracturing are formation permeability,
the in-situ stress distribution, reservoir fluid
viscosity, skin factor, reservoir pressure, reservoir
depth and the condition of the wellbore. The skin
factor refers to whether the reservoir is already
stimulated or, perhaps is damaged. If the skin
factor is positive, the reservoir is damaged and
could possibly be an excellent candidate for
stimulation.
The best candidate wells for hydraulic fracturing
treatments will have a substantial volume of oil
and gas in place, and will have a need to increase
the productivity index. Such reservoirs will have
(1) a thick pay zone, (2) medium to high pressure,
(3) in-situ stress barriers to minimize vertical
height growth, and (4) either be a low
permeability zone or a zone that has been
damaged (high skin factor). For coalbed methane
reservoirs, the ideal candidate, in addition to the 4
factors listed above, will be a thick coal seam
containing both (1) a large volume of sorbed gas
and (2) abundant coal cleats to provide
permeability.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-2
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Reservoirs that are not good candidates for
hydraulic fracturing are those with little oil or gas
in place due to thin reservoirs, low reservoir
pressure, or small aerial extent. Reservoirs with
extremely low permeability may not produce
enough hydrocarbons to pay all the drilling and
completion costs even if successfully stimulated;
thus, such reservoirs would not be good
candidates for stimulation. In coal seam
reservoirs, the number, thickness and location of
the coal seams must be considered when deciding
if the coals can be completed and stimulated
economically. If the coal seams are too thin or
too scattered up and down the hole, the coals may
not be ideal candidates for stimulation by
hydraulic fracturing.
Developing Data Sets
For most petroleum engineering problems,
developing a complete and accurate data set is
often the most time consuming part of solving the
problem. For hydraulic fracture treatment design,
the data required to run both the fracture design
model and the reservoir simulation model can be
divided into two groups. One group lists the data
that can be "controlled" by the engineer. The
second group reflects data that must be measured
or estimated, but cannot be controlled.
The primary data that can be controlled by the
engineer are the well completion details,
treatment volume, pad volume, injection rate,
fracture fluid viscosity, fracture fluid density,
fluid loss additives, propping agent type, and
propping agent volume. The data that must be
measured or estimated by the design engineer are
formation depth, formation permeability, in-situ
stresses in the pay zone, in-situ stresses in the
surrounding layers, formation modulus, reservoir
pressure, formation porosity, formation
compressibility, and the thickness of the
reservoir. There are actually three (3) thickness
that are important to the design engineer: the
gross thickness of the reservoir; the net thickness
of the oil or gas producing interval; and the
permeable thickness that will accept fluid loss
during the hydraulic fracture treatment.
The most critical data for the design of a fracture
treatment are, roughly in order of importance, (1)
the in-situ stress profile, (2) formation
permeability, (3) fluid loss characteristics, (4)
total fluid volume pumped, (5) propping agent
type and amount, (6) pad volume, (7) fracture
fluid viscosity, (8) injection rate, and (9)
formation modulus. Since most engineers have
more work to do than time to do the work, the
design engineer should focus most of his/her time
on the most important parameters. In hydraulic
fracture treatment design, by far, the two most
important parameters are the in-situ stress profile
and the permeability profile of the zone to be
stimulated and the layers of rock above and
below the target zone.
In new fields or reservoirs, most operating
companies are normally willing to spend money
to run logs, cut cores and run well tests to
determine important factors such as the in-situ
stress and the permeability of the major reservoir
layers. By using such data, along with fracture
treatment records and production records,
accurate data sets for a given reservoir in a given
field can normally be compiled. These data sets
can be used on subsequent wells to optimize the
fracture treatment designs. It is normally not
practical to cut cores and run well tests on every
well. Thus, the data obtained from cores and well
tests must be correlated to log parameters so the
logs on subsequent wells can be used to compile
accurate data sets.
To design a fracture treatment, most engineers
use pseudo 3-dimensional (P3D) models. Full 3-
D models exist; however, the use of full 3-D
models is currently limited to supercomputers and
research organizations. To use a P3D model, the
data must be input by reservoir layer. Fig. 1
illustrates the profiles of important input data
required by a P3D model. For the situation in
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-3
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Fig. 1, the fracture treatment would be initiated in
the sandstone reservoir. The fracture would
typically grow up and down until a barrier is
reached to prevent vertical fracture growth. In
many cases, thick marine shale will be a barrier to
vertical fracture growth. In some cases, coal
seams will prevent fractures from growing
vertically. Many coal seams are highly cleated,
and when the fracture fluid enters the coal seam,
it remains contained within the coal seam. In
thick, highly cleated coal seams, the growth of
the hydraulic fracture will normally be limited to
the coal seam.
Table 1 - Sources of Data
Porosity
(frac)
0.10
0.12
0.18
0.06
0.06
0.10
Shale
Shale
£
— »
Si
GR
(API)
Shale
Sand .
Sand
tstone
Siltstone
Resistivity
_». ^_ (OHMM)
'Thickness
(ft)
200'
"\ 5Q[in.
• ' ^ sot
100'
1001
Perm
(md)
0.0001
0.01
0.03
0.003
0.003
In-situ
Stress
(psi)
7200
6100
6140
6550
6650
Shale
200'
0.0001 7650
Fig. 1 - Typical input data for a PSD model.
The data used to design a fracture treatment can
be obtained from a number of sources, such as
drilling records, completion records, well files,
open hole geophysical logs, cores and core
analyses, well tests, production data, geologic
records, and other public records, such as
publications. In addition, service companies
provide data on their fluids, additives and
propping agents. Table 1 illustrates typical data
needed to design a fracture treatment and possible
sources for the data.
Fracture Treatment Optimization
The goal of every design engineer should be to
design the optimum fracture treatment for each
and every well. In 1978, Hoi ditch et al.2 wrote a
paper concerning the optimization of both the
Data
Formation Permeability
Formation Porosity
Reservoir Pressure
Formation Modulus
Formation
Compressibility
Poisson's Ratio
Formation Depth
In-situ Stress
Formation Temperature
Fracture Toughness
Water Saturation
Net Pay Thickness
Gross Pay Thickness
Formation Lithology
Wellbore Completion
Fracture Fluids
Fracture Proppants
Units
md
%
psi
psi
psi
ft
psi
°F
psi - Vin
%
Ft
Ft
Sources
Cores, Well Tests,
Correlations,
Production Data
Cores, Logs
Well Tests, Well Files,
Regional Data
Cores, Logs,
Correlations
Cores, Logs,
Correlations
Cores, Logs,
Correlations
Logs, Drilling Records
Well Tests, Logs,
Correlations
Logs, Well Tests,
Correlations
Cores, Correlations
Logs, Cores
Logs, Cores
Logs, Cores, Drilling
Records
Cores, Drilling
Records, Logs,
Geologic Records
Well Files, Completion
Prognosis
Service Company
Information
Service Company
Information
3,4
propped fracture length and the drainage area
(well spacing) for low permeability gas
reservoirs. Fig. 2 illustrates the methodology
used to optimize the size of a fracture treatment
Fig. 2 clearly shows the following:
As the propped length of a fracture increases,
the cumulative production will increase, and
the revenue from hydrocarbon sales will
increase,
As the fracture length increases, the
incremental benefit ($ of revenue per foot of
additional propped fracture length) decreases,
As the treatment volume increases, the
propped fracture length increases,
As the fracture length increases, the
incremental cost of each foot of fracture ($ of
cost per foot of additional propped fracture
length) increases, and
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-4
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
• When the incremental cost of the treatment is
compared to the incremental benefit of
increasing the treatment volume, an optimum
propped fracture length can be found for
every situation.
Additional economic calculations can be made to
determine the optimum fracture treatment design.
However, in all cases, the design engineer must
consider the effect of the fracture upon flow rates
and recovery, the cost of the treatment, and the
investment guidelines of the operator of the well.
Field Considerations
After the optimum fracture treatment has been
designed, it must be pumped into the well
successfully. A successful field operations
requires planning, coordination and cooperation
of all parties. Treatment supervision and the use
of quality control measures will improve the
successful application of hydraulic fracturing.
Safety is always the primary concern in the field.
Safety begins with a thorough understanding by
all parties on their duties in the field. A safety
meeting is always held to review the treatment
procedure, establish a chain of command, be sure
everyone knows his/her job responsibilities for
the day, and to establish a plan for emergencies.
The safety meeting should also be used to discuss
the well completion details and the maximum
allowing injection rate and pressures, as well as
the maximum pressures to be held as backup to
an annulus. All casing, tubing, wellheads, valves,
and weak links, such as liner tops, should be
thoroughly tested prior to rigging up the
fracturing equipment. Mechanical failures during
a treatment can be costly and dangerous. All
mechanical problems should be repaired prior to
pumping the fracture treatment.
Prior to pumping the treatment, the engineer-in-
charge should conduct a detailed inventory of all
the equipment and materials on location. The
inventory should be compared to the design and
the prognosis. After the treatment has concluded,
the engineer should conduct another inventory of
all the materials left on location. In most cases,
the difference in the two inventories can be used
to verify what was mixed and pumped into the
wellbore and the hydrocarbon bearing formation.
Cum.
Prod.
L,= 1,500
$
Revenue
Time
Fracture Length
$ Revenue
Less
$ Cost
Treatment
Volume
$
Cost
Fracture Length
Fracture Length Fracture Length
Fig. 2 - Fracture treatment optimization process.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-5
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
In addition to an inventory, samples of the base
fracturing fluid (usually water) should be taken
and analyzed. Typically, a water analysis is done
on the base fluid to determine the minerals
present and the type of bacteria in the water. The
data from the water analysis can be used to select
the additives required to mix the viscous fracture
fluid required to create a wide fracture and to
transport the propping agent into the fracture.
Table 2 shows the typical compositions for mix
waters used in different fracturing situations. In
addition to testing the water, samples of the
additives used during a treatment and the fracture
fluid after all additives have been added should
be taken during the job and saved for future
analyses, if required.
Table 2 - Fracturing Fluids and Conditions
for Their Use
Base Fluid
Water Based
Foam Based
Oil Based
Fluid Type
Linear Fluids
Crosslinked
Fluids
Water Based
Foam
Acid Based Foam
Alcohol Based
Foam
Linear Fluids
Crosslinked
Fluids
Water External
Emulsions
Main
Composition
Gelled Water,
GUAR< HPG,
HEC, CMHPG
Crosslinker +
GUAR, HPG,
CMHPG, CMHEC
Water and
Foamer+ N2orCO2
Acid and Foamer
+ N2
Methanol and
Foamer + N2
Oil, Gelled Oil
Phosphate Ester
Gels
Water + Oil +
Emulsifier
Used For
Short Fractures,
Low Temperatures
Long Fractures,
High Temperatures
Low Pressure Formations
Low Pressures, Water
Sensitive Formations
Low Pressure Formations
With Water Blocking Problems
Water Sensitive Formations,
Short Fractures
Water Sensitive Formations,
Long Fractures
Good For Fluid Loss Control
Formation temperature is one of the main factors
concerning the type of additives required to mix
the optimum fracturing fluid. In deep, hot
reservoirs (>250°F), more additives are required
than in shallow, low temperature reservoirs.
Since most coal seams are very shallow, fewer
additives are normally required to mix the
optimum fracture fluid.
2.0 Fracture Mechanics
Fracture mechanics has been part of mining
engineering and mechanical engineering for
hundreds of years. No one is more interested in
underground rock fractures than a miner working
in an underground mine. In petroleum
engineering, we have only used fracture
mechanics theories in our work for about 50
years. Much of what we use in hydraulic
fracturing theory and design has been developed
by other engineering disciplines many years ago.
However, certain aspects, such as poroelastic
theory, are unique to porous, permeable
underground formations. The most important
parameters are in-situ stress, Poisson's ration, and
Young's modulus.
In-situ Stresses
Underground formations are confined and under
stress. Fig. 3 illustrates the local stress state at
depth for an element of formation. The stresses
can be divided into 3 principal stresses. In Fig. 3,
<Ji is the vertical stress, 02 is the maximum
horizontal stress, while 03 is the minimum
horizontal stress, where <Ji>a2>a3. This is a
typical configuration for coalbed methane
reservoirs. However, depending on geologic
conditions, the vertical stress could also be the
intermediate (02) or minimum stress (03). These
stresses are normally compressive and vary in
magnitude throughout the reservoir, particularly
in the vertical direction (from layer to layer). The
magnitude and direction of the principal stresses
are important because they control the pressure
required to create and propagate a fracture, the
shape and vertical extent of the fracture, the
direction of the fracture, and the stresses trying to
crush and/or embed the propping agent during
production.
A hydraulic fracture will propagate perpendicular
to the minimum principal stress (03). If the
minimum horizontal stress is 03, the fracture will
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-6
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
be vertical and, we can compute the minimum
horizontal stress profile with depth using Eq. 1.
where plate tectonics or other forces increase the
horizontal stresses.
°3
Fig. 3 - Local in-situ stress at depth.
<5m;n =
Where:
a ar
'ext
Eq. 1
V =
Gob =
a =
the minimum horizontal stress (in-situ
stress)
Poissons' ratio
overburden stress
Biot's constant
reservoir fluid pressure or pore
pressure
tectonic stress
Poisson's ratio can be estimated from acoustic log
data or from correlations based upon lithology.
For coal seams, the value of Poisson's ratio will
range from 0.2 - 0.4. The overburden stress can
be computed using density log data. Normally,
the value for overburden pressure is about 1.1 psi
per foot of depth. The reservoir pressure must be
measured or estimated. Biot's constant must be
less than or equal to 1.0 and typically ranges from
0.5 to 1.0. The first two (2) terms on the right
hand side of Eq.l represent the horizontal stress
resulting from the vertical stress and the
poroelastic behavior of the formation. The
tectonic stress term is important in many areas
Poroelastic theory can be used to determine the
minimum horizontal stress in tectonically relaxed
areas.8'9 Poroelastic theory combines the
equations of linear elastic stress-strain theory for
solids with a term that includes the effects of fluid
pressure in the pore space of the reservoir rocks.
The fluid pressure acts equally in all directions as
a stress on the formation material. The "effective
stress" on the rock grains is computed using
linear elastic stress-strain theory. Combining the
two sources of stress results in the total stress on
the formation, which is the stress that must be
exceeded to initiate fracturing.
In many areas, however, the effects of tectonic
activity must be included in the analyses of the
total stresses. To measure the tectonic stresses,
injection tests are conducted to measure the
minimum horizontal stress. The measured stress
is then compared to the stress calculated by the
poroelastic equation to determine the value of the
tectonic contribution.
Basic Rock Mechanics
In addition to the in-situ or minimum horizontal
stress, other rock mechanical properties are
important when designing a hydraulic fracture.
Poisson's ratio is defined as "the ratio of lateral
expansion to longitudinal contraction for a rock
under a uniaxial stress condition".10 The value of
Poisson's ratio is used in Eq. 1 to convert the
effective vertical stress component into an
effective horizontal stress component. The
effective stress is defined as the total stress minus
the pore pressure.
The theory used to compute fracture dimensions
is based upon linear elasticity. To apply this
theory, the modulus of the formation is an
important parameter. Young's modulus is
defined as "the ratio of stress to strain for uniaxial
stress".10 The modulus of a material is a measure
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-7
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
of the stiffness of the material. If the modulus is
large, the material is stiff. In hydraulic fracturing,
a stiff rock will result in more narrow fractures.
If the modulus is low, the fractures will be wider.
The modulus of a rock will be a function of the
lithology, porosity, fluid type, and other
variables. Table 3 illustrates typical ranges for
modulus as a function of lithology.
Table 3. Typical Ranges of Young's Modulus for
Various Lithologies
Litholoqy
Soft Sandstone
Hard Sandstone
Limestone
Coal
Shale
Younq's Modulus
2-5x106psi
6-10x106psi
8-12x106psi
0.1-1 x106psi
1-10x106psi
Because coal is highly cleated, the modulus of the
coal seam in-situ may be very low. In very low
modulus, highly cleated coal seams, it is likely
that most fractures will be wide and short, that is,
not penetrating far into the formation from the
well bore.
Fracture Orientation
A hydraulic fracture will propagate perpendicular
to the least principle stress (Fig. 3). In some
shallow formations the least principal stress is the
overburden stress; thus, the hydraulic fracture
will be horizontal. Nielsen and Hansen published
a paper where horizontal fractures in coal seam
reservoirs were documented u. In reservoirs
deeper than 1000 ft or so, the least principal stress
will likely be horizontal; thus, the hydraulic
fracture will be vertical. The azimuth orientation
of the vertical fracture will depend upon the
azimuth of the minimum and maximum
horizontal stresses. Lacy and Smith provided a
detailed discussion of fracture azimuth in SPE
Monograph 12.12
Injection Tests
The only reliable technique for measuring in-situ
stress is by pumping into a reservoir, creating a
fracture, and measuring the pressure at which the
fracture closes 13. The well tests used to measure
the minimum principal stress are as follows: in-
situ stress tests; step-rate/flow back tests; mini-
fracture tests; and step-down tests. For most
fracture treatments, mini-fracture tests and step-
down tests are pumped ahead of the main fracture
treatment. As such, accurate data are normally
available to calibrate and interpret the pressures
measured during a fracture treatment. In-situ
stress tests and step-rate/flow back tests are not
run on every well. However, it is common to run
such tests in new fields or new reservoirs to help
develop the correlations required to optimize
fracture treatments for subsequent wells.
An in-situ stress test (or micro-frac) can be either
an injection-falloff test or an injection-flow back
test. The in-situ stress test is conducted using
small volumes of fluid (a few barrels), injected at
low injection rates (gals/min), normally using
straddle packers to minimize well bore storage
effects, into a small number of perforations (1-2
ft). The objective is to pump a thin fluid (water
or nitrogen) at a rate barely sufficient to create a
small fracture. Once the fracture is open, then the
pumps are shut down, and the pressure is
recorded and analyzed to determine when the
fracture closes. Thus, fracture closure pressure is
synonymous with in-situ stress and with
minimum horizontal stress. When the pressure in
the fracture is greater than the fracture closure
pressure, the fracture is open. When the pressure
in the fracture decreases below the fracture
closure pressure, the fracture is closed. Fig. 4
illustrates a typical wellbore configuration for
conducting an in-situ stress test. Fig. 5 shows
typical data that are measured. Multiple tests are
conducted to ensure repeatability. The data from
any one of the injection-falloff tests can be
analyzed to determine when the fracture closes.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-8
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Fig. 6 illustrates how one such test can be
analyzed to determine in-situ stress.
Falloff
~T
1
1
J_
:
:
^7
A
•»•*.
, -
H C
• •
• •
•
LU
.-*«•.
• •
n<
***"
--x.
:
:
Electric wireline
SRO gauge
O *" "1
Perforated joint
MRO gauge
r^ £ j. _i i~
MRO gauge
_
--,
, — •
,
• •
• •
•
-ii
.-•v.
• •
>n
s /
A
.***•
--*,
:
:
~p
. I
: 30 ft
_i_
!
:
Fig. 4 - Cased hole test configuration.
Stage 1
Stage 2
Breakdown
c
o
*o
•I
Time
Fig. 5 - Typical stress test pump-in/shut-in.
0123456
SQUARE-ROOT-OF-SHUT-IN-TIME, sqrt(min)
Fig. 6 - Closure pressure analysis.
Mini-fracture tests are run to reconfirm the value
of in-situ stress in the pay zone and to estimate
the fluid loss properties of the fracture fluid. A
mini-fracture test is run using fluid similar to the
fracture fluid that will be used in the main
treatment. Several hundred barrels of fracturing
fluid are normally pumped at fracturing rates. In
coal seams, because the fracture height will
usually be small, the mini-fracture test will often
be eliminated or pumped with only a small
volume of fracturing fluid. The purpose of the
injection is to create a fracture that will be of
similar height to the one created in the main
fracture treatment. After the mini-fracture has
been created, the pumps are shut down and the
pressure decline is monitored. The pressure
decline can be used to estimate the fracture
closure pressure and the total fluid leak-off
coefficient. Data from mini-fracture treatments
can be used to alter the design of the main
fracture treatment if the data determined during
the mini-fracture test is substantially different that
the data used to design the main fracture
treatment.
For an inject!on-falloff test to be conducted
successfully, it is necessary to have a clean
connection between the wellbore and the created
fracture. The purpose of in-situ stress tests and
mini-fracture tests are to determine the pressure
in the fracture when the fracture is open, and the
pressure when the fracture is closed. If there is
excess pressure drop near the wellbore, due to
poor connectivity between the wellbore and the
fracture, the interpretation of in-situ stress test
data can be difficult. In coal seam reservoirs, due
to the highly cleated nature of the coal, multiple
fractures that follow tortuous paths are often
created during injection tests.14 When these
tortuous paths are created, the pressure drop in
the "near-wellbore" region can be very high,
which complicates the analyses of the pressure
falloff data. As such, in-situ stress test data and
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-9
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
data from mini-fracture tests in coal seams are
very difficult to measure and interpret.
The design engineer needs data from well tests to
design the optimum fracture treatment. It is
common for an operator to spend a lot of money
and time running injection tests to determine
values of in-situ stress, formation permeability,
and leak-off coefficient. Fracture treatment
theory is well grounded in science and
engineering and, in most cases, data are collected
from logs, cores and well tests to assure that
designs are as accurate as possible.
3. Fracture Propagation Models
The first fracture treatments were pumped just to
see if a fracture could be created and if sand
could be pumped into the fracture. In 1955,
Howard and Fast15 published the first
mathematical model that an engineer could use to
design a fracture treatment. The Howard and Fast
model assumed the fracture width was constant
everywhere, allowing the engineer to compute
fracture area based upon fracture fluid leakoff
characteristics of the formation and the fracturing
fluid.
2D Fracture Propagation Models
The Howard and Fast model was a two-
dimensional (2D) model. In the following years,
other 2D models were published.16"19 When
using a 2D model, the engineer fixes one of the
dimensions (normally the fracture height), then
calculates the width and length of the fracture.
With experience and accurate data sets, 2D
models can be used with confidence because the
design engineer can accurately estimate the
created fracture height beforehand.
Figs. 7 and 8 illustrate two of the most common
2D models used in fracture treatment design. The
PKN geometry (Fig. 7) is normally used when the
fracture length is much greater than the fracture
height, while the KGD geometry (Fig. 8) is used
if fracture height and length are similar 20. Either
of these two models can be used successfully to
design hydraulic fractures. The key is to use
models to make decisions. The design engineer
must always compare actual results with the
predictions from model calculations. By
"calibrating" the 2D model with field results, the
2D models can be used to make design changes
and improve the success of stimulation
treatments.
Fig. 7 - PKN geometry.
Area of Largest
Flow Resistance
•Approximately Elliptical
Sjiape of Fracture
Fig. 8 - KGD geometry.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-10
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
If the correct value of fracture height is used in a
2D model, the model will give reasonable
estimates of created fracture length and width,
provided, of course, that other parameters, such
as in-situ stress, Young's modulus, formation
permeability and total leakoff coefficient are also
entered correctly. Engineers had to use 2D
models for years due to the lack of computing
power. Today, with high-powered computers
available to most engineers, Pseudo 3-
Dimensional (PSD) models are used by most
fracture design engineers. PSD models are better
than 2D models for most situations because the
PSD model computes the fracture height, width
and length distribution using the data for the pay
zone and all the rock layers above and below the
perforated interval.
3D Fracture Propagation Models
Clifton21 provides a detailed explanation of how
3-Dimensional fracture propagation theory is
used to derive equations for programming 3D
models, as well as PSD models. Figs. 9 and 10
illustrate typical results from a PSD model. PSD
models give more realistic estimates of fracture
geometry and dimensions, which can lead to
better designs and better wells. PSD models are
used to compute the shape of the hydraulic
fracture as well as the dimensions.
s :
".a
— r
— I
S3=£
3
3
J
1-
k
t
I1
6
5
3
2
1
Upper
Height
, r
' L
Lower
Height
1
Layer 1
Fig. 9 - Width from a PSD model.
O)
300
200-
100-
0-
-100-
-200-
-300-
Fig.
Stress
Contrast
(psi)
1200
300
1000
0
250
500 750 1000 1250 1500
Fracture Half-Length
10- Width and height from PSD model.
4. Fracturing Fluids and Additives
To create the fracture, a fluid is pumped into the
wellbore at high rate to increase the pressure in
the wellbore at the perforations to a value greater
than the breakdown pressure of the formation.
The breakdown pressure is generally believed to
be the sum of the in-situ stress and the tensile
strength of the rock. Once the formation is
broken down, and the fracture is created, then the
fracture can be propagated at a pressure called the
fracture propagation pressure. The fracture
propagation pressure is equal to the sum of the in-
situ stress, plus the net pressure drop, plus the
near wellbore pressure drop. The net pressure
drop is equal to the pressure drop down the
fracture due to viscous fluid flow in the fracture.
The near wellbore pressure drop can be a
combination of the pressure drop of the viscous
fluid flowing through the perforations and/or the
pressure drop due to tortuosity between the
wellbore and the propagating fracture. Thus, the
fracturing fluid properties are very important in
the creation and propagation of the fracture.
Properties of a Fracturing Fluid
The ideal fracturing fluid should be compatible
with the formation rock, compatible with the
formation fluid, generate enough pressure drop
down the fracture to create a wide fracture, be
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-11
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
able to transport the propping agent in the
fracture, break back to a low viscosity fluid for
clean up after the treatment, and be cost effective.
The family of fracture fluids available consist of
water base fluids, oil base fluids, acid base fluids
and foam fluids. Table 2 lists the types of
fracturing fluids that are available and the general
use of each type of fluid. For most reservoirs,
water base fluids with appropriate additives will
be the best fluid. In some cases, foam generated
using N2 or CC>2 can be used to successfully
stimulate shallow, low-pressure zones. When
water is used as the base fluid, the water should
be tested for quality. Table 4 presents generally
accepted levels of water quality for use in
hydraulic fracturing.
Table 4 - Acceptable Levels for Mix Water
PH
Iron
Oxidizing Agents
Reducing Agents
Carbonate*
Bicarbonate*
Bacteria
Cleanliness
6-8
< 10 ppm
None
None
< 300 ppm
< 300 ppm
None
Reasonable
*Higher Carbonate/Bicarbonate Content Will
Require Further Pilot Testing on Gel Break,
and Crosslinking
The viscosity of the fracture fluid is important.
The fluid should be viscous enough (normally
50-1000 cp) to create a wide fracture (normally
0.2-1.0 in) and transport the propping agent into
the fracture (normally 10s to 100s of feet). The
density of the fluid is also important. Water
based fluids have densities near 8.4 ppg. Oil base
fluids, although never used to fracture treat coal
seam reservoirs, will have densities that are 70-
80% of the water based fluids. Foam fluids can
have densities that are 50% or less those of water
based fluids. The density affects the surface
injection pressure and the ability of the fluid to
flow back after the treatment. In low pressure
reservoirs, low density fluids, like foam, can be
used to assist in the fluid clean up.
A fundamental equation used in all fracture
models is that the fracture volume is equal to the
total volume of fluid injected minus the volume
of fluid that leaks off into the reservoir. The fluid
efficiency is the percentage of fluid that is still in
the fracture at any point in time, when compared
to the total volume injected at the same point in
time. The concept of fluid loss was used by
Howard and Fast to determine fracture area 15. If
too much fluid leaks off, the fluid has a low
efficiency (say 10-20%) and the created fracture
volume will be only a fraction of the total volume
injected. However, if the fluid efficiency is too
high (say 80-90%), the fracture will not close
rapidly after the treatment. Ideally, a fluid
efficiency between 40-60% will provide an
optimum balance between creating the fracture
and having the fracture close down after the
treatment.
In most low permeability reservoirs, fracture fluid
loss and efficiency is controlled by the formation
permeability. In high permeability formations, a
fluid-loss additive must be added to the fracture
fluid to reduce leak-off and improve fluid
efficiency. In highly cleated coal seams, the leak-
off can be extremely high, with efficiencies down
in the 10-20% range. To fracture treat these
highly cleated coal seams, the treatment must
often be pumped at high injection rates using
fluid loss additives. In general, the objective of
most fracture treatments in coal seams is to create
a short, wide fracture to connect the coal cleat
system to the well bore vs. creating long
hydraulic fractures that penetrate deeply into the
coal seam. Therefore, water with very few
additives, pumped at medium to high injection
rates is commonly used to stimulate coal seam
reservoirs.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-12
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Fracture Fluid Additives
Typical additives for a fracture fluid have been
described in detail by Ely 22. Typical additives
for a water based fluid are briefly described
below.
• Polymers - used to viscosify the fluid
Crosslinkers - used to change the viscous
fluid to a pseudo-plastic fluid
• Biocides - used to kill bacteria in the mix
water
• Buffers - used to control the pH of the
fracture fluid
• Surfactants - used to lower the surface
tension
• Fluid loss additives - used to minimize fluid
leak-off into the formation
• Stabilizers - used to keep the fluid viscous at
high temperature
• Breakers - used to break the polymers and
crosslink sites at low temperature
Additional information on additives is presented
in Table 5.
Table 5 - Summary of Chemical Additives
Type of
Additive
Biocide
Breaker
Buffer
Clay stabilizer
Diverting agent
Fluid loss
additive
Friction reducer
Iron Controller
Surfactant
Gel stabilizer
Function
Performed
Kills bacteria
Reduces fluid
viscosity
Controls the pH
Prevents clay
swelling
Diverts flow of fluid
Improves fluid
efficiently
Reduces the
friction
Keeps iron in
solution
Lowers surface
tension
Reduces thermal
degradation
Typical Products
Gluteridehyde
carbonate
Acid, oxidizer,
enzyme breaker
Sodium bicarb.,
fumaric acid
KCI, NH CL, KCI
substitutes
Ball sealers, rock
salt, flake boric-
acid
Diesel, particulates,
fine sand
Anionic copolymer
Acetic & citric acid
Fluorocarbon,
Nonionic
MEOH, sodium
thiosulphate
The owner of the oil or gas well normally does
not own the equipment or the additives required
to pump a fracture treatment. The operator will
hire a service company to pump the fracture
treatment. Each service company has their own
research department for developing fracture fluids
and additives. Each service company obtains
their additives from various suppliers. As such,
there is no set of rules one can use to select the
proper additives for a fracture fluid, without first
consulting with the service company that will mix
and pump the fluid into the well. Many times,
pilot tests of the fracture fluids must be conducted
to be certain all the additives will work properly
at the temperature in the reservoir and for the
duration of the treatment.
All operating and service companies are
concerned with protecting the environment and
the USDW. As such, research is being conducted
in developing "green additives" to use in
hydraulic fracturing, especially in shallow
formations like coal seam reservoirs. It costs a
lot of money to handle additives and dispose of
fracturing fluids that are either left over after the
treatment or produced back from the well bore.
The development of new, green additives will be
a new technology that will benefit all parties.
5. Propping Agents and Fracture
Conductivity
Propping agents are required to "prop-open" the
fracture once the pumps are shut down and the
fracture begins to close. The ideal propping agent
will be strong, resistant to crushing, resistant to
corrosion, have a low density, and readily
available at low cost.23 The products that best
meet these desired traits are silica sand, resin-
coated sand, and ceramic proppants.
Types of Propping Agents
Silica sand is obtained from sand mining
operations. There are several sources in the
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-13
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
United States and a few outside the US. The sand
must be tested to be sure it has the necessary
compressive strength to be used in any specific
situation. Generally, sand is used to prop open
fractures in shallow formations. For coal seam
reservoirs, sand is usually the best choice for a
propping agent and virtually every fracture
treatment in a coal seam reservoir uses sand.
Sand is much less expensive per pound than the
resin-coated sand or the ceramic proppants.
Resin-coated (epoxy) sand is stronger than sand
and is used where more compressive strength is
required to minimize proppant crushing. Some
resins can be used to form a consolidated sand
pack in the fracture, which will help to eliminate
proppant flow back into the wellbore. Resin
coated sand is more expensive than sand.
Ceramic proppants consist of sintered bauxite,
intermediate strength proppant (ISP), and light
weight proppant (LWP). The strength of the
proppant is proportional to its density. Also, the
higher strength proppants, like sintered bauxite,
cost more than ISP and LWP. Ceramic proppants
are used to stimulate deep (>8,000 ft) wells where
large values of in-situ stresses will apply large
forces on the propping agent.
Factors Affecting Fracture Conductivity
The fracture conductivity is the product of
propped fracture width and the permeability of
the propping agent, as illustrated in Fig. 11. The
permeability of all the propping agents, sand,
resin-coated sand, and the ceramic proppants, will
be 200+ darcies when no stress has been applied
to the propping agent. However, the conductivity
of the fracture will be reduced during the life of
the well because of increasing stress on the
fracture, stress corrosion affecting the proppant
strength, proppant crushing, proppant embedment
into the formation, and damage due to gel residue
or fluid loss additives.
Fracture Conductivity, wkf
wkf = fracture width x fracture permeability
Fracture
Width
f WellV
• Permeability •
• Propped fracture width is primarily a function of
proppant concentration
Fig. 11 - Definition of fracture conductivity.
The effective stress on the propping agent is the
difference between the in-situ stress and the
flowing pressure in the fracture, as illustrated in
Fig. 12. As the well is produced, the effective
stress on the propping agent will normally
increase because the value of the flowing bottom
hole pressure will be decreasing. However, as
can be seen by examining Eq. 1, the in-situ stress
will decrease with time as the reservoir pressure
declines. This phenomenon of decreasing in-situ
stress as the reservoir pressure declines was
proven conclusively by Salz.8 In shallow coal
seam reservoirs, the effective stress on the
propping agent is always low and does not
normally affect the fracture conductivity.
• The stress on proppant (Peff) increases
as the flowing bottomhole pressure
decreases
AP =o — p
off imitu wf
Fig. 12 - Effective stress on proppant.
Fig. 13 illustrates the differences is fracture
conductivity vs. increasing effective stress on the
propping agent for a variety of commonly used
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-14
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
propping agents. The data in Fig. 13 clearly show
that for shallow wells, where the effective stress
is less than 4000 psi, sand can be used to create
high conductivity fractures. As the effective
stress increases to larger and larger values, then
the higher strength, more expensive propping
agents must be used to create a high conductivity
fracture.
total fluid volume pumped, (5) propping agent
type and amount, (6) pad volume, (7) fracture
fluid viscosity, (8) injection rate, and (9)
formation modulus. The two most important
parameters are the in-situ stress profile and the
permeability profile of the zone to be stimulated
and the layers of rock above and below the target
zone.
There is a structured methodology followed by
the engineer to design, optimize, execute,
evaluate and re-optimize the fracture treatments
in any reservoir. The first step is always the
construction of a complete and accurate data set.
Table 1 lists the sources for the data required to
run fracture propagation and reservoir models.
Notice that the design engineer must be capable
of analyzing logs, cores, production data, well
test data, and digging through well files to obtain
all the information needed to design and evaluate
a well that is hydraulically fracture treated.
2,000
4,000 6,000 8,000 10,000 Design Procedures
Effective Stress, psi
Fig. 13 - Effect of stress on conductivity.
6. Fracture Treatment Design
Data Requirements
In Section 1 of this paper, the data required by
the engineer to design a hydraulic fracture
treatment was discussed. The data were divided
into two groups: (1) data that must be measured
or estimated and (2) data that can be controlled by
the design engineer. The primary data that can be
controlled by the engineer are the well
completion details, treatment volume, pad
volume, injection rate, fracture fluid viscosity,
fracture fluid density, fluid loss additives,
propping agent type, and propping agent volume.
As stated earlier, the most important data are (1)
the in-situ stress profile, (2) formation
permeability, (3) fluid loss characteristics, (4)
To design the optimum treatment, the engineer
must determine the effect of fracture length and
fracture conductivity upon the productivity and
the ultimate recovery from the well. As in all
engineering problems, sensitivity runs need to be
made to evaluate uncertainties, such as formation
permeability and drainage area. In coal seam
reservoirs, uncertainties can also exist in variables
such as the gas content and the desorption rate.
The production data obtained from the reservoir
model should be used in an economics model to
determine the optimum fracture length and
conductivity. Then a fracture treatment must be
designed using a P3D fracture propagation model
to achieve the desired length and conductivity at
minimum cost. The most important concept is to
design a fracture using all data and appropriate
models that will result in the optimum economic
benefit to the operator of the well.
A P3D hydraulic fracture propagation model
should be run to determine what needs to be
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-15
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
mixed and pumped into the well to achieve the
optimum values of propped fracture length and
fracture conductivity. The base data set should
be used to make a base case run. Then, the
engineer determines which variables are the most
uncertain. Many times, the values of in-situ
stress, modulus, permeability, fluid loss
coefficient, for example, are not known with
certainty and have to be estimated. The design
engineer acknowledges these uncertainties and
makes sensitivity runs with the PSD model to
determine the effect of these uncertainties on the
design process. As databases are developed, the
number and magnitude of the uncertainties will
diminish.
In effect, the design engineer should fracture treat
the well many times on his or her computer
screen. Making these sensitivity runs will (1)
lead to a better design and (2) educate the design
engineer on how certain variables affect the
ultimate values of both the created and the
propped fracture dimensions. Such designs will
be comprehensive, will consider uncertainties,
and will be developed using professional
processes.
Fracturing Fluid Selection
A critical decision by the design engineer is the
selection of the fracture fluid for the treatment.
Economides et al. 24 developed a flow chart that
can be used to select the category of fracture fluid
on the basis of factors such as reservoir
temperature, reservoir pressure, the expected
value of fracture half-length, and a determination
if the reservoir is water sensitive. Their fluid
selection flow chart for a gas well is presented in
Fig. 14.
Most productive coal seam reservoirs are less
than 5000 ft deep. The permeability in highly
cleated coal seams decreases with increasing
depth and overburden stress. At depths greater
than about 5000 ft, in most cases, the coal seam
does not have enough permeability to be
economically developed.
BorT
X-Linked
Guar/HPG
70-75 Quality
or Low pH
X-Linked
+
25% CO
2
, 1
B, T orZ
X-Linked
Guar/HPG
L
X
2
B,T orZ
X-Linked
HPG/CMHPG
Fig. 14 - Selecting a fracture fluid.
Because most productive coal seams are shallow,
low temperature reservoirs, then the choice of
fracturing fluid (according to Fig. 14) will be (1)
N2 foam for low pressure reservoirs, (2) linear
water based fluids if all you need is a short, low
conductivity fracture, or (3) cross-linked gel if
you need a wide or long fracture. Holditch et
al.14 discussed the criteria for selecting a
fracturing fluid in the Gas Research Institute's
Coal Seam Stimulation Manual.
For thick highly cleated coals, a crosslinked fluid
should be used to create wide fractures and place
as much proppant as possible in the fractures
close to the wellbore. The purpose of the
treatment is to link up the cleats to the wellbore
using the hydraulic fracture and the proppant.
The fluid should use the minimum amount of gel
possible and breaker should be used to minimize
damage to the fracture, and to assist in cleanup.
If the fracture is intended to connect up several
thin coal seams that are vertically scattered up
and down the wellbore, then coil tubing can be
used to selectively stimulate each coal seam. Fig.
15 illustrates how coil tubing can be used to
stimulated multiple intervals, one at a time.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-16
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Single or multiple fracturing stimulation using coiled
tubing as a conduit for both the isolation and the
treatment.
Bottom Hoi
Assembly
Fig. 15 - Fracturing using coil tubing.
In low-pressure coal seams, N2 foam can be used
as the fracture fluid. Foamed fracture fluids will
create wide fractures, can transport the propping
agent, and are easier to clean up than fluids that
do not contain N2.
Propping Agent Selection
Economides et al. 24 also produced a flow chart
for selecting propping agents. Their chart is
included as Fig. 16. Because most productive
coal seams are shallow, sand is always used as
the propping agent. In certain cases, where
proppant flow back becomes a problem, then
resin-coated sand is sometimes used. Special
care must be used to design such treatments,
because at low temperature, it may be difficult to
get the resin to set and to create the consolidated
sand pack needed to prevent proppant flow back.
Fig. 16 - Proppant selection based on closure
pressure.
To determine the optimum fracture conductivity,
the design engineer should use the dimensionless
conductivity (Cr) concept published by Cinco-
Ley 25.
P,KLf
wkf
Eq. 2
where w is the fracture width (ft), kf is the
proppant permeability (md), k is the formation
permeability (md), and Lf is the fracture half-
length. To minimize the pressure drop down the
fracture, the value of Cr should be approximately
equal to ten (10).
For example, in a coal seam, if the formation
permeability is 25 md, and the optimum fracture
half-length is 50 ft, then the optimum fracture
conductivity would be 3,927 md-ft. The engineer
needs to design the treatment to create a fracture
wide enough, and pump proppants at
concentrations high enough to achieve the high
conductivity required to optimize the treatment.
Some engineers tend to compromise fracture
length and conductivity in an often-unsuccessful
attempt to prevent damage to the formation
around the fracture. Hoi ditch26 showed that
substantial damage to the formation around the
fracture can be tolerated as long as the optimum
fracture length and conductivity are achieved.
Ideally, the design engineer can create the
optimum fracture length and conductivity while
minimizing damage to the formation. If the
opposite occurs, that is, the formation is not
damaged, but the fracture is not long enough or
conductive enough, then the well performance
will be disappointing.
The operator of the well should always evaluate
the risks such as mechanical risks, product price
risks and geologic risks. Uncertainties in the
input data can be evaluated by making sensitivity
runs using both the reservoir models and the
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-17
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
fracture propagation models. One of the main
risks in hydraulic fracturing is that the entire
treatment will be pumped and/or paid for (i.e. the
money is spent), but for whatever reason, the well
does not produce at the desired flow rates nor
recovers the expected cumulative recovery.
Many times, mechanical problems with the well
or the surface equipment cause the treatment to
fail. Other times, the reservoir does not respond
as expected.
To evaluate the risk of mechanical or reservoir
problems, the design engineer can use 100% of
the costs on only a fraction of the revenue in the
economic analyses. For example, say one (1) in
every five (5) fracture treatments in a certain
formation is not successful. Then one can use
80% of the expected revenue and 100% of the
expected costs to determine the optimum fracture
length. An illustration of how such an analyses
can alter the desired fracture length is presented
in Fig. 17.
8
Optimal
>
o.
No Risk
Risk
Adjusted
150
300
450
600
Half-length,ft
Fig. 17 - Economic analysis.
Finally, after the optimum, risk adjusted fracture
treatment has been designed, it is extremely
important to be certain the optimum design is
pumped correctly into the well. For this to occur,
the design engineer and the service company
should work together to provide quality control
before, during and after the treatment is pumped.
The best engineers tend to spend sufficient time
in the office to design the treatment correctly,
then go to the field to help supervise the field
operations (or provide on-site advice to the
supervisor).
7. Post-Fracture Well Behavior
The original fracture treatments in the 1950's
were designed to increase well productivity.
These treatments were normally pumped to
remove damage in moderate to high permeability
wells. McGuire and Sikora27 and Prats28
published equations that were used for many
years to design fracture treatments that resulted in
desired folds of increase in the productivity index
of a well. The productivity index of an oil well is
J =
and for a gas well is
J =
Eq. 3
Eq. 4
J is the productivity index in terms of barrels per
psi per day or mcf per psi squared per day. The
viscosity and compressibility are included in the
equation for productivity index of a gas well,
because they are pressure dependent.
Assuming J is the productivity index for a
fractured well at steady state flow, and Jo is the
productivity index of the same well under radial
flow conditions, Prats28 found that
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-18
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
J
Eq. 5
In
0.5
for a well containing an infinite conductivity
fracture whose fracture half-length is Lf. Prats
found that a well with a fracture half-length of
100 ft will produce as if the well had been drilled
with a 100 ft diameter drill bit. In other words,
the hydraulic fracture, if conductive enough, acts
to extend the wellbore and stimulate flow rate
from the well. If the dimensionless fracture
conductivity, Cr (Eq. 2), is equal to 10 or greater,
the hydraulic fracture will essentially act as if it is
an infmately conductive fracture.
In coal seam reservoirs, the gas diffuses through
the coal into the cleat system. If the cleat system
is poorly developed and the permeability of the
coal is low («lmd), then the coal reservoir will
probably not be economic to produce because it is
almost impossible to create long, conductive
fractures in thin coal seams. Thus, most
commercial coal seam reservoirs are highly
cleated, moderate permeability (5md<k<100md)
reservoirs. As such, short, conductive fractures
are required and large volumes of fluids are not
needed to stimulate highly cleated coal seam
reservoirs. The object of a hydraulic fracture in a
highly cleated coal seam is to connect the cleat
system with the well bore using the hydraulic
fracture fluids and proppants.
8.0 Fracture Diagnostics
Fracture diagnostics involves analyzing the data
before, during and after a hydraulic fracture
treatment to determine the shape and dimensions
of both the created and propped fracture.
Fracture diagnostic techniques have been divided
into several groups.29
Group 1 - Direct far field techniques
Direct far field methods are comprised of
tiltmeter fracture mapping and microseismic
fracture mapping techniques. These techniques
require delicate instrumentation that has to be
emplaced in boreholes surrounding and near the
well to be fracture treated. When a hydraulic
fracture is created, the expansion of the fracture
will cause the earth around the fracture to deform.
Tiltmeters can be used to measure the
deformation and to compute the approximate
direction and size of the created fracture. Surface
tiltmeters are placed in shallow holes surrounding
the well to be fracture treated and are best for
determining fracture orientation and approximate
size. Downhole tiltmeters are placed in vertical
wells at depths near the location of the zone to be
fracture treated. As with surface tiltmeters,
downhole tiltmeter data can be analyzed to
determine the orientation and dimensions of the
created fracture, but are most useful for
determining fracture height. Tiltmeters have been
used on an experimental basis to map hydraulic
fractures in coal seams.11
Microseismic fracture mapping relies on using a
downhole receiver array of accelerometers or
geophones to locate microseisms or micro-
earthquakes that are triggered by shear slippage in
natural fractures surrounding the hydraulic
fracture. The principle of microseismic fracture
mapping29 is illustrated in Fig. 18. In essence,
noise is created in a zone surrounding the
hydraulic fracture. Using sensitive arrays of
instruments, the noise can be monitored,
recorded, analyzed and mapped.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-19
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Receiver Detects Ground
Motion From Microseism
V
Elastic
Waves
Emmited
Leakoff
Region
Fig. 18 - Principle of microseismic fracture
mapping.
Tiltmeters have been used extensively in the oil
and gas industry for more than 10 years, although
it has only been recent that the technology has
been available to look at fractures at depths
greater than 4,000ft. Current surface tiltmeter
technology can see below 10,000ft.
Microseismic monitoring has traditionally been
too expensive to be used on anything but research
wells, but its cost has dropped dramatically in the
past few years, so although still expensive (on the
order of $50,000 to $100,000), it is being used
more commonly throughout the industry. As
with all monitoring and data collection
techniques, however, the economics of marginal
wells makes it difficult to justify any extra
expense. If the technology is used at the
beginning of the development of a field, however,
the data and knowledge gained are often used on
subsequent wells, effectively spreading out the
costs.
Group 2 - Direct near-wellbore techniques
Direct near-wellbore techniques are run in the
well that is being fracture treated to locate or
image the portion of fracture that is very near
(inches) the wellbore. Direct near-wellbore
techniques consist of tracer logs, temperature
logging, production logging, borehole image
logging, downhole video logging, and caliper
logging. If a hydraulic fracture intersects the
wellbore, these direct near-wellbore techniques
can be of some benefit in locating the hydraulic
fracture.
However, these near-wellbore techniques are not
unique and can not supply information on the size
or shape of the fracture once the fracture is 2-3
wellbore diameters in distance from the wellbore.
In coal seams, where multiple fractures are likely
to exist, the reliability of these direct near-
wellbore techniques are even more speculative.
As such, very few of these direct near-wellbore
techniques are used on a routine basis to look for
a hydraulic fracture.
Group 3 - Indirect fracture techniques
The indirect fracture techniques consist of
hydraulic fracture modeling of net pressures,
pressure transient test analyses, and production
data analyses. Because the fracture treatment
data and the post-fracture production data are
normally available on every well, the indirect
fracture diagnostic techniques are the most
widely used methods to determine the shape and
dimensions of both the created and the propped
hydraulic fracture.
The fracture treatment data can be analyzed with
a PSD fracture propagation model to determine
the shape and dimensions of the created fracture.
The PSD model is used to history match the
fracturing data, such as injection rates and
injection pressures. Input data, such as the in-situ
stress and permeability in key layers of rock can
be varied (within reason) to achieve a history
match of the field data.
Post-fracture production and pressure data can be
analyzed using a 3D reservoir simulator to
estimate the shape and dimensions of the propped
fracture. Values of formation permeability,
fracture length and fracture conductivity can be
varied in the reservoir model to achieve a history
match of the field data.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-20
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
The main limitation of these indirect techniques is
that the solutions are not very unique and require
as much fixed data as possible. For example, if
the engineer has determined the formation
permeability from a well test or production test
prior to the fracture treatment, so that the value of
formation permeability is known and can be fixed
in the models, the solution concerning values of
fracture length become more unique. Most of the
information in the literature concerning post-
fracture analyses of hydraulic fractures has been
derived from these indirect fracture diagnostic
techniques.
Limitations of fracture diagnostic techniques
Warpinski discussed many of these same fracture
diagnostic techniques.30 Table 6, from
Warpinski's paper, lists certain diagnostic
techniques and their limitations. In general,
fracture diagnostics is expensive and only used in
research wells. Fracture diagnostic techniques do
work and can provide important data when
entering a new area or a new formation.
However, in coal seam wells, where costs must
be minimized to maintain profitability, fracture
diagnostic techniques are rarely used and are
generally cost prohibitive.
Table 6 - Limitations of Fracture Diagnostic
Techniques
Parameter
Fracture
Height
Fracture
Height
Fracture
Height
Fracture
Height
Fracture
Height
Fracture
Height
Fracture
Length
Fracture
Length
Fracture
Length
Fracture
Length
Fracture
Azimuth
Fracture
Azimuth
Fracture
Azimuth
Fracture
Azimuth
Technique
Tracer logs
Temperature
logs
Stress profiling
PSD models
Microseismic
Tiltmeters
PSD models
Well testing
Microseismic
Tiltmeters
Core techniques
Log techniques
Microseismic
Tiltmeters
Limitation
Shallow depth of
investigation; shows height
only near the wellbore
Difficult to interpret; shallow
depth of investigation;
shows height only near
wellbore
Does not measure fracture
directly; must be calibrated
with in-situ stress tests
Does not measure fracture
directly; estimates vary
depending on which model
is used
Optimally requires nearby
offset well; difficult to
interpret; expensive
Difficult to interpret;
expensive and difficult to
conduct in the field
Length inferred, not
measured; estimates vary
greatly depending on which
model is used
Large uncertainties
depending upon
assumptions and lack of
prefracture welltest data
Optimally requires nearby
offset well; difficult to
interpret; expensive
Difficult to interpret;
expensive and difficult to
conduct in the field
Expensive to cut core and
run tests; multiple tests must
be run to assure accuracy
Requires open hole logs to
be run; does not work if
natural fractures are not
present
Analysis intensive;
expensive for determination
of azimuth
Useful only to a depth of
5000 ft; requires access to
large area; expensive
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-21
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
9.0 Nomenclature
CMHPG = Carboxymethylhydroxypropyl-
guar
HEC = Hydroxyelthycellulose
HPG = Hydroxypropylguar
ISIP = Instantaneous shut-in pressure
ISP = Intermediate strength proppant
k = Formation permeability, md
KCL = Potassium chloride
KGD = Kristonovich, Geertsma, Daneshy
Lf = Fracture half-length, ft
LWP = Light weight proppant
MEOH = Methanol
MRO = Memory readout gauge
NFLjCL = Ammonium chloride
PKN = Perkins, Kern, Nordgren
RSC = Resin coated sand
SRO = Surface Readout gauge
wkf = Fracture conductivity, md-ft
a = Biot's constant
v = Poissons' ratio
tfext = Tectonic stress
<3min = Minimum horizontal stress (in-situ
stress)
(Job = Overburden stress
<jp = Reservoir fluid pressure or pore
pressure
<Ji = Vertical (overburden) stress
<J2 = Minimum horizontal stress
as = Maximum horizontal stress
10. References
1. Gidley et al: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 1
2. Holditch, S.A. et al: "The Optimization of Well
Spacing and Fracture Length in Low Permeability
Gas Reservoirs", paper SPE 7496 presented at the
1978 SPE Annual Technical Conference and
Exhibition, Houston, Oct. 1-4.
3. Veatch, R.W., Jr.: "Overview of Current
Hydraulic Fracture Design and Treatment
Technology - Part I", JPT (April 1983) 677-87.
4. Britt, L.K.: "Optimized Oilwell Fracturing of
Moderate-Permeability Reservoirs", paper SPE
14371 presented at the 1985 SPE Annual
Technical Conference and Exhibition, Las Vegas,
Sept. 22-25.
5. Gidley et al.: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 57
6. Hubbart, M.K. and Willis,D.G.: "Mechanics of
Hydraulic Fracturing", Trans., AIME (1957) 210,
153.
7. Whitehead, W. S., Hunt, E. R., and Holditch, S.
A.: "The Effects of Lithology and Reservoir
Pressure on the In-Situ Stresses in the Waskom
(Travis Peak) Field," SPE 16403 presented at the
1987 Low Permeability Reservoir Symposium in
Denver, CO, May 18-19.
8. Salz, L.B.: "Relationship Between Fracture
Propagation Pressure and Pore Pressure", paper
SPE 6870 presented at the 1977 SPE Annual
Technical Conference and Exhibition, Denver,
Oct. 7-12.
9. Veatch, R. W. Jr. and Moschchovidis, Z. A.: "An
Overview of Recent Advances in Hydraulic
Fracturing Technology", paper SPE 14085
presented at the 1986 International Meeting on
Petroleum Engineering, Beijing, March 17-20,
10. Gidley et al.: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 62-63
11. Nielsen, P. E. and Hanson, M. E.: "Analysis and
Implications of Three Fracture Treatments in
Coals at the USX Rock Creek Site Near
Birmingham, Alabama", paper presented at the
1987 Coalbed Methane Symposium, Tuscaloosa,
AL (Nov. 16-19, 1987).
12. Gidley et al.: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 341
13. Gidley et al.: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 58
14. Holditch, S. A., Ely, J. W., and Carter, R. H.:
"Development of a Coal Seam Fracture Design
Manual," paper 8976 presented at the 1989
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-22
image:
EPA816-R-04-003
Appendix A
Hydraulic Fracturing White Paper
Coalbed Methane Symposium in Tuscaloosa, AL,
April 17-20.
15. Howard, G. C. and Fast, C. R.: "Optimum Fluid
Characteristics for Fracture Extension", Drilling
and Production Practice, New York, API, (1957)
24, 261-270 (Appendix by E.D. Carter).
16. Perkins, T.K. and Kern, L.R.: "Widths of
Hydraulic Fractures," JPT, (September 1961) 13,
No. 9, 937-949.
17. Geertsma, J. and de Klerk, F.: "A Rapid Method
of Predicting Width and Extent of Hydraulically
Induced Fractures", JPT, (December 1969) 21,
1571-1581.
18. Nordgren, R.P.: "Propagation of a Vertical
Hydraulic Fracture", SPE Journal, (August 1972)
12, No. 8, 306-314.
19. Daneshy, A. A.: "On the Design of Vertical
Hydraulic Fractures", JPT (January 1973) 83-93;
Trans., AIME, 255.
20. Geertsma, J. and Haafkens, R.: "A Comparison of
the Theories to Predict Width and Extent of
Vertical, Hydraulically Induced Fractures",
Trans., AIME (March 1979) 101, 8.
21. Gidley et al: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 95
22. Gidley et al.: Recent Advances in Hydraulic
Fracturing, SPE Monograph 12, Richardson,
Texas, (1989), 131
23. Holditch, S. A.: "Criteria of Propping Agent
Selection", prepared for the Norton Company
(1979a).
24. Economides, M.J. and Nolte, K.G.: Reservoir
Stimulation, Third Edition, John Wiley & Sons,
LTD, West Sussex, England, 2000.
25. Cinco-Ley, H., Samaniego-V., F. and Dominquez,
N.: "Transient Pressure Behavior for a Well with
a Finite-Conductivity Vertical Fracture", SPE
Journal (August 1978) 18, 253-264.
26. Holditch, S. A.: "Factors Affecting Water
Blocking and Gas Flow From Hydraulically
Fractured Gas Wells," Journal of Petroleum
Technology, (Dec. 1979) pp. 1515-1524.
27. McGuire, W.J. and Sikora, V.T.: "The Effect of
Vertical Fractures on Well Productivity", JPT
(October 1960) 12, 72-74; Trans. AIME (1960)
219,401-403.
28. Prats, M.: "Effect of Vertical Fractures on
Reservoir Behavior-Incompressible Fluid Case",
SPE Journal (June 1961) 1, No. 1, 105-118;
Trans. AIME (1961)222.
29. Cipolla, C. L. and Wright, C. A.: "State-of-the-Art
in Hydraulic Fracture Diagnostics", SPE paper
64434 presented at the SPE Asia Pacific Oil and
Gas Conference held in Brisbane, Australia (Oct.
16-18,2000).
30. Warpinski, Norman R.: "Hydraulic Fracture
Diagnostics", Journal of Petroleum Technology,
(Oct. 1996) pp. 907-910.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. A-23
image:
EPA 816-R-04-003 Appendix B
QA Plan
Appendix B
Quality Assurance Plan:
Evaluation of Impacts to Underground Sources of Drinking Water by
Hydraulic Fracturing of Coalbed Methane Reservoirs
The U.S. Environmental Protection Agency (EPA), bases environmental protection
efforts on the best available scientific information and sound science. The credibility of
the resulting policy decision depends, to a large extent, on the strength of the scientific
evidence on which it is based. Sound science can be described as organized
investigations and observations conducted by qualified personnel using documented
methods and leading to verifiable results and conclusions (SETAC, 1999).
This Quality Assurance Plan for data collection and evaluation describes the procedures
the Agency used for a systematic and well-documented, graded approach to realizing the
goal for the "Evaluation of Impacts to Underground Sources of Drinking Water by
Hydraulic Fracturing of Coalbed Methane Reservoirs." The goal of Phase I of EPA's
hydraulic fracturing study was to assess the potential for contamination of USDWs due to
the injection of hydraulic fracturing fluids into CBM wells and to determine based on
these findings, whether further study is warranted. This Quality Assurance Plan
(developed following the guidelines of EPA publication 240/B-01/003) guides the
production of a set of data and scientific findings that are sound, with conclusions
supported by the data.
1.0 Project Management
This section of the Quality Assurance Plan addresses the basic area of project
management, including the project history and objectives, and roles and responsibilities
of the participants.
1.1 Project and Task Organization
Overall project management was provided by the EPA's Office of Water, Groundwater
and Drinking Water (OGWDW), Groundwater Protection Division. Data was gathered
by an EPA OGWDW contractor.
The contractor compiled the gathered data into a draft summary report, reviewed the draft
report, and submitted the draft report to EPA and other federal agencies for review. After
the contractor addressed comments from EPA and other federal agencies, EPA submitted
the draft report to a Peer Review Panel for their comments (see Table B-l for a list of the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-l
image:
EPA816-R-04-003
Appendix B
QA Plan
members of the Peer Review Panel). Following receipt of comments from the Peer
Review Panel, EPA and its contractors responded to those comments. The availability of
the report for stakeholder review and comment was announced in the Federal Register on
August 28, 2002.
Table B-1: Peer Review Panel
Name
Affiliation
Education
Experience
Morris Bell
Engineer, Colorado
Oil and Conservation
Commission
Engineering
Degree, University
of Oklahoma
Closely involved with coalbed methane
development in the San Juan and Raton
Basins. Has investigated water well
complaints and directed projects to test
water wells. Worked for Amoco as a
production engineer, drilling and
completing tight gas wells. Also worked
as a consultant, specializing in the
completion and evaluation of coalbed
methane wells.
Peter E.
Clark
Associate Professor,
Dept. of Chemical
Engineering and
Material Science,
University of
Alabama
Ph.D., University of
Oklahoma State
University
Specializes in complex fluid flows and
hydraulic fracturing. Has taught several
courses in the Chemical Engineering,
Mineral Engineering, Engineering
Mechanics, and Civil Engineering
Departments. These courses included
fluid mechanics, petroleum rock and
fluids, well completion, drilling, and
natural gas engineering.
David Hill
Manager,
Engineering
Resources, Gas
Technology Institute
(GTI)
MBA, Northwestern
University; BS,
Marietta College,
Petroleum
Engineering
Expertise includes unconventional
reservoirs (e.g., coalbed methane, gas
shales, tight sands); hydraulic fracturing;
and reservoir evaluation in technical,
managerial, and marketing aspects of
technology development, deployment,
and commercialization. Has authored
and co-authored over 40 articles about
oil- and gas-related research and
development, and field-based operations.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. B-2
image:
EPA816-R-04-003
Appendix B
QA Plan
Table B-1: Peer Review Panel
Name
Affiliation
Education
Experience
Buddy
McDaniel
Technical Advisor for
Production
Enhancement
Technology,
Halliburton
B.S., Chemical
Engineering,
University of
Oklahoma
Specializes in applications for highly
deviated and horizontal wellbores and
understanding of reservoir response to
fracturing applications. Conducted
research related to laboratory
measurement of fracture conductivity of
proppants under simulated reservoir
conditions. Was actively involved in
design and application of hydraulic
fracturing treatments in soft chalks,
deviated and horizontal wellbores, gas
storage wells, geothermal wells, and
conventional hydrocarbon reservoirs.
Jon Olson
Asst. Professor,
Dept. of Petroleum
and Geosystems
Engineering,
University of Texas
at Austin
Ph.D., Stanford
University, Applied
Earth Sciences
Worked in the areas of fracture
mechanics and coal geology and has
published several papers on these
subjects. Was employed by Mobil
Exploration for several years as research
engineer in the areas of rock mechanics,
structural geology, and well performance.
Ian Palmer
Senior Petroleum
Engineer, BP Amoco
Ph.D., University of
Adelaide in
Australia
Has worked extensively in coalbed
methane extraction, including fracture
design and prediction, rock mechanisms
of coal, and openhole cavity completions.
Also developed hydraulic fracturing
models.
Norm
Warpinski
Distinguished
Member of Technical
Staff, Sandia
Laboratories
Ph.D., University of
Illinois, Mechanical
Engineering
Authority on hydraulic fracturing,
geomechanics, poroelasticity, in situ
stresses, and production mechanisms.
Has expertise ranging from theoretical
modeling and laboratory testing to field
and in situ mineback experiments.
Serves as project manager and lead
scientist for a program to develop
hydraulic fracture diagnostic technology
for use in industry fracturing applications.
Has published extensively on subject of
hydraulic fracturing.
1.2 Problem Definition and Background
Hydraulic fracturing is a half century-old technology used in oil and natural gas
production. The hydraulic fracturing process uses very high hydraulic pressures to
initiate a fracture. A hydraulically induced fracture acts as a conduit in the rock or coal
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
App. B-3
image:
EPA 816-R-04-003 Appendix B
QA Plan
formation that allows the oil or coalbed methane to travel more freely from the rock pores
(where the oil or methane is trapped) to the production well that can bring it to the
surface.
After a well is drilled into a reservoir rock that contains oil, natural gas, and water, every
effort is made to maximize the production of oil and gas. One way to improve or
maximize the flow of fluids to the well is to connect many pre-existing fractures and flow
pathways in the reservoir rock with a larger fracture. This larger, man-made fracture
starts at the well and extends out into the reservoir rock for as much as several hundred
feet. To create or enlarge fractures, a thick fluid, typically water-based, is pumped into
the coal seam at a gradually increasing rate and pressure. Eventually the coal seam is
unable to accommodate the fracturing fluid as quickly as it is injected. When this occurs,
the pressure is high enough that the coal fractures along existing weaknesses within the
coal. Along with the fracturing fluids, sand (or some other propping agent or "proppant")
is pumped into the fracture so that the fracture remains "propped" open even after the
high fracturing pressures have been released. The resulting proppant-containing fracture
serves as a conduit through which fracturing fluids and groundwater can more easily be
pumped from the coal seam.
To initiate coalbed methane production, groundwater and some of the injected fracturing
fluids are pumped out (or "produced" in the industry terminology) from the fracture
system in the coal seam. As pumping continues, the pressure eventually decreases
enough so that methane desorbs from the coal, flows toward, and is extracted through the
production well.
EPA is conducting a study to assess the potential for contamination of underground
sources of drinking water (USDWs) due hydraulic fracturing fluid injection into coalbed
methane wells. The study focuses on hydraulic fracturing used specifically for enhancing
coalbed methane production. EPA, through its contractors and subcontractors, gathered
information on the hydraulic fracturing process and requested comment from the public
on contamination allegedly due to hydraulic fracturing practices. In this Phase I effort,
EPA did not incorporate new, scientific fact finding, but used existing sources of
information, and consolidated pertinent data in a summary report to serve as the basis for
the study. EPA decided if additional research was required based on the findings from
this effort.
1.3 Project and Task Description
The purpose of this project is to assist EPA in assessing the potential for contamination of
USDWs from the injection of hydraulic fracturing fluids into coalbed methane wells, and
to determine based on these findings if further study is warranted. EPA will use the
information from this study in any regulatory or policy decisions regarding hydraulic
fracturing. The first step in investigating the potential for hydraulic fracturing to affect
the quality of USDWs was to define mechanisms by which contamination could occur.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-4
image:
EPA 816-R-04-003 Appendix B
QA Plan
EPA defined two hypothetical mechanisms by which hydraulic fracturing of coalbed
methane wells could potentially impact USDWs:
1. Direct injection of fracturing fluids into a USDW in which the coal is located,
or injection of fracturing fluids into a coal seam that is already in hydraulic
communication with a USDW (e.g., through a natural fracture system).
2. Creation of a hydraulic connection between the coalbed formation and an
adjacent USDW.
The objective of the project is to consider these two mechanisms, based on existing
literature and data, when evaluating whether hydraulic fracturing fluid injection into
coalbed methane wells could contaminate USDWs.
Information was collected regarding the geology and hydrogeology of the coalbed
methane production regions, the processes used to hydraulically fracture coalbed methane
production wells, and the fluids used in the fracturing process. EPA also evaluated water
supply incidents possibly related to hydraulic fracturing of coalbed methane production
wells. EPA relied on currently available literature and data as the primary source of
information for project efforts.
1.4 Quality Objectives and Criteria
To ensure that findings are valid, the following quality assurance questions will be
addressed for all sources of data:
• What was the purpose of the study?
• Whose data are they?
• What is their source?
• Are the data reliable?
• Is the interpretation biased?
This Quality Assurance Plan establishes a set of guidelines and general approaches to
assess available data and information in a clear, consistent, and explicit manner. Data
collection and review according to this process will make conclusions more transparent,
and thus more readily understood and communicable to stakeholders.
The objectives of the systematic expert review of data and information are transparency,
avoidance of bias, validity, replicability, and comprehensiveness. Following a data and
information review protocol can ensure a common understanding of the task and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-5
image:
EPA 816-R-04-003 Appendix B
QA Plan
adherence to a systematic approach. The components of this Quality Assurance Plan are
as follows:
• Specification of the hypotheses to be addressed;
• Justification of the expertise represented in the expert investigators team;
• Specification of the methods to be used for identification of relevant studies,
assessment of evidence of the individual studies, and interpretation of the
entire body of available evidence (WHO, 2000);
• Review process; and
• Communication of findings.
Revisions to the Quality Assurance Plan may be necessary as new aspects of the task
emerge during the study development process.
1.5 Special Training and Certification
To provide authoritative assessments of data and information, it is important to rely on
expert investigators to evaluate the evidence, draw conclusions on the existence of actual
and/or potential hazard, and estimate the magnitude of the associated risk. The team of
expert investigators, that evaluated the evidence associated with this study, possesses the
following qualifications:
• Formal training in basic scientific principles applicable to the project;
• Basic knowledge of the subject or the body of technical information
pertaining to it;
• Experience in scientific review of technical data and information;
• Ability to use descriptive and analytical tools appropriately;
• Ability to design studies to test hypotheses;
• Ability to communicate results accurately to decision-makers and
stakeholders; and
• Experience coordinating multiple tasks and disciplines to ensure timely and
accurate delivery of study components.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-6
image:
EPA 816-R-04-003 Appendix B
QA Plan
The above-listed qualifications ensure that the project team was able to fulfill the
objectives of this project.
1.6 Documents and Records
Documents produced for the project and submitted to EPA included the draft and final
summary reports (hard copy and digital format). Information and records included in the
data report package following completion of the project included:
• Maps (hard copies);
• Scientific literature (hard copies);
• Books (hard copies);
• Database search results (hard copies);
• Logbooks (hard copies); and
• Site visit notes and photographs (hard copies).
All the above-listed materials are maintained by the EPA OGWDW.
2.0 Data Generation and Acquisition
Processes and methods used to collect the data and information must be clear, explicit,
and based on valid practice. It is important to adhere to a rigorous and thorough
approach to the processes of data collection and data logging.
In Phase I, EPA did not incorporate new, scientific fact finding, but instead used existing
sources of information, and consolidated pertinent data in a summary report to serve as
the basis for the study. EPA decided if additional research is required based on the
findings from this effort. As such, this Quality Assurance Plan does not cover areas of
sampling process design, sampling methods, sample handling and custody, analytical
methods, quality control, instrument/equipment testing, inspection, and maintenance,
instrument/equipment calibration and frequency, and inspection/acceptance of supplies
and consumables.
2.1 Non-Direct Measurements
All information summaries and conclusions developed during the course of this project
were based on non-direct measurements. Available literature and data were used as the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-7
image:
EPA 816-R-04-003 Appendix B
QA Plan
primary source of information for the summary report. An extensive literature search
was conducted using the Engineering Index and GeoRef on-line reference databases.
Searches will be guided by subject topics and key words within the following areas:
• Hydrogeology of the coalbed methane basins;
• Hydraulic fracturing practices;
• Fracture behavior;
• Hydraulic fracturing fluids and additives; and
• Information regarding water quality incidents.
All search results were printed, catalogued, and surveyed for pertinent journal articles,
books and conference proceedings that may contain information meeting the specific data
needs of the summary report. Most pertinent articles were acquired from the University
of Texas Library in Austin, Texas, as this library's holdings include an extensive
collection of oil and gas-related publications. References from the articles were
researched and documents relevant to the study were acquired. All papers collected for
the study were archived by topic for future reference.
To verify facts extracted from the literature, state regulatory agencies, geological surveys,
gas companies, service companies and other relevant organizations were contacted by
telephone. Dated telephone logs were used to document all communications. Personal
conversations with the employees of the various organizations yielded additional
information in the form of literature, figures and maps. These were collected and
referenced in conjunction with literature identified in the literature searches.
Internet-based searches were used to locate additional information. Relevant web sites
were located using various search engines such as Google™, Yahoo®, and Alta Vista®.
More specialized search engines, such as those provided on state geological survey web
sites, also were searched. All relevant web sites were logged and referenced
appropriately. Efforts were made to acquire the most recent literature. EPA offered state
drinking water agencies and the public an opportunity to provide information to EPA on
any impacts to groundwater believed to be associated with hydraulic fracturing by a
request for public comment. Submissions were reviewed by EPA staff for information
pertinent to this report. In addition, a request to provide information and comments
regarding incidents of public and private well impacts that could potentially be associated
with hydraulic fracturing was published in the July 30, 2001 Federal Register (Federal
Register: July 30, 2001; Volume 66; Number 146; Page 39395-39397).
Details on specific methods used to collect information for each of the major report
chapters is included in Chapter 2 of this report.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-8
image:
EPA 816-R-04-003 Appendix B
QA Plan
2.2 Data Management
Gathered information and data was managed to facilitate finding any one piece of
gathered data. To achieve this goal, the following data management procedures were
used:
• All telephone interviews were recorded in labeled log books;
• All scientific literature, published maps, existing water quality data,
conference proceedings, and trade journal articles were filed by coal basin;
• Material safety data sheets and product literature were filed separately;
• Trip folders (to contain notes and photographs) were generated for each site
visit;
• Computer database searches were filed separately; and
• Internet websites were referenced in the summary report.
Most data was stored in hard copy format. Wherever possible, data was stored digitally
on compact disc.
3.0 Assessment and Oversight
The quality assurance review process provides a means to examine if the results and
conclusions are verifiable. The review process results in a determination of whether the
conclusions are directly supported by the data or evidence gathered and can be
independently validated by others. This quality assurance review process is hierarchical
and includes four review levels:
• Weighted emphasis on data based on source;
• Cross referencing of data sources when possible;
• EPA and other federal agencies review; and
• Review by a Peer Review Panel.
EPA's review was accomplished by the Work Assignment Manager in conjunction with
other EPA headquarter offices and with other EPA Underground Injection Control
regional offices involved with coalbed methane or hydraulic fracturing. Other federal
agencies asked to review work products produced by this project, included the United
States Geological Survey and the Department of Energy.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-9
image:
EPA 816-R-04-003 Appendix B
QA Plan
EPA assembled a peer review panel consisting of experts in hydraulic fracturing or
associated subjects. The panelists provided comments to EPA regarding the sources of
data used in the study, the data themselves, and the conclusions drawn from those data.
Comments were requested to assist the investigators in making the study as sound as
possible and to ensure that the study met EPA standards for objectivity, evidence, and
responsiveness to the study charge. Reviewer comments and objections were preserved
and made a part of the record for the study. Issue papers were written containing detailed
explanations of responses to comments and objections. Reasons for proceeding or not
proceeding with the study were clearly explained.
4.0 Data Validation and Usability
This section describes activities that occurred after the initial collection of data. These
activities determined whether or not the gathered data were useful and helpful to the
project.
4.1 Data Review, Verification, and Validation
Subsequent to the data logging process, those reports potentially providing useful
information underwent a selection process to evaluate quality of the information and
usefulness to the study. Systematic evaluation of the validity of individual studies, data,
and information included assessment of the following:
• Source of the data and information;
• Qualitative review of the literature;
• Qualitative review of data and information collected;
• Scientific strength of the data and information;
• Geographical, geological, geochemical, spatial, and temporal relevance;
• Relevance to determining baseline conditions;
• Validity of extrapolation to the scope of the study;
• Characteristics of associations, plausibility, alternative explanations;
• Consistency and specificity of the results;
• Scientific uncertainties, limitations, and confounding variables; and
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-10
image:
EPA 816-R-04-003 Appendix B
QA Plan
• Other evaluation parameters, as appropriate.
A scale or rating of the data and information with respect to a level of proof required to
support conclusions is specifically not proposed as part of this quality assurance process.
Establishing a specific level of scientific evidence required to justify a subsequent
conclusion would generate significant controversy. Instead, expert judgment was used to
evaluate and weigh available data and information.
A variety of technical methods and tools were utilized to sort through the pertinent
information and decipher the meaning of the data. These data analysis methods may
include:
• Quantitative review of selected data and information collected;
• Tabulating valid data and information;
• Constructing geologic cross sections;
• Evaluating current and historical site operations;
• Review of consistencies between studies;
• Review of sources of discrepancies between studies and information; and
• Other methods/tools as appropriate.
All assumptions were explicitly documented, the basis for the use of any models
explained, lack of evidence noted, and scientific uncertainties described as precisely as
possible.
4.2 Reconciliation with User Requirements
This sub-section describes how the gathered and validated data and information were
used to meet the requirements of this project and EPA.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-l 1
image:
EPA 816-R-04-003 Appendix B
QA Plan
4.2.1 Drawing Conclusions
Drawing conclusions from evaluated, analyzed, and summarized data and information
involve judgment as to whether observations are consistent with the study
hypotheses/objectives, or, whether some alternative is suggested. The expert
investigators drew upon all evaluated and appropriately summarized data and
information; however, no checklist or formula was applied to arrive at conclusions.
Instead, critical scientific reasoning and judgment was used to draw conclusions. The
process of scientific reasoning and judgment was made explicit by describing and
documenting how investigators:
• Assessed completeness of data and information;
• Accounted for lack of evidence and limitations, and impacts on the
conclusions;
• Assessed and accounted for bias in original data and/or information;
• Used applicable guidelines and rationales;
• Used any ranges of estimates to arrive at conclusions, where appropriate and;
• Incorporated assumptions into assessments and accounted for the implications
of those assumptions in their conclusions.
Conclusions were drawn within the boundaries of the data and the scope of the study.
Lack or absence of evidence was addressed. The relative strength or weakness of
available information to support conclusions, limitations on where a conclusion may
apply, and alternative interpretations of data, was recognized. Any qualification on the
use of the data and factors that contribute to uncertainty was conveyed.
Much of the information obtained from public response to the Federal Register Notice or
from other sources cannot be confirmed through review of peer-reviewed publications or
other data sources. However, the information was reviewed and contrasted to evaluate
the extent of complaints received and any trends in the complaints within and between
individual coalbed methane production basins.
4.2.2 Communication of Findings
This Quality Assurance Plan is reflected in the communication of scientific findings in a
clear, accurate, and complete manner to interested parties. Investigators communicated:
• The body of technical information that was considered;
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-12
image:
EPA 816-R-04-003 Appendix B
QA Plan
• The manner for evaluating, and drawing conclusions from, collected data and
information; and
• Conclusions that address the hypotheses/objectives, supported by the results
of data evaluation and analysis.
The use of presentation tools such as charts, diagrams, and computer-generated displays
was based on sufficient, valid, and defensible data.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-13
image:
EPA 816-R-04-003 Appendix B
QA Plan
REFERENCES
Breslin, K. 2000. Airing on the Side of Caution or Pulling Standards Out of Thin Air?
Environmental Health Perspectives, 108(4), April 2000.
http://ehpnetl.niehs.nih.gov/members/2000/108-4/spheres.html.
Clay, R. 1999. Still Moving Toward Environmental Justice, Environmental Health
Perspectives, 107(6), June 1999.
http://ehpnetl.niehs.nih.gov/members/1999/107-6/spheres.html.
Manuel, J. 2000. Truth in Numbers, Environmental Health Perspectives, 108(8), August
2000. http://ehpnetl.niehs.nih.gov/docs/2000/108-8/niehsnews.html.
Society of Environmental Toxicology and Chemistry (SETAC). 1999. Sound Science
Technical Issue Paper. Society of Environmental Toxicology and Chemistry
(SET AC), SET AC Press, Pensacola, FL, 1999. http://www.setac.org/sstip.html.
US Environmental Protection Agency. 2000. Strengthening Science at the U.S.
Environmental Protection Agency: Research-Management and Peer-Review
Practices. National Academy Press, Washington, D.C. (2000).
http://www.nap.edu/catalog/9882.html.
US Environmental Protection Agency. 2001. EPA Requirements for Quality Assurance
Project Plans (QA/R-5). EPA/240/B-01/003, March 2001.
http://www.epa.gov/quality/0uality Assurance Planps.html.
World Health Organization (WHO). 2000. Evaluation and Use of Epidemiological
Evidence for Environmental Health Risk Assessment: WHO Guideline
Document. Environmental Health Perspectives, 108(10), October 2000. World
Health Organization European Centre for Environment and Health, Bilthoven
Division, A. van Leeuwenhoeklaan 9, Bilthoven, The Netherlands.
http://ehpnetl.niehs.nih.gov/members/2000/108p997-
1002kryzanowski/ab stracts. html.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs App. B-14
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Attachment 1
The San Juan Basin
The San Juan Basin covers an area of about 7,500 square miles across the Colorado/New
Mexico line in the Four Corners region (Figure Al-1). It measures roughly 100 miles
long in the north-south direction and 90 miles wide. The Continental Divide trends
north-south along the east side of the basin, and land surface elevations within the basin
range from 5,100 feet on the western side to over 8,000 feet in the northern part.
The San Juan Basin is the most productive coalbed methane basin in North America.
Coalbed methane production in the San Juan Basin totaled over 800 billion cubic feet
(Bcf) in 1996 (Stevens et al., 1996). This number rose to 925 Bcf in 2000 (GTI, 2002).
The coals of the Upper Cretaceous Fruitland Formation range from 20 to over 40 feet
thick. Total net thickness of all coalbeds ranges from 20 to over 80 feet throughout the
San Juan Basin. Coalbed methane production occurs primarily in coals of the Fruitland
Formation, but some coalbed methane is trapped within the underlying and adjacent
Pictured Cliffs Sandstone, and many wells are completed in both zones. Coalbed
methane wells in the San Juan Basin range from 550 to 4,000 feet in depth, and about
2,550 wells were operating in 2001 (CO Oil and Gas Conservation Commission and NM
Oil Conservation Division, 2001).
1.1 Basin Geology
The San Juan Basin is a typical asymmetrical, Rocky Mountain basin, with a gently
dipping southern flank and a steeply dipping northern flank (Figure Al-2) (Stone et al.,
1983). The Fruitland Formation is the primary coal-bearing unit of the San Juan Basin
and the target of most coalbed methane production. Geologic cross sections showing
generalized relationships between the Fruitland Formation and adjacent are shown in Al-
4 through Al-6. The Fruitland coals are thick, with individual beds up to 80 feet thick.
The Fruitland Formation is composed of interbedded sandstone, siltstone, shale, and coal.
The stratigraphy of the Fruitland Formation is predictable throughout the basin, as
follows:
• The thickest coalbeds are always found in the lower third of the formation;
• Pictured Cliffs Sandstone occurs immediately below the formation;
• Sandstone content is greater in the lower half; and
• Siltstone and shale predominate in the upper half (Choate et al., 1993).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs Al-1
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
The San Juan Basin may be subdivided into three unique regions, based on similar
geologic, hydrologic, and production characteristics (Figure Al-7). These regions are
denoted as Area 1, Area 2, and Area 3, and are described in more detail below (Kaiser
and Ayers, 1994).
Area 1 consists of the northwestern quarter of the basin. Area 1 is characterized by the
thickest (>20 feet) and highest-rank coal deposits in the San Juan Basin (Ayers et al.,
1994). Most wells produce more than 1,000 cubic feet per day and several wells produce
more than 15,000 cubic feet per day. Almost 90 percent of total methane production
from the Fruitland Formation comes from three fields in a region of Area 1 known as the
"Fairway" (Young et al., 1991; Ayers et al., 1994). Area 1 is an area of active recharge
and in most places is hydrostatically over-pressured (greater than 0.50 pounds per square
inch per foot). Wells in Area 1 usually produce moderate to large volumes of water,
some of which meet the quality criteria of less than 10,000 milligrams per liter (mg/L)
total dissolved solids (TDS) for an underground source of drinking water (USDW)
(Kaiser et al., 1994).
Area 2 (the west-central region of the San Juan Basin) is hydrostatically under-pressured
(0.30 to 0.50 pounds per square inch per foot) and is an area of regional groundwater
discharge. Coalbeds are usually 7 to 15 feet thick, and occur primarily in northwest-
trending belts that extend to the southwestern margin of the basin. Methane production
from wells can be more than 100 thousand cubic feet per day, and a few wells produce
200 to 500 thousand cubic feet per day. Methane gas is produced water-free in this area
as a consequence of the hydrostratigraphy and trapping mechanisms (Kaiser and Ayers,
1994). Additionally, Kaiser and Ayers (1994) suggest that water may be less mobile in
the hydrophilic and low permeability coals. The Fruitland Formation in this area where it
is under-pressured generally shows the presence of saline-type waters (Kaiser et al.,
1994) that most likely have TDS concentrations greater than 10,000 mg/L, which does
not meet the criteria for a USDW.
Area 3, the eastern region of the San Juan Basin, is hydrostatically under-pressured, and
features low permeability and low hydraulic gradient, which suggests slow water
movement within most of the aquifer. Only a few coalbed gas wells are located in this
part of the basin, and they produce up to 8,000 cubic feet of methane per day, with little
or no water content (Kaiser and Ayers, 1994). Produced waters from the Fruitland
Formation in most of Area 3 have a high-salinity, resembling seawater (Kaiser and
Ayers, 1994) in which TDS are too high to meet the water quality criteria of a USDW.
However, along the southern margin of Area 3, TDS concentrations are less than 10,000
mg/L (Kaiser et al., 1994).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-2
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
1.2 Basin Hydrology and USDW Identification
Tertiary sandstones and Quaternary alluvial deposits are present at the surface over much
of the basin interior. These serve as the primary drinking water aquifers in the basin
(Figure A1-2), and produced 55 million gallons per day in 1985 (Wilson, 1986).
Cretaceous sandstones are an important source of water on the basin's periphery (Choate
et al., 1993). The Paleocene Ojo Alamo Sandstone yields as much as 30 gallons per
minute of potable water (Hale et al., 1965) and is mentioned as one of the primary
drinking water aquifers of the region (Brown and Stone, 1979). Cleats and larger
fractures in the Fruitland coals and the presence of interbedded permeable sandstones
make the Fruitland Formation an aquifer and source of drinking water along the northern
margin of the basin where TDS in the groundwater are less than 10,000. In most of Area
1, both the Fruitland Formation and the underlying upper Pictured Cliffs Sandstone act as
a single hydrologic unit (Kaiser et al., 1994). The Fruitland and upper Pictured Cliffs
Sandstone aquifer is underlain and confined by the low-permeability main Pictured Cliffs
Formation and is overlain and partly confined by the Kirtland shale, which is up to 1,000
feet thick in the central basin. Overlying the Kirtland Formation is the Ojo Alamo
Sandstone, (Figures Al-4, Al-5 and Al-6) which has been suggested as a possible source
of groundwater for the municipality of Bloomfield (Stone et al., 1983). At Bloomfield,
the coal and gas bearing Fruitland is separated from the Ojo Alamo aquifer by the
Kirtland shale.
In the northern part of the basin, the Fruitland Formation and the underlying upper
Pictured Cliffs Sandstone can be considered a single hydrogeologic unit on a regional
scale because they exhibit the same hydraulic head and water quality characteristics and
are the source of both the water and gas in the Pictured Cliffs sand tongues (Ayers and
Zellers, 1994; Ayers et al., 1994). At the local scale, however, the two formations appear
to exhibit poor hydraulic continuity, as evidenced by areas of over-pressuring (greater
than 0.5 pounds per square inch per foot), abrupt changes in potentiometric surface
(Figure Al-8), and upward flow (Kaiser et al., 1994). Discrete flow within individual
units here is likely due to pinch out of thick, laterally extensive coal seams and truncation
and offset of the beds by faults.
In general, groundwater is recharged along the Fruitland outcrops at the elevated, west,
northern, and northwestern margins of the basin, and lateral flow converges primarily
from the northeast and southeast toward upward discharge to the San Juan River valley
(Kaiser et al., 1994). In the north, the Fruitland and upper Pictured Cliffs Sandstone
aquifer system is confined by the overlying Kirtland shale, but it is poorly confined by
the Kirtland in the central and southern portions of the basin. Water from the Fruitland
discharges in the western part of the basin and migrates upward across the Kirtland shale
into the Animas and San Juan Rivers (Stone et al., 1983). Generalized groundwater
movement in the Fruitland system is shown in cross-section and plan view in Figures Al-
9 and Al-10 (Kaiser and Swartz, 1988). The results of groundwater flow modeling for
the entire basin (Kaiser et al., 1994) are shown in Figure Al-11.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-3
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
In most of Area 1, the Fruitland system produces water containing less than 10,000 mg/L
TDS, the water quality criteria for a USDW. Groundwater is usually freshest at the
outcrop in recharge areas. The water dissolves salts and mixes with formation water as it
flows, and the groundwater becomes increasingly saline as distance from the recharge
source increases. The presence of low-salinity water at given locations in the San Juan
Basin usually marks close proximity to the recharge source or the most permeable flow
paths and implies a dynamic, active aquifer system (Kaiser et al., 1994). Figure Al-12
shows the chloride concentration of groundwater in the Fruitland Formation, and
indicates that water nearest the northern recharge areas has a low dissolved solids and
chloride content. Kaiser et al. (1994) reported that wells in the northern part of Area 1
produced water containing from 180 to 3,015 mg/L TDS. This was found to be the case
over large portions of Area 1, especially within freshwater plumes resulting from areas of
high permeability or fracture trends (Kaiser and Swartz, 1990; Oldaker, 1991).
Kaiser et al. (1994) conducted a water-quality sampling program in the San Juan Basin.
Analyses taken from Fruitland coal wells in Area 1 show that the majority of wells (16 of
27 wells) produce water containing less than 10,000 mg/L TDS, (Figures Al-13a and Al-
13b), although some nearby wells thought to be in less permeable zones produce water
with higher TDS concentrations up to 23,000 mg/L (Kaiser et al., 1994). The boundary
between waters with more and less than 10,000 TDS has not been published. Another
group of wells throughout the same area was also sampled, but these wells were
completed (constructed) in the adjacent and underlying Pictured Cliffs Sandstone bodies,
which are in hydrologic communication with the Fruitland system (Kaiser et al., 1994).
Although from the above information it would seem that the Fruitland would be
classified a USDW, the following additional information about disposal of brackish water
produced along with the methane would seem to indicate that most of the water in the
Fruitland would not meet the TDS criteria for USDW. Coalbed methane wells in the San
Juan Basin produced from 0 to over 10,500 gallons of water per day, which contain from
less than 300 mg/L TDS to over 25,000 mg/L (Kaiser et al., 1994; Kaiser and Ayers,
1994). Brackish water of various TDS concentrations and brine are produced in the over-
pressured Area 1 of the basin while virtually no water is produced from coalbed methane
wells in Areas 2 and 3 of the basin. Cox (1993) reported "Water disposal in the San Juan
basin is a significant, long-term issue." In 1992, coalbed methane wells produced over 5
million gallons of water per day, and production was expected to increase to over 7.5
million gallons per day by 1995 (Cox, 1993). Produced water is disposed of by means of
evaporation ponds, or, more commonly, by Class II injection into deeper zones such as
the Entrada and Bluff sandstones, Morrison Formation, and Mesa Verde sandstone
(Kaiser and Ayers, 1994). The authors estimated that injection wells cost up to $2
million each and Cox (1993) reported that 51 of them had been constructed in the basin
by 1993.
Area 2 is primarily an area of groundwater discharge. The Fruitland coals and Pictured
Cliffs Sandstone in Area 2 are in hydraulic communication and behave as a single
aquifer. The aquifer is under-pressured (less than 0.50 pounds per square inch per foot),
transmits groundwater from the northeast and southeast, and eventually discharges to the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-4
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Animas and San Juan rivers. The TDS of most samples from Area 2 ranges from 10,000
to 16,000 mg/L (Kaiser et al., 1994).
The Fruitland system in most of Area 3 contains slow-moving water with salinity
approximately equal to that of seawater, greater than 25,000 mg/L TDS, (Kaiser and
Ayers, 1994). In Area 3, the Fruitland and Pictured Cliffs are separate, confined aquifers.
In the southeastern one-third of Area 3, the Kirtland shale is absent because of Tertiary-
age erosion, and the Fruitland and Ojo Alamo Sandstone could be in hydraulic
communication with one another (Figure Al-6). In this area Tertiary rocks, including the
Ojo Alamo, are mapped by the United States Geological Survey (Figure Al-14) as an
aquifer having water with TDSs ranging from 500 to 1,000 mg/L (Lyford, 1979).
At the basin's southern margin in Area 3, downward flow occurs from the Ojo Alamo
through the Kirtland shale to the poorly confined Fruitland aquifer through which it then
moves southward to outcrops at a lower elevation and northward to the San Juan River
Valley (Kaiser et al., 1994) (Figure Al-11). Twenty-four of 26 water samples from the
Fruitland/Pictured Cliffs aquifer system in the south margin of the basin reported by
Kaiser and Swartz (1994) had less than 9,000 mg/L TDS (Figure Al-13e & Al-13f).
Groundwater in the Fruitland Formation at the southern margin of the basin has less than
10,000 mg/L TDS because most recharge there comes from above the Kirtland formation,
rather than from southward throughput from the Fruitland Formation.
1.3 Coalbed Methane Production Activity
Coalbed methane production occurs primarily in coals of the Fruitland Formation.
However, some methane is absorbed in the underlying and adjacent Pictured Cliffs
Sandstone, therefore many wells are completed in both zones. About 2,550 wells were
operating in the San Juan Basin in 2001 (CO Oil and Gas Conservation Commission and
NM Oil Conservation Division, 2001). All wells are vertical wells that range from about
500 to 4,000 feet in depth, and were drilled using water or water-based muds. Almost
every well has been fracture-stimulated, using either conventional hydraulic fracturing in
perforated casing or cavitation cycling in open holes (Palmer et al., 1993b). Total gas
production was 925 Bcf in 2000 (GTI, 2002).
Cavitation cycling is a fracturing method unique to a small area of the north-central San
Juan Basin called the "Sweet Spot," or Fairway, of Area 1 (Figure Al-15). Almost half
of all San Juan wells are located within the Fairway area and utilize open-hole
completions (no casing across the production interval) and cavitation cycling. Cavitation
cycling is used in this area because coals are: 1) very thick (individual coals over 40 feet
thick); 2) hydrostatically over-pressured (0.5 to 0.7 pounds per square inch per foot); and
3) relatively more permeable than the rest of the basin (and coals in other basins) (Palmer
et al., 1993b). This method uses several mechanisms to link the wellbore to the coal
fracture system. Cavitation cycling:
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-5
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
• Creates a physical cavity in the coals of the open-hole section (up to 10 feet in
diameter);
• Propagates a self-propping, vertical, tensile fracture that extends up to 200
feet away from the wellbore (parallel to the direction of least stress); and
• Creates a zone of shear stress-failure that enhances permeability in a direction
perpendicular to the direction of least stress (Palmer et al., 1993a;
Khodaverian and McLennan, 1993) (Figure Al-16).
Cavitation is accomplished by applying pressure to the well using compressed air or
foam, and then abruptly releasing the pressure. The over-pressured coal zones provide a
pressure surge into the wellbore (a "controlled blowout"), and the resulting stress causes
dislodgement of coal chips and carries the chips up the well. These cycles of pressure
and blowdown are repeated many times over a period of hours or days, and the repeated,
alternating stress-shear failure in the coal formation creates effects that extend laterally
from the wellbore (Kahil and Masszi, 1984). The resulting vertical fracture is tensile in
origin, that is, it results from a "pulling" force rather than the compressive forces that
create conventional hydraulic fractures. Because the fracture is tensile in origin, the
height of the fracture does not usually extend out of the target coal seam (Logan et al.,
1989).
Wells outside the Fairway area utilize cased-hole, perforated completions that employ
conventional hydraulic fracturing (Hoiditch, 1990). Palmer et al. (1993a) reported that
hydraulic fracturing in the San Juan Basin uses between 55,000 to 300,000 gallons of
stimulation and fracturing fluids and between 100,000 to 220,000 pounds of sand
proppant. In the San Juan Basin, geologic conditions in conjunction with fracturing
techniques usually produce vertical fractures much longer than they are high, for
example, up to 400 feet radially and less than 150 feet high (e.g., Colorado 32-7 No. 9
well, La Plata County, CO; Mavor et al., 1991). The primary reasons for the controlled
height of San Juan coalbed fractures are the thickness and close spacing of coal seams
(obviating the need for excessive height), and the presence and petro-physical properties
of the overlying Kirtland shale (which prevents inadvertent fracture excursion out of the
Fruitland) (Jeu et al., 1988; Logan et al., 1989; Palmer and Kutas, 1991). Holditch
(1993) reports that where the coal seam is not overlain by shale, hydraulic fractures in the
San Juan Basin can grow into overlying beds.
Fassett (1991) found that coalbed methane could migrate into overlying USDWs near the
northern outcrop, in areas where confining shale layers are absent. Because of these
factors, hydraulic fracturing in the San Juan Basin may indirectly impact overlying
USDWs near the Fruitland outcrop at the basin margins, where USDWs are in closer
proximity and the Kirtland shale may be eroded. Near the northern and northwestern
recharge zones, groundwater usually contains less than 3,000 mg/L TDS (Kaiser et al.,
1994; Cox etal., 1995).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-6
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Fracturing and stimulation fluids utilized in the northern San Juan Basin include (Figure
Al-17 and Table Al-1):
• Hydrochloric acid (12% to 28% HC1);
• Plain water;
• Slick water (water mixed with solvent);
• Linear gels (water and a thickener such as guar-gum or a polymer);
• Cross-linked gels with breakers (gels with additives to prevent fluid leak-off
from the fracture, and "breaker" chemicals to reduce viscosity so that the gel
can be produced back from the well after treatment); and
• Nitrogen and CO2 foam (75 percent gas, 25 percent water or slick water, plus
a foaming agent) since about 1992 (Harper et al., 1985; Jeu et al., 1988;
Holditch et al., 1989; Palmer et al., 1993a; Choate et al., 1993)
Oilfield service companies supply the stimulation fluid used to fracture the well as part of
the service. The chemical composition of many fracturing fluids may be proprietary, and
EPA was unable to find complete chemical analyses of any fracturing fluids in the
literature. Table Al-1 presents some data from the literature concerning the general
chemical makeup of common San Juan fracturing fluids (Economides and Nolte, 1989;
Penny et al., 1991). In addition, most gel fluids utilize a breaker compound (usually
borate or persulfate compounds or an enzyme, at 2 pounds/1,000 gallons) to allow post-
treatment thinning and easier recovery of gels from the fracture (e.g., Jeu et al., 1988;
Palmer et al., 1993a; Pashin and Hinkle, 1997).
Many of the compounds listed in Table Al-1 are quite hazardous in their undiluted form.
However, these compounds are substantially diluted prior to injection. Coalbed methane
development by fracturing, and stimulation in the San Juan Basin are regulated by the
Colorado Oil and Gas Conservation Commission and the New Mexico Oil and Gas
Board. Based on an analysis of current regulations, neither agency regulates the type or
amount of fluids used for fracturing (Colorado State Oil and Gas Board Rules and
Regulations 400-3, 2001; New Mexico Energy, Minerals and Natural Resources
Department, Oil Conservation Division Regulations Title 19, Chapter 15, 2001).
About half of the coalbed methane wells in Area 1 are located in the Fairway zone and
feature "cavitation-cycling" completions (Palmer et al., 1993a) (Figure Al-15).
Therefore, about half of the wells in Area 1 have probably been stimulated using
conventional fracture treatments. Based on the well density of Area 1 in 1990 (Figure
Al-18) compared to the 2001 well population (2,550 wells), it is estimated that between
700 and 1,000 coalbed methane wells have been fracture-stimulated in the USDW of
Area 1.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-7
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
It has been shown that methane can migrate from gas wells into aquifers along the
northern margin of the basin, but this condition was remediated with improved gas well
construction (Cox et al., 1995). In addition, wells completed in other aquifers in the
outcrop area have been shown to produce water chemically and isotopically similar to
Fruitland wells, implying communication between the formations (Cox et al., 1995).
1.4 Summary
Coalbed methane development and hydraulic fracturing in some of the northern portions
of the San Juan Basin take place within a USDW. The waters of the Fruitland-upper
Picture Cliffs aquifer and producing zone in Area 1 usually contain less than 10,000
mg/L TDS. Most waters in the northern half of Area 1 contain less than 3,000 mg/L, and
wells near the outcrop produce water that contains less than 500 mg/L.
Each fracture stimulation treatment may inject, on average, approximately 55,000 to
300,000 gallons of stimulation and fracturing fluid per treatment. There are no state
controls on the type, composition, or volume of fracturing fluid employed in each well or
treatment. In contrast to conventional gas formations, the anisotropic nature of fracture
permeability, the volume of treatment fluids employed, and the height and proppant
distribution in coalbed fractures may prevent the effective recovery of fracturing fluids
during subsequent production.
The potential for fracturing to cause or allow degradation of water in aquifers adjacent to
the producing zones seems relatively remote in the currently active gas producing fields,
but the potential for such degradation varies in different parts of the basin. It has been
shown that methane can migrate from gas wells into aquifers along the northern margin
of the basin, but this condition was corrected with improved gas well construction. There
is little potential for fracturing to create communication between the Fruitland-upper
Picture Cliffs aquifer and the Ojo Alamo aquifer over much of the basin because the
aquifers are separated by the poorly permeable Kirkland shale. However, the Kirkland
varies greatly in thickness and forms a leaky hydrogeologic barrier when it is thinner. In
the eastern part of the basin, the Kirkland Formation has been eroded and the Ojo Alamo
lies disconformably and directly upon the Fruitland Formation, potentially allowing
fracturing to cause hydraulic communication between the saline waters of the Fruitland
and the fresh waters (500 to 1,000 mg/L) of the Ojo Alamo.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-8
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
Table Al-1. Chemical Components of Typical Fracture/Stimulation Fluids Used
for San Juan Coalbed Methane Wells
Type of
Stimulation Fluid
Hydrochloric acid
"Slick" water
Diesel oil
Nitrogen and CC>2 foam
Gels1
R-F
Pfizer Flocon 4800
Marathon MARCH
DuPont LuDox SM
CPAM crosslinked with
Pfizer Floperm 500
Drilling Specialties
HE-100 Crosslinked
Composition
12% to 28% HC1 water solution
miscible or immiscible solvent as
viscosity reducer (% unknown)
NA
75 % gas, 25 % water or slick water,
plus a foaming agent)
3% resorcinol, 3% formaldehyde,
0.5% KC1, 0.4% NaHCO3
0.4% xanthan, 154 ppm Cr3+
(asCrC!3), 0.5%KC1
1.4% polyacrylamide (HPAM), 636 ppm
Cr3+ (as acetate), 1% NaCl
10% colloidal silica, 0.7% NaCl
0.4% cationic polyacrylamide (CPAM),
1520 ppm glyoxal 2% KC1
pH
NA
NA
NA
0.3% HP AM-AMPS, 100 ppm Cr
(as acetate), 2% KC1
3+
6.5
4.0
6.0
8.2
7.3
5.0
Dowell YF-230
Hydroxypropylguar (HPG) x-linked
with borate, persulfate with amine
NA
1 Gels are typically mixed at a ratio of 40 Ibs. per 1000 gal. water; compositions shown are "as
mixed".
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-9
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
IO«»
Regional Tectonic Setting of the San Juan Basin
(Laubach & Tremain, 1994)
7/14/01 kl, 1027-SJ-f o
Figure A1-1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-10
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ?••
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. c^
C3 O
OQ g
o ri
^+1 rt
-"3.O-&O-1
Generalized Hydrogeologic Cross-Section of the San Juan Basin (Stone etal., 1983)
FipureAI-2
oo
K^
ON
H
"
>
I
td
image:
EL a EL
C? 3.
ft B
P. R-
& 35 *>
111
2. B- C
a i a
fj- fO
0 OQ
^ S
3 §
o »
3.
3
OQ
o
s
•
o
QJ
T3
S?Q
S -41
<* C
CD »
3 Co
So.
rt -^ __*
°- -* QJ
•> s S*
3 3 5'
s- 3- 3
O.
C
fi
r 5
> a
o
CQ
a
en
oo
K^
ON
H
>
I
td
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
li°
g p? y
ft ?t ng
«^ ^
2. B- C
a i a
-
0 OQ
^ 3
3 §
o id
3. oo
C3 O
OQ g
o ri
«5 rt
Area 1
O6 3M *S E75
i i
• Area 3
•'•> &-
? s'v-t ,
J ^ t t*,.
•-• «• ^ : •-> ii. •-.,•'.
1 | ) / .^ f I ) "^ip?f^T>
V6" 2345
J I I i
EXPLANATION
.1 Ba«4sK>r* " ^-~
/*•
Zslffl ls?>5)iiU,
1jrte C rffss S^a'i^iiDf^ * ^tcgrtf *-jj
.
/OMI »«
^ / s ,
t
—" >SiS•r->--.^ii*«*h,\,.;r:, •'_•'• ••^,t41_i-iJ_ii,' * • I3qr^l-if «>^>**.—^
^ti***|=^f-f • -->?-si- y -J|;^r^-'4«^^ij:iC;"^
s;:
TO »3
5w ^^
Cross Section S-10 from Figure A1-3, a Stratigraphic Strike Section (Ayers and Ambrose, 1990}
oo
K^
ON
H
>
I
td
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ~-
en ^ en
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. oo
C3 O
OQ g
o ri
«5 rt
'.«_-. *. ;,__- ;- i, .•£. —:> r *,_; / /_ -, *.__?_ j.,,, J.WTM/, ^ ^
•, ^- ""-j ~:'-- j' V-—SL*—"•**—-*i"«5" - ~ -% ";'- r;- 4^"^**=is***f"*lll**l'"*«*—•#
. ...i.TV^'Tkia' **Tl!!^£S^BW.« -,*s£^(^rf& =s«^sr""JT1^r^J^>s^^&3"^^SSSd?M--^;"r"- J^^^laT'^'CjyS'fe*"*?lL,'3wf'*"^TJS£*4i..'i^'«3™^?W»«f*!5v-' Vv5*,-**»f "^*Z&" ~ *,-&
^^^^^^S^^^^^^^^^K^^-^-^^. "
> \ ™* *^"—^C ~ ™- * ^ff— >* "*•? - . 1 ' *O- »'* "" t i*1 * **""". '* *™*™^- •*
i * ' i "n" 4T™* *E!%^j "*"»./ V " f - ! , ^ * » * *, I ^ '
""£ '" " ' ~ i ' * f ~ •"^°"I:TU ""J* « "* * ''"T ^ ,$*~*^ r"*. * ~" "* - \ \ I
"^ '" '• i ', •' "'^---""r-c,'-. ' < " '-- '' -§-»i>;«r^ { ,' " , « %
\'- < i- i ' i l*-*"^-*! .' ^ "" '• 'f<-r» — •• '"
fXPLAK*TK)M
^Vr'*J Qji di^mc ^n^Kiur*
[.•;,./,] uapB» P'c>jr«a curd
l&'ITl f"le'. nt C-llffs Sa^^ll
__
ifl-.r*fidy.el!^B^ tfi*«F^'i? f ^gwB Siifl^t ^J
V--". '; < '
~rrc":V~"v>-.,-
.--^~-f ,
I
Cross Section D-20 torn Figure A1-3, a Stratigraphic Dip
Section from the Fruttland to the Ojo Alamo (Ayers and Ambrose, 1990)
vise; a a« F A1 -5
oo
K^
ON
H
>
I
td
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ?••
en ^ en
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. oo
C3 O
OQ g
o ri
«5 rt
Wtsl
Area 3
Emt
:£•
•i* -.- : '
Bst.
We'-
tt In
E-W from Figure A1-3, a Stratigraphic Strike Section from the Southeastern San
Showing the Erostonai Fruitland-Ojo Alamo Contact
The Kirtland Shale has been erade-d and the Ojo Alart'o rests uncontjrrt'flWyon the Fruitland
(Ayers and Ambrose, 1990)
oo
K^
ON
H
>
I
td
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
\ f La Plata
«Dyrango 1
Montezyma
Overpressure
.1.0)
Underpressure
Areas of the San Juan Basin that Exhibit Similar Characteristics for
Production, Coal Properties, and Hydrologic Pressure
(New Mexico Bureau of Mines and Minerals, 1993)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-16
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
L
(Sato point
CCMLito
of the of the at.,
&
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-17
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
en ^ en
o p. c!
5ii
"'aq
cf |
3. en
Oq g
O O
^+J ft
C/l
£ |
K- O
OO -1^
Scuft
North
EXPLANATION
huh k fccie-s
General Ground Water Flow in the Fruitland/Pictured Cliffs Acpffer System, San Juan Basin
(Kaiser and Swartz, 1988)
iCI?SJ.»,n>r F(gUFBA1-8
OO
K^
ON
H
>
I
td
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
SOT^M? $-3CC#i'>*< C.iri ri^rntie'
lrrg/L!
*'k» S
\
TidA* )-MI ztr.e
Generalized Flow Paths of the Fruitland/Pictured Cliffs Aquifer System, San Juan Basin
(Kaiser et al., 1994)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-19
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
li°
g p? y
ft ?t ng
«^ ^
2. B- C
a i a
-
0 OQ
^ 3
3 §
o id
3. oo
C3 O
OQ g
o ri
«5 rt
1
K»
O
O
South
North
w*if •' * ' "L**, l^IiHSiwlL' '*/ i ^Ji^'f'Cr'rii'^ ' fjir
' -sri* ,' >„ aT *r>?Br y"-™.1*^ j.^*-1"!-^.i-Ti^*^ *'aLt<ir ..,*
Equipotentials and Flow Paths from Ground Water Flow Modeling of the San Juan Basin
(Kaiser at al., 1994)
.ittar Rgure A1 -11
oo
K^
ON
H
>
I
td
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
Ijn'crred irorti htat not
m Ssji*&<!'. Jtt-Mob.': 36-i
• 53-jtherr, U?« 1^24
^?1*
Mf KsCO
Chloride Concentration Map (rng/L)of Waters of the Fmitland Aquifer, San Juan Basin
(Kaiser and Swartz, 1988)
*
Figure A1-12
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-21
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
£
it
&• "•
i%
%
n
n_ 4
O
i
1
11
a
B
D *
1 *
0
fiiiiyid Okie ' i: ' i
NoTlf Hsr.tra Das-- j
« i i » ,
j B! n*?7 f | c
^ : :• .2=-
1 ! E
| ,j ^ : ? '
5 f-! --t .-"-^ JS
P ' * * *' ' E
1 .,j 1- : ^^ ^^^^.^ » : a 9
I b: I ; [I fn :: j ; SB '
.;" i i Tjl i p H [TH
S 5tW B55£ 'S-SSS 1S.CJS IB.SW 31 .ISM "».flC5 C
TBS <mfl'k).
. ill
NfifSfi «:*w!irrij b^A-irt
... s^
r,,1to IS *
i ^
p— . : -
S r : es
: 1 : e *
I - i
1
1=5—, ; § z
! i z
"n .- . -in
TDS <mg.'L)
; Seutti «wjm (ei
: ' * *"
1^
p>«^| v
41 *"
i I
! ! «
«—*.«.!, .. rfl-H.-H.".-!| SJ
•^ ! l_ E4
H Th m : 3
• t | . . J [ . i W4- ... 1 : i
: ^ » 27 ;
: ~~j :
r™1 r^i ^i r*i
.Him h
< istt 5000 iso: 4SK sco:-
„ . :
•• No-lN-cmriital ba«.-f: ''' :
i o - -5
I i( . ^
j f 5^5 ?^CfC 9X1 R *MO -K'Sl
CI img:L>
:
"I 5QU3I Iflfgil '.11;
I h n n ci j J
TBS (m|.'i4 Ci Cmfl.'M
Histograms of Water Analyses (mg/L) from the Fruitland/Pictured Cliffs Aquifer System in the
North Central and South-Margin Areas of the San Juan Basin
(Kaiser and Swartz. 1994)
w«c r-..,:a.'-s,- ;.,.. Figure A 1 13
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-22
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
15*
//v -
^X/5 Outnros of
tijxj Ttrtlsff feet
:<s>olved-sc't \at
ft i re c t i a-A '0 f
.. _ _ _ 7
0 f It II WKJifliC ffCWt.
i i i i C.H. Oane a.Kl S.C. laehsnan
(tJSS) a«d 0. Twats (19/6;
Direction of Ground Water Flow and Dissolved Solids Concentration
in Tertiary Rocks (Lyford, 1979)
i*
«,a» FipjrsA1-14
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-23
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
San Juan Co. Rio Arriba Co
Red dats
Outline of the Fairway Zone of Area 1 of the San Juan Basin
i we:5s ds-og ooventoia f'actjrig treatments, aid cnnpty dots 'ep'esent cavlat'oT-sjc-ng oonnp:Bicns
(Palmer et al., 1193)
*
FjgureA1-15
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-24
image:
EL a EL
C? 3.
ft B
P. R-
oo
K^
ON
& 35 *>
111
2. B- C
a i a
ti- rt>
0 OQ
^ 3
3 §
o id
-
3.
s
OQ
o
cr,
\
Fracture
of
1
K<
O
O
H
Conceptual Schematic (Plan View) of Tensile Fracture and Shear in Formed by
Gpenhofe Cavitation Cycling
(Khcdaverian and McLennan, 1993)
>
I
A1-16
td
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
S'linked Gelled Sand Fr.ic
Woler volmr.o
SanJ volume
'[ vrv fluid
T \rvsor.d
Slick Water Sind Fwe
TV PL Hind
'f '/Ps* !->.lLHl
A\ '.,;. iiut'Ctiii'O tote
Miiv sip-d concert rnluui
Fr«if (gelled watrr'i
'.Vatt-r \\'*]f;rr«;-
3\«d vulume
t r-si
Foam Frac (slick water)
W.Mor loUime
Hand %\>li;me
Tvpc Ikiid
23UU g.il/ni of piy
^v.tii Iht;;-'t of pav
>is bo rote cm?---Imki'd :.;e;
J:i/7l'mesli n>l )l):<- ot job i.jr tlnisi
12/20 :iv?b ithortMrtt.'!-.'
u-ju g,tl/?t ,»f pav
HUG lbs/!l of pav
>•!'« hiw.itt-1 ctii*,N-i]]iki\l gel
UO ir.osV- Jl-t 14'"; nl johtor^'lujJ
27'- of total fluid
SmuM-idli'M
2or>0 gal/ft of pay
?!!u(i )b«./tt c-f pav
fu'sb ivo'.rr w./ tricNfi'i redii-,XT
4.1.'7t'! T110'.J1 I 1st 10'"- O(' ;Ob !\>I
2J ••''4'1 iriv'-li itfu'rvarler)
Table of Fracture Stimulation Treatments in the Fruitland Formation
of the San Juan Basin (Palmeret aL 1993)
Rgure A1-17
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-26
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ?••
en ^ en
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. oo
C3 O
OQ g
o ri
^+1 rt
1
K»
O
O
flt ja^ : w^a * «u»vw *-* ^^
Density of Welts in the Northern Portion of Area 1 in the San Juan Basin, as of 12/31/1990
(New Mexico Bureau of Mines, 1993}
A1-1B
oo
K^
ON
H
"
>
I
td
image:
EPA816-R-04-003
Attachment 1
The San Juan Basin
Fruitland Net Coal Map.
Note: Crass Sections D20 ar-di S10 are Shown r; Figures A1-£
(Ayers and Ambrose, 1990)
FiguraA1-19
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
Al-28
image:
EL a EL
C? 3.
ft B
P. R-
& ^ ^
111
2. B- C
a i a
fj- fO
0 OQ
^ 3
3 §
o id
3.
s
OQ
o
1
K<
O
O
\
.— — — ^
^ I I X*
/ L v
L = fracture >•
t L = 1SQto30Qfeet
Leakoff
«*"-"W"*l* g^'1 i^-LmiiH^ \ 1 — — — — " ™- '
==•
ill \^'" I I 1 *
[
\ D = 50to1O3ft
•s^ ^
*
)
/
p
mm
/
-— * , » —
Plan View of a Vertical, Two- Winged Goalbed Methane Fracture Showing the Reservoir
by Fracturing Fluid Leakoff (Palmer et a!., 1991),
7114^1 * 1M:T
\
\
I
Reg ton Invaded
aj.no,
A1- 20
EPA 816-R-04-003 Attachment 1
The San Juan Basin
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
REFERENCES
AAPG = American Association of Petroleum Geologists
SPE = Society of Petroleum Engineers
Ayers, W.B., Ambrose, W.A., and Yeh, J.S. 1994. Coalbed methane in the Fruitland
Formation - depositional and structural controls on occurrence and resources.
New Mexico Bureau of Mines and Minerals Bulletin 146: Coalbed methane in the
upper Cretaceous Fruitland Formation, San Juan Basin, New Mexico and
Colorado, pp. 13-40.
Ayers, W.B. and Zellers. 1994. Coalbed methane in the Fruitland Formation, Navajo
Lake area - geologic controls on occurrence and producibility. New Mexico
Bureau of Mines and Minerals Bulletin 146: Coalbed methane in the upper
Cretaceous Fruitland Formation, San Juan Basin, New Mexico and Colorado, pp.
63-86.
Brown, D.R. and Stone, W.J. 1979. Hydrogeology of the Aztec quadrangle, San Juan
county, New Mexico. New Mexico Bureau of Mines and Mineral Resources
(Sheet 1).
Choate, R., Lent, T., and Rightmire, C.T. 1993. Upper Cretaceous geology, coal, and the
potential for methane recovery from coalbeds in the San Juan Basin - Colorado
and New Mexico. AAPG Studies in Geology, 38:185-222.
Colorado State Oil and Gas Board Rules and Regulations 400-3, 2001.
Colorado Oil and Gas Conservation Commission and New Mexico Oil Conservation
Division. 2001. Personal communication with staff.
Cox, D.O. 1993. Coal-seam water production and disposal, San Juan Basin. Quarterly
Review of Methane from Coal Seams Technology, 11(2): 26-30 (December).
Cox, D.O., Young, G.B.C., and Bell, M.J. 1995. Well testing in coalbed methane (cbm)
wells: an environmental remediation case history. Society of Petroleum
Engineers Paper No. 30578, Proceedings 1995 SPE Technical Conference
(Dallas), pp. 467-500.
Economides, M.J. andNolte, K.G. 1989. Reservoir Stimulation, Second Edition,
Prentice-Hall, New Jersey.
Fassett, J.E. 1991. The mystery of the escaping gas: forensic geology in the northern
San Juan Basin, La Plata County, Colorado. USGS - AAPG Association
Roundtable, p. 1223.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs Al-30
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Gas Technology Institute (GTI) Web site, 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Hale, W.E., Reiland, L.J., and Beverage, J.P. 1965. Characteristics of the water supply
in New Mexico. New Mexico State Engineer, Technical Report 31.
Harper, T.R, Hagans, J.T., and Martins, J.P. 1985. Fracturing without proppant. SPE
13858, Proceedings SPE Low Permeability Reservoirs Symposium (Denver), p.
83.
Holditch, S.A., Ely, J.W., Semmelbeck, M.E., Carter, R.H., Hinkle, 1, and Jeffrey, R.G
1989. Enhanced recovery of coalbed methane through hydraulic fracturing. SPE
Paper 18250, Proceedings 1988 SPE Annual Technical Conference and
Exhibition (Production Operations and Engineering), p. 689.
Holditch, S.A. 1990. Completion methods in coal seam reservoirs. SPE 20670,
Proceedings 65th SPE Annual Technical Conference (New Orleans), p. 533.
Holditch, S.A. 1993. Completion methods in coal-seam reservoirs. Journal of Petroleum
Technology, 45(3): 270-276 (March).
Jeu, S.J., Logan, T.L., and McBane, R.A. 1988. Exploitation of deeply buried coalbed
methane using different hydraulic fracturing techniques. SPE Paper 18253,
Proceedings 63rd Annual Technical Conference (Houston).
Kahil, A. and Masszi, D. 1984. Cavity stress-relief method to stimulate demethanation
boreholes. SPE Paper No. 12843, Proceedings 1984 SPE Unconventional Gas
Recovery Symposium (Pittsburg).
Kaiser, W.R. and Swartz, T.E. 1988. Hydrology of the Fruitland Formation and coalbed
methane producibility, In Geologic evaluation of critical production parameters
for coalbed methane resources, Part 1: San Juan Basin. Annual Report to the Gas
Research Institute, GRI-88/0332.1, pp. 61-81.
Kaiser, W.R. and Swartz, T.E. 1990. Hydrodynamics of the Fruitland Formation. In
Geologic Evaluation of critical production parameters for coalbed methane
resources, Part 1: San Juan Basin. Annual Report for 1990, Gas Research
Institute, GRI-90/0014.1, pp. 99-126.
Kaiser, W.R. and Ayers, W.B. Jr. 1994. Coalbed methane production, Fruitland
Formation, San Juan Basin: geologic and hydrologic controls. New Mexico
Bureau of Mines and Minerals Bulletin 146: Coalbed methane in the upper
Cretaceous Fruitland Formation, San Juan Basin, New Mexico and Colorado, pp.
187-207.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-31
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Kaiser, W.R., Swartz, T.E., and Hawkins, GJ. 1994. Hydrologic framework of the
Fruitland Formation, San Juan Basin. New Mexico Bureau of Mines and Minerals
Bulletin 146: Coalbed methane in the upper Cretaceous Fruitland Formation, San
Juan Basin, New Mexico and Colorado, pp. 133-164.
Khodaverian, M. and McLennan. 1993. Cavity completions: a study of mechanisms and
applicability. Proceedings of the 1993 International Coalbed Methane Symposium
(Univ. of Alabama/Tuscaloosa), pp. 89-97.
Logan, T,L, Clark, W.F., and McBane, R.A. 1989. Comparing open-hole cavity and
cased hole hydraulic fracture completion techniques, San Juan Basin, New
Mexico. SPE Paper 19010, Proceedings SPE Low Permeability Reservoirs
Symposium (Denver).
Lyford, F.P., 1979. Ground Water in the San Juan Basin, New Mexico and Colorado,
USGS Water-Resources Investigations 79-73, 22p.
Mavor, M.J., Dhir, R., McLennan, J.D., and Close, J.C. 1991. Evaluation of the
hydraulic fracture stimulation of the Colorado 32-7 No. 9 well, San Juan Basin.
Rocky Mountain Association of Geologists Guidebook, "Coalbed methane of
Western North America", Fall Conference and Field Trip, pp. 241-249.
New Mexico Bureau of Mines and Minerals. 1993. Atlas of Rocky Mountain Gas
Reservoirs, p. 122.
New Mexico Energy, Minerals and Natural Resources Department, Oil Conservation
Division Regulations Title 19, Chapter 15,
http://www.emnrd.state.nm.us/ocd/OCDRules/Oil&Gas/rulebook/rulebook.pdf,
2001.
Oldaker, P.R. 1991. Hydrogeology of the Fruitland Formation, San Juan Basin,
Colorado and New Mexico. In Coalbed methane of Western North America.
Rocky Mountain Association of Geologists, pp. 61-66.
Palmer, ID. and Kutas, G.M. 1991. Hydraulic fracture height growth in San Juan Basin
coalbeds. SPE 21811, Proceedings SPE Low Permeability Reservoirs Symposium
(Denver).
Palmer, ID., Lambert, S.W., and Spitler, J.L. 1993a Coalbed methane well completions
and stimulations. Chapter 14 in AAPG Studies in Geology 38, pp. 303-341.
Palmer, ID., Mavor, M.J., Spitler, J.L., Seidle, J.P., and Volz, R.F. 1993b. Openhole
cavity completions in coalbed methane wells in the San Juan Basin. Journal of
Petroleum Technology, 45(11): 1072-1080 (November).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs Al-32
image:
EPA 816-R-04-003 Attachment 1
The San Juan Basin
Pashin, J.C. and Hinkle, F. 1997. Coalbed Methane in Alabama. Geological Survey of
Alabama Circular 192, 71pp.
Penny, G.S, Conway, M.W., McBane, R. 1991. Coordinated laboratory studies in
support of hydraulic fracturing of coalbed methane. Proceedings, 1991 SPE
Annual Technical Conference and Exhibition (Sigma Reservoir Engineering), 66,
pp. 231-246.
Stevens, S.H., Kuuskraa, J.A., and Schraufnagel, R.A. 1996. Technology spurs growth
of U.S. coalbed methane. Oil and Gas Journal, pp. 56-63 (January).
Stone, W.J., Lyford, P.P., Frenzel, P.P., Mizell, N.H. and Padgett, E.T. 1983.
Hydrogeology and water resources of San Juan Basin, New Mexico. New Mexico
Bureau of Mines and Mineral Resources, Hydrologic Report 6, 70 p.
Wilson, B. 1986. Water Use in New Mexico. New Mexico State Engineer Technical
Report 46, 84 p.
Young, G.B.C., McElhiney, J.E., Paul, G.W., and McBane, R.A. 1991. An analysis of
Fruitland coalbed methane production, Cedar Hill field, northern San Juan Basin;
SPE Paper No. 22913, Proceedings SPE Annual Technical Conference and
Exhibition (Dallas).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A1-3 3
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
Attachment 10
The Sand Wash Basin
The Sand Wash Basin is in northwestern Colorado and southwestern Wyoming. It is part
of the Greater Green River Basin, which includes the Washakie Basin, the Great Divide
(Red Desert) Basin, and the Green River Basin (Figure A10-1). These sub-basins are
separated by uplifts caused by deformation of the basement rock. The Cherokee Arch, an
anticlinal ridge that runs east to west along the Colorado/Wyoming border, separates the
Sand Wash Basin from the adjacent Washakie Basin. The Greater Green River Basin, in
total, covers an area of approximately 21,000 square miles. The Sand Wash Basin covers
approximately 5,600 square miles, primarily in Moffat and Routt Counties of Colorado.
Coalbed methane resources in the Sand Wash Basin have been estimated at 101 trillion
cubic feet (Tcf). Approximately 90 percent of this resource is within the Williams Fork
Formation (Kaiser et al., 1993). Despite this ample resource, economic viability of
recovery of the gas is limited by the presence of large volumes of water in most coalbeds.
Presently, there appears to be no commercial production (GTI, 2002); however,
approximately 120 permits for drilling within Moffat County were issued between
February 2000 and August 2001 (Colorado Oil and Gas Commission, 2001). It is not
clear exactly how many of these permits were related to coalbed methane exploration and
production.
10.1 Basin Geology
The geologic history of the Sand Wash Basin is relatively complex, characterized by
periods of deposition followed by deformation related to tectonic activity. This activity
has impacted depositional patterns, coal occurrence and maturity, and hydrology (Tyler
and Tremain, 1994). A very thorough discussion of the geologic history of the Sand
Wash Basin is available in Tyler and Tremain (1994).
The coal-bearing formations in the region include the lies, Williams Fork, Fort Union,
and the Wasatch Formations (Figure A10-2). These formations were deposited, from
bottom to top, during the Upper Cretaceous, Paleocene and upper Paleocene periods. The
total thickness of the coal seams in these formations can measure up to 150 feet
(Quarterly Review, 1993). Basement rock formations in the Sand Wash Basin can be as
deep as 17,000 feet (Tyler and Tremain, 1994). A map of the coal and geologic features
is presented in Figure A10-3a and a conceptual cross-section is presented in Figure A10-
3b.
The Sand Wash Basin was near the western edge of the Western Interior Seaway that
spreads across what is now central North America during the Upper Cretaceous (Figure
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-1
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
A10-4). During the late Cretaceous the seaway retreated to the northeast. Intermontane
basins developed during the Laramide, and coal-bearing fluvial-lacustrine sediments were
deposited (Quarterly Review, 1993). The coal in the Sand Wash Basin was formed from
peat deposited in swamps along a broad coastal plain. Sediments that eroded from
nearby uplift formations covered the peat beds (Tyler and Tremain, 1994). The
alternating deposition of organic material and sands was repeated many times creating
layers of coal interbedded with layers of sandstone and other sedimentary rocks that filled
the basin.
Cretaceous or Mesaverde Group coal in the Sand Wash Basin ranges in rank from sub-
bituminous along the basin margins to high volatile A bituminous coal in the deeper parts
of the basin. These ranks are indicative of moderately mature to well-developed mature
coal formed under high pressure and high heat. Within the Mesaverde Group, the most
important potential coalbed methane resource in the basin (Kaiser et al., 1993), the coal
ranks from sub-bituminous along the basin margins to medium volatile bituminous in the
basin center (Kaiser et al., 1993). The methane in these coals formed both biogenically
(by bacterial action on organic matter), and thermogenically (under high temperature).
The average gas content of 261 coal samples collected during two studies was 147
standard cubic feet of methane per ton of coal (Boreck et al., 1977; Tremain and Toomey,
1983). Some samples from the Sand Wash Basin have been found to contain as much as
540 standard cubic feet of methane feet per ton. Gas content has generally been found to
increase somewhat with depth. At depths of less than 1,000 feet, gas content is typically
less than 20 standard cubic feet per ton, which has been taken to indicate that gas
probably leaked out of the shallow coalbeds into the atmosphere. Analysis of gas
samples has indicated that the gas is typically 90 percent methane, the remainder being
mostly nitrogen and carbon dioxide (Scott, 1994). Carbon dioxide content ranges from 1
to more than 25 percent (Scott, 1994).
Of all the coal-bearing formations, the Upper Cretaceous Williams Fork is the most
significant unit because it contains the thickest and most extensive coalbeds. The
Williams Fork Formation is within the Mesaverde Group that also includes the Almond
Formation along the Wyoming state line (Tyler and Tremain, 1994). The Almond
Formation is shown (Figure A10-2) as a separate formation overlying the Williams Fork
(Tyler and Tremain, 1994), but is also reported (Kaiser et al., 1993) to be a lateral
equivalent of the upper Williams Fork Formation found in the southern Sand Wash
Basin. For more information relative to this apparent conflict see Kaiser et al. (1993, p.
29). The coal-bearing Williams Fork Formation outcrops along the southern and eastern
margins of the basin, and may be deeper than 8,000 feet in the deepest part of the basin
(Figure A10-3b). The coals are interbedded with sandstones and shale. The thickest total
coal deposits in the Williams Fork Formation, up to 129 feet, are centered near Craig,
CO. This total is made up of several separate coalbeds up to 25 feet thick interbedded
with sedimentary rock.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-2
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
Stratigraphically above the Williams Fork Formation, the Paleocene Fort Union
Formation, which includes sandstone, siltstone, shale, and coal, is also a potentially
productive zone for coalbed methane production. The Fort Union outcrops at the
Elkhead Mountains east of the basin and along the southern and western parts of the
basin. The bottom of the Fort Union Formation is about 7,000 feet below the surface.
Net coal thickness can be up to 80 feet with as many as nine individual beds. Individual
beds up to 50 feet thick have been identified.
The Wasatch Formation includes beds of shale and sandstone and minor amounts of coal.
It can extend as deep as 2,000 feet below the surface. The Wasatch Formation has not
been targeted for coalbed methane development because of the small quantity of coal.
10.2 Basin Hydrology and USDW Identification
Regional groundwater flow in the Sand Wash Basin is from east to west and to the
northwest towards the center of the basin. Water enters the aquifers at the exposed
outcrops along the southern and eastern margins of the basin and moves northwestward.
Vertical movement of groundwater, including potential artesian conditions, is dependent
on local geologic conditions. Kaiser and Scott (1994) summarized their extensive
investigation of groundwater movement within the Fort Union and Mesaverde Group.
The Mesaverde Group is a highly transmissive aquifer. The coalbeds along with
associated sandstone beds within the group may be the most permeable part of the
aquifer. The Williams Fork Formation contains sandstone beds that are reported to be
excellent aquifers (Brownfield, 2002). Lateral flow within the Fort Union Formation is
slower, in part, owing to less permeable fluvial sandstones in the unit.
Total dissolved solids (TDS) concentrations of groundwater in the Mesaverde Group
were investigated by Kaiser and Scott (1994) (Figure A10-5). They found that chloride
concentrations ranged from 290 milligrams per liter (mg/L) in the eastern area of the
basin near the outcrops where water enters the aquifers, to more than 26,000 mg/L in the
central part of the basin. Calcium showed a similar pattern of distribution with the lowest
concentrations near the outcrops, increasing toward the basin center. Calcium
concentrations ranged from 10 mg/L to over 2900 mg/L. Based on the chloride and
calcium concentrations presented by Kaiser and Scott (1994), the water in the aquifers
near the recharge areas at the basin margins meets the water quality criteria for an
underground source of drinking water (USDW) of less than 10,000 mg/L, but the water in
the deeper central part of the basin does not (Figure A10-5). The mapped outcrop area
(Figure A10-3a) of the Mesaverde Group indicates that the coal seam lies within a
USDW where it is relatively shallow and close to the eastern and southern margins of the
basin.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-3
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
10.3 Coalbed Methane Production Activity
Coalbed methane resources in the Sand Wash Basin have been estimated at 101 Tcf.
Approximately 90 percent of this gas is in the Williams Fork Formation (Kaiser et al.,
1993). Approximately 24 Tcf of coalbed methane are located at depths less than 6,000
feet below ground surface (Kaiser et al., 1994). Despite this ample resource, economic
viability of recovery of the gas is limited by the presence of large volumes of water in
most coalbeds. Exploration in the 1980s and 1990s led to limited commercial use of the
resource. Records from the Colorado Oil and Gas Commission indicate that
approximately 31 million cubic feet of coalbed methane was produced in Moffat County
during 1995 (Colorado Oil and Gas Commission, 2001). From 1996 to 1999 (the last
year that data are available), no further gas was produced in this County (Colorado Oil
and Gas Commission, 2001). However, Colorado Oil and Gas Commission records
indicate that approximately 120 permits for drilling within Moffat County were issued
during the period from February 2000 through August 2001 (Colorado Oil and Gas
Commission, 2001). It is not clear exactly how many of these permits were related to
coalbed methane exploration and production, but a handful of the permits were issued to
gas companies, and the permits are listed as targeting known coalbeds within specific
methane producing formations (Colorado Oil and Gas Commission, 2001).
At Craig Dome in Moffat County, Cockrell Oil Corporation drilled a 16-well
development for exploration in the Williams Fork Formation. According to the Colorado
Geological Survey, Craig Dome is located along the Cedar Mountain fault system
(Colorado Geological Survey, 2002). The wells were abandoned a short time later
because of excessive water. The Colorado Geological Survey indicated that the fault
system may act as a conduit for anomalously high water migration from the outcrop. An
average total of 40 feet of high-volatile bituminous coal was encountered in beds up to 15
feet thick. Gas content was tested at 10 to 350 cubic feet per ton of coal. Wells were
cased through the target coalbed, perforated, and hydraulically fractured using water and
sand. The wells yielded large volumes of fresh water with TDS levels measuring less
than 1,000 mg/L, but little gas (Colorado Oil and Gas Commission, 2001). Water was
removed at an average of 21,756 gallons per day per well during testing. Based on
records from the Colorado Oil and Gas Commission, Cockrell Oil Corp does not appear
to be involved currently with coalbed methane production in this region (Colorado Oil
and Gas Commission, 2001).
The Colorado Geological Survey also indicated that faults in Trout Creek Canyon
southeast of Craig are on trend with (and thus are likely to be related to) the Cedar
Mountain fault system (Colorado Geological Survey, 2002). In addition, KLT Gas Inc.
has a pilot program southwest of Craig Dome on the Breeze lease which is on trend with
the Cedar Mountain fault system. If a fracture propagates into and along a fault plane, it
may contaminate a USDW (Colorado Geological Survey, 2002.)
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-4
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
Limited commercial success has been experienced in the basin. As of 1993, only one
commercial operator, Fuelco, was working in the basin. Fuelco was operating 11 wells
along Cherokee Arch at 40 to 80 acre spacing. Well depths were to 2,500 feet. A total of
40 feet of coal was found in the Almond Formation (Mesaverde Group) between 810 to
2,360 feet. All wells were cased through the coal, selectively perforated, and stimulated
using water and sand. Gas production averaged a total of 50,000 cubic feet per day from
four wells. The highest producing well peaked at 100,000 cubic feet per day (Quarterly
Review, 1993). Total production of gas through 1993 from the Dixon Field, the only
producing field in this region, was about 84 million cubic feet (Kaiser et al., 1993). Total
water production for the four wells was high at 126,000 gallons per day due to the high
permeability of the coal (Quarterly Review, 1993). Water pumped from the wells
contained 1,800 mg/L of TDS and was discharged to the ground with a National
Pollution Discharge Elimination System permit (Quarterly Review, 1993).
The Sand Wash Basin has been used by the University of Texas Bureau of Economic
Geology in the development of its Coalbed Methane Producibility Model (Kaiser et al.,
1994). The development of the model was based on a comparison of basins that included
the Sand Wash Basin and the San Juan Basin of southwestern Colorado and northwestern
New Mexico. The San Juan Basin has proven to be a very productive coalbed methane
resource. The Sand Wash Basin was used as an example of a basin with low potential for
productivity (Figure A10-6) (Kaiser et al., 1994).
Hydraulic fracturing has been used in the Sand Wash Basin to improve the flow of gas
into the wells. Hydraulic fracturing fluids have typically consisted of water with sand
used as a proppant. However, very little information was available regarding specific
types and volumes of fluids and proppants used. No indication of the use of other
materials was noted in the sources reviewed (Colorado Oil and Gas Commission, 2001).
10.4 Summary
Coalbeds containing methane gas are present within the Sand Wash Basin at accessible
depths. Some investigation and very limited commercial development of this resource
have occurred, mostly in the late 1980s and early 1990s. There appears to be no
commercial production at present. Development of coalbed methane resources in the
Sand Wash Basin has been slower than in many other areas due to limited economic
viability. The need for extensive dewatering in most wells has been a limiting factor,
compounded by relatively low gas recovery.
Between 1996 and 1999, no coalbed methane was produced in Moffat County. Permits
for new gas wells have been issued indicating that there may be some continued interest
in this area (Colorado GIS, 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-5
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
Groundwater quality in the basin varies greatly. Typically, chloride and calcium
concentrations within the coal-bearing Mesaverde Group are low and potentially within
potable ranges in the eastern and southern parts of the basin, implying the existence of a
USDW, and therefore the potential for impacts. Concentrations increase as the water
migrates toward the central and western margins of the basin. IDS concentrations
significantly higher than the 10,000 mg/L USDW water quality standard have been
detected in the western portion of the basin.
Compared to other potentially productive areas of the country, very little information has
been published regarding current developments, groundwater location and conditions,
drilling techniques, etc. The level of information available seems to be commensurate
with the amount of commercial activity.
The use of fracturing fluids, specifically water and sand proppant, has been reported for
this basin. No record of any other fluid types has been noted. Although variable, the
water quality within the fractured coals indicates the presence of USDWs within the
coalbeds.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-6
image:
o g, w
§ O ps
a- 3. ET
2. & C
^ 2 s
1/1 s &
R'oq
^ 3
ll
3 c0
S o
OQ g
o R
^+1 ft
9 8
COALSED METHANE PROJcd'8
1 Butronwood Pelrcfaun inc.
2 Oil Cow
3 Oi C&
5 Metfyai ire,
6 Co.
7 Texaco ira.
fl Tntafi Oil fliVf G^s Corp.
•i* 6
*?
•31!"'
\
!
/Sand Wash
Basin
2* •••'<-,
Lix-.at €-1 o; tne Sari: VVas'i Das M a* Col?jfa«.io ai-J '/v'yoni eg tQujrterly Review, 1993}
Figure A1C-1
w
oo
ON
O
UJ
Cd 3
|. 5
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
? 3' 8
s^
en ^ en
2. B- C
a i a
fj- ft
0 OQ
^ 3
3 §
5 o
OQ g
O O
«5 ft
VI
> i
£ ^
9 8
.T D-'v'OE
AND
•*•&! »AKIC OA5lt,S
&« J» '
of Geolc3g-€ Forrnatidns Wilhin the Sand Wash Basin and NtMjhboring Basins
Tyler anc Tremam, IWM'i
oo
K^
ON
i1
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
? 3' 8
s^
en ^ en
2. B- C
a i a
-
0 OQ
^ 3
3 §
o id
-
5 o
OQ g
O O
«5 ft
VI
> i
^ ^
9 8
! r-i;
M.ip :if i ..">nl ,-i" t > -..•v» nfjir - nnli.rp1
thr: 5"i'".J '</<.'"*
oo
K^
ON
i1
-I
*l
image:
Q2 C^ ^
a- 3. £T
ft B fa
S & 5 '
<f B B
g. °^, 0
5* ^ ^
<t ft »73
ft *-> r*-
VI "* VI
^ iT"! r^~
5 *~< ^
o &. C
S'3 »
1/1 B &
£• ft
° ^
3 §
0 B
3. t/3
OQ g
O O
VI
c^
> fD
r- ) tO
i 0
i—1 O
t — ^ r~~
C
Southwest
i n Don ..
sate
'•"= ^ , „
T-i .1-t
R v *„ •
*' ->
c f"«.B . '" %* ^ \^
£ ^V \ 7^
I r.-n- ^1^
;'
r< iv*
i i
-
_' Ti -
1 WJi i
**u,.. J
I OKI '
^
*•
f
- —
- «.,
*f""«t ""»''•«-•
^?¥^*
*•* °°i«JL, '• •'
*""•„•
j,
'.L-T r.* , *. M
. » i " ? *
T - -
"7"" •;
-„_ ,
1 ,,.,.
""~^jrl
*t?
8^
.
ul '
J '
- . _
- , _
"cJ"
.-', a
1--~.
•=
AQ U ' ~A R D S AQ U I r L
rrrrrrrr.'i
f: :;;; ;' | i fth
[ 1 ViincclB Shttle !
- - "
„:;;:!; Wei
. lea
f
C'
Norlheasl
Hiid''ai<jy
* *, " r 'JO'S
/ifX
,^J ff ' '
~ " - t ~ " K'JJ? ^
"• ^~ * e ^i 'fl3 ^
f f'r'"°wG ' ^
' l&'-'f "' I
" } W ' ift" " ^
_ , , _^_ ^ -- " / ' .; /IjT J
- ' / l-'-M? f
~ " ' ~~ ^ " ? F *"s
/" ' ' £ft'?} /
/ "" t /
/" .* • ^g$ f
~~ " ~^'" ** ^^'
_- -T • _yl! "Jf^ /
' ' - - '"^^y
:r- .. ~':' \"\..«'&!P<J&£Z^ ^*~
t , x.V i^' "*"-«•- ••'" "''"''
feSP^>^
it •
9?<5
T J &S"
"" *• 4^',,'
2' 35
1 Ftjf! ^
~:
-
- 1 i^lKl 1
r
0 ! -
S!0
. :-«:-.
- s« «>«
oO-s
, 1 1
.-.I ?fi l,,t,
jj-r;
T 1 |£XV- P**fl',»-f'«
irk For.iiatori - nn^n'inmclr'-'nijrra'-n ""^^ Ho« pdti.
Forrnation _ "H^vi- M.i^r r.cvn'-.'. O l-au'-i p.iffllfrt Urn".
, i 1 1 i
Conceptual Cross Sect on C-C' Refer to Figu-a A10-3a)
(Tylor Jf-u T't-!fdiin. 1997}
-ICI.T •%•!.. it
^
oo
ON
^
f'
O
UJ
S
ft
in
§ »>
^|
« B
&" ft
2. 0
image:
EPA816-R-04-003
Attachment 10
The Sand Wash Basin
Crefoceous Pole
40'
" " ,* ." "„-
•":i
30°
-; Lonti oreo
Location of the Sand Wash Basin in Relation la the Western Interior Seaway
of Upper Cretaceous Times ("Tyler and Tremain,
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A10-11
image:
o g, w
I, a EL
c? 3. a
ro B ?5
*lf
I^P
« ^ «
o &
a
3 c
3. gj
cro g
s,S
East
I I'MiKM I PK
1 ! I
I I1STN l''».'l • Utl. ! 1 'UN ! "-."I I '•''• '<
' ..-—-
r_ r -,
Grojnd Watef Quality Trends in the Sand Wash Basin ^Sc:ott, 1994}
w
oo
ON
O
U)
ro
c«
^
o
Cd 3
2- o
image:
EPA816-R-04-003
Attachment 10
The Sand Wash Basin
CD
L
Vl
if
It
07
r
E
1C
DC
L.
K
crt
.
G. _
4 ^
fe "UJ
x: <-
c
O
P
L
(T.
a;
u.
O
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A10-13
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
REFERENCES
Brownfield, Michael. 2002. USGS, Denver Federal Center, Denver, CO. Personal
(written) communication.
Boreck, D. L., Jones, D. C., Murray, D. K., Schultz, J. E., and Suek, D. C. 1977.
Colorado coal analyses, 1975 (analyses of 64 samples collected in 1975):
Colorado Geological Survey Information series 7, 112 p.
Colorado Geological Survey. 2002. Public Comment OW-2002-0002-0086 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Colorado Oil and Gas Commission Website-Colorado GIS. 2001. Approved Drilling
Permits: http://cogccweb.state.co.us/cogis/DrillingPermitsList.asp.
Gas Technology Institute (GTI) website. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org.
Kaiser, W. R., Scott, A. R., Hamilton, D. S., Tyler, Roger, McMurry, R. G., Zhou,
Naijiang, and Tremain, C. M. 1993. Geologic and hydrologic controls on
coalbed methane: Sand Wash Basin, Colorado and Wyoming: The University of
Texas at Austin, Bureau of Economic Geology, topical report prepared for the
Gas Research Institute under contract no. 5091-214-2261, GRI-92/0420, 151p.
Kaiser, W.R. and Scott, A.R. 1994. Hydrologic setting of the Fort Union Formation,
Sand Wash Basin. Report of Investigations - Geologic and Hydrologic Controls
on Coalbed Methane, Texas, University, Bureau of Economic Geology, 220, pp.
115-125.
Kaiser, W.R., Scott, A., Zhou, N., Hamilton, D.S., and Tyler, R. 1994. Resources and
Producibility of Coalbed Methane in the Sand Wash Basin. Report of
Investigations - Geologic and Hydrologic Controls on Coalbed Methane, Texas
University, Bureau of Economic Geology, 220, pp. 129-145.
Quarterly Review of Methane From Coal Seams Technology. 1993. Greater Green
River Coal Region Wyoming and Colorado, pp. 13-17.
Scott, Andrew R. 1994. Coal Rank, Gas Content, and Composition and Origin of
Coalbed Gases, Mesaverde Group, Sand Wash Basin. Bureau of Economic
Geology and Colorado Geological Survey, Resource Series 30, pp. 51-62.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-14
image:
EPA 816-R-04-003 Attachment 10
The Sand Wash Basin
Tremain, C. M., and Toomey, J. 1983. Coalbed methane desorption data: Colorado
Geological Survey Open-file Report 81-4, 514 p.
Tyler, R., and Tremain, C.M. 1994. Tectonic evolution, stratigraphic setting, and coal
fracture patterns of the Sand Wash Basin. Report of Investigations - Geologic and
Hydrologic Controls on Coalbed Methane, Texas University, Bureau of Economic
Geology, 220, pp. 3-19.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A10-15
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
Attachment 11
The Washington Coal Region (Pacific and Central)
The Pacific Coal Region (Figure Al 1-1) is approximately 6,500 square miles and lies
along the western and eastern flanks of the Cascade Range from Canada into northern
Oregon. The coals along the western flank lie within the Puget downwarp. Bellingham,
Seattle, Tacoma, and Olympia in the State of Washington, and Portland, Oregon lie in or
adjacent to the sub-basins. Choate et al. (1980) estimated coalbed methane resources for
four target sub-basins (Figure Al 1- 1) representing 1,800 square miles of the 6,500
square mile Pacific Coal Region to be 0.3 trillion to 24 trillion cubic feet (Tcf). The
Central Coal Region (Figure Al 1-2) is primarily the Columbia Plateau, between the
Cascade Range to the west and the Rocky Mountains in Idaho, to the east. The Region
extends from the Okanogan Highlands in the north to the Blue Mountains to the south,
and encompasses approximately 63,320 square miles. Pappajohn and Mitchell (1991)
estimated the coalbed methane potential of the Central Coal Region to be more than 18
billion cubic feet (Bcf) per square mile. According to the available literature, there were
no producing fields in either the Pacific Coal Region or the Central Coal Region in
Washington as of 2000 (GTI, 2001).
11.1 Basin Geology
A series of discontinuous coal fields lie along the western flank of the Cascade Range
(Figure Al 1-3). The Roslyn and Taneum-Manastash fields are located on the eastern
flank of the Cascade Range (Figure Al 1-3). The coal-bearing sediments were formed in
a swampy fluvial-deltaic coastal plain depositional environment in the Paleocene to late
Eocene Eras. In the Columbia Plateau Region, the Cretaceous to Eocene coal-bearing
rocks are buried beneath a thick sequence of extrusive basalts.
The coal-bearing deposits of the Pacific and the Central Coal Regions are Cretaceous to
Eocene Age and formed within fluvial and deltaic depositional environments prior to the
uplift of the Cascade Mountain Range. The coalbeds of the Pacific and Central Basins
are thought to result from peat accumulations in poorly drained swamps of the lower
deltas while the thinner coalbeds probably formed in the better drained upper deltas
(Buckovic, 1979 as cited in Choate et al., 1980). During the Oligocene, Cascade volcanic
activity buried the deltaic sediments and compression caused some deformation of the
sediments. During the Miocene, extensive volumes of basalt poured out in central
Washington and covered the coal-bearing fluvial deposits. During the late Pliocene, the
Coast Range and the Cascades continued to be uplifted, separating the Pacific Coal
Region from the Central Coal Region, and causing extensive tectonic deformation,
folding and faulting, of the coal-bearing sediments.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -1
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
Deformation of the coal-bearing rocks increases toward the Cascade front. Fracturing
may enhance porosity and permeability of the coalbeds, allowing greater methane storage
and production (Pappajohn and Mitchell, 1991). On the other hand, however, fracturing
may also increase the porosity and permeability of confining beds, allowing methane to
escape up the stratigraphic section over time and dissipate in the atmosphere. Continuing
deformation, primarily faulting, may be a limiting factor controlling methane production
in the Pacific Coal Region as well.
11.1.1 Pacific Coal Region Geology
In the Pacific Coal Region, deformation has increased geologic complexity making it
difficult to follow or correlate coalbeds, especially across faults. Geothermal heating
along the western flank of the Cascades created a thermally altered zone of increased coal
rank ranging into the bituminous and anthracite ranks. The maturation to bituminous
rank increases potential methane yields (Walsh and Lingley, 1991; Pappajohn and
Mitchell, 1991).
The major coal-bearing areas are in, from north to south, Whatcom, Skagit, King, Pierce,
Kittitas, Thurston, Lewis, and Cowlitz Counties in Washington (Figure Al 1-3). The
discussion of regional geology presented here illustrates the geologic conditions in the
Green River district in King County, the Wilkerson-Carbonado coalfield in Pierce
County, and the Centralia-Chehalis district in northern Lewis and southern Thurston
Counties, and does not attempt to provide a detailed description of every coalfield. For
more detailed information on the Bellingham area, Whatcom County, the reader is
referred to Beikman et al. (1961), for Whatcom and Skagit Counties to Jenkins (1923 and
1924), and for the Roslyn coal area to Walker (1980). Other areas not discussed but
important within the Pacific Region are the Toledo-Castle Rock District, and the Roslyn-
Cle Elum and Teneum-Manastash fields. The stratigraphy for three sub-basins (Green
River, Wilkerson-Carbonado, and Centralia-Chehalis) of the Pacific Coal Region is
presented in Figure Al 1-4. The general setting and geology of each sub-basin is unique
and complex.
The coal deposits of King County are located southeast of Seattle (Figure Al 1-3). The
Green River district is the largest and most extensively mined coal-bearing area in King
County. The King County coals occur in the Puget Group of Eocene Age (Figure Al 1-
4). Evans (1912) divided the Puget Group into 3 coal zones, which, from oldest to
youngest, are the Bayne, Franklin, and Kummer. Deformation has been moderate and
most of the coalbeds dip less than 35 degrees. In parts of the Green River district the
deformation has been more intense, and dips of 50 degrees or more are common. The
King County coals range in rank from subbituminous to high-volatile bituminous.
Within the Green River District, the Puget Group is estimated to be at least 6,500 feet
thick and contains at least 15 coalbeds up to 40 feet thick (Beikman et al., 1961). The
principal coalbeds are located in the Franklin and Kummer zones in the Puget Group
(Vine, 1969). Coal has been mined in the Green River District since about 1883, and it
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -2
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
has produced more than 25 million tons of coal. Currently there is no coal production in
the district.
The Wilkerson-Carbonado coalfield is located in Pierce County, southeast of Seattle
(Figure Al 1- 3). The Pierce County coals occur in the Eocene Carbonado Formation
(Beikman et al., 1961). The Carbonado consists of more than 5,000 feet of interbedded,
layered lenses of sandstone, siltstone, mudstone, and shale with carbonaceous shale and
coal (Figure Al 1-4). At least 10 coalbeds have been identified in the area. Coalbeds
range in thickness from 1 to 5 feet with the maximum thickness of 15 feet. The coals
range in rank from high-volatile bituminous to low-volatile bituminous. The Wilkerson-
Carbonado coals have the highest rank of any major coal-bearing area in Washington
State. Throughout the field, deformation has been intense. Dips of 60 degrees or more
are common, and fault displacements range from a few feet to more than 1,500 feet.
Although these areas have recently been targets of coalbed methane exploration, there is
currently no production.
The coal deposits of Lewis and Thurston Counties occur in the Skookumchuck Formation
(Figure Al 1-3) of late Eocene Age (Snavely et al., 1958). The Centralia-Chehalis district
is located in northern Lewis and southern Thurston Counties (Figure Al 1-3).
Deformation of the Skookumchuck is moderate resulting in tightly folded anticlines and
broad open synclines. The coal deposits have been cut by a series of high angle reverse
faults roughly paralleling the fold axes. The faults dip to the northeast, with the
southwest block downthrown, and have displacements ranging from 200 to 500 feet. The
coal rank ranges from lignite to anthracite. The central part of the Centralia-Chehalis
district contains as many as 14 subbituminous coalbeds ranging from a few inches to over
40 feet in thickness. The district contains more than half of the calculated coal reserves
of the State. The Trans Alta Centralia Mining Company continues to operate a major strip
mine centered about 5 miles northeast of Centralia, where it is anticipated that 9,400
acres will be stripped over 35 years. Within the Centralia mine, the Big Dirty bed is
more than 40 feet thick. To the west of Centralia, the Vader coal area contains several
lignite beds with thickness up to 20 feet, which may correlate in part with the coals in the
Centralia-Chehalis area.
In Whatcom and Skagit counties (Figure Al 1-3), the Chuckanut Formation contains as
many as 15 coalbeds, ranging from 1 to 15 feet thick and ranking from lignite to
anthracite, but generally bituminous. The rank of the coal increases eastward towards the
crest of the Cascades Range.
The rank of Pacific Region coals varies greatly from place to place, ranging from lignite
to anthracite, but generally rank increases toward the crest of the Cascade Range. The
coal rank is used to identify bituminous coal-target areas where gas yields may be
greatest. While the structural geology is very complex, the thermally-altered
metamorphic zone is rather predictable. Both of these factors will play a major role in
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -3
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
the design of any exploration and development plans for coalbed methane in the Pacific
Coal Region.
The complex stratigraphy and structural deformation of the lenticular coals in the Pacific
Coal Region are major obstacles to the exploration and development of coalbed methane
fields. Predicting the location of coalbeds is a complex and difficult process because the
geology in the area has been modified by intense deformation. Additionally, the faulting
that commonly occurs along the axes of anticlines may form conduits for the escape of
methane through overlying confining beds. Steeply dipping beds of coal have presented
difficulties in controlling drill bit directions and in development and stimulation for
coalbed methane production.
Choate et al. (1980) estimated coalbed methane resources for four target sub-basins
(Figure Al 1-1), representing 1,800 square miles of the 6,500 square mile Pacific Coal
Region, to be 0.3 trillion to 24 Tcf. Methane had been encountered in 67 oil and gas
exploration wells drilled in this region by 1984. Methane gas was found at depths of less
than 500 feet in 25 wells, less than 1,000 feet in 38 wells, and less than 2,000 feet in 50
wells. In western Whatcom County, methane has been found in unconsolidated glacial
drift capped by impervious clay beds. East of Ferndale, methane gas reportedly has been
produced commercially from unconsolidated deposits at depths ranging from 166 to 193
feet at flow rates ranging from 750,000 to 5,000,000 cubic feet per day (Choate et al.,
1980).
11.1.2 Central Coal Region Geology
The Central Coal Region refers to the coal-bearing formations east of the Cascade Range.
The Columbia River Basalt Group, primarily the Grande Ronde Basalt, Wanapum Basalt,
and Saddle Mountains Basalt bury the Cretaceous to Eocene coal-bearing formations of
the Central Coal Region. In this region, methane is entrained in groundwater from
confined aquifers in the basalts. Interbedded with the flood basalts are epiclastic and
volcaniclastic sediments. The less fractured zones of basalt appear to act as aquitards
(Johnson et al., 1993). Johnson et al. (1993) have concluded that the greatest volume of
methane is derived from upward migration from the underlying Eocene coals. They also
suggest that faults through the underlying sediments and basalts provide conduits for the
migration of gas-bearing groundwater into the confined zones.
The Yakima fold belt lies between the confluence of the Snake and Columbia Rivers and
the Cascade Range, and is a series of broad asymmetric anticlines and synclines whose
axes generally trend west northwest to east southeast (Figure Al 1-5). The anticlinal
ridges are typically cut by thrust faults that are inclined and steepen with depth (Reidel et
al., 1989). While the anticlines may form structural traps for methane in the source
coalbeds, the thrust faults in the anticlines may form conduits for the upward migration of
methane through overlying confining beds. The fold structures are very flat and broad
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -4
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
and do not result in the steeply dipping strata that are characteristic of the Pacific Coal
Region west of the Cascades.
11.2 Basin Hydrology and USDW Identification
Surficial deposits of Pleistocene glacial outwash locally form aquifers capable of
sustaining public drinking water supplies in the Pacific and Central Washington Regions.
In the Central Coal Region, aquifers in the basalts are extensively developed for
irrigation. Public water supplies in Pierce County (Olympia area) and King County
(Seattle area) of the Puget Sound Region (Pacific Coal Basin) are obtained from the
glacial drift aquifer (Dion, 1984) that overlies Eocene sediments, which may contain coal
and methane. Water quality information from four gas test wells indicates the presence
of 1,330 to 1,660 milligrams per liter (mg/L) total dissolved solids (TDS) in water within
the coalbeds of Pierce County (Dion, 1984). This meets the water quality requirements
of an underground source of drinking water (USDW). The Washington Department of
Ecology and the EPA deemed this water to be of sufficient quality to permit its discharge
to surface waters of the Carbon River (Pappajohn and Mitchell, 1991).
The Columbia River Basalt Group is identified as a major regional multi-aquifer province
(Lindholm and Vaccaro, 1988; Dion, 1984). The aquifer is used extensively for
irrigation, but may also be used as a source of drinking water. Wells in the Basalts
commonly yield 150 to 3,000 gallons per minute. TDSs in the water produced generally
range from 250 to 500 mg/L (Dion, 1984).
The occurrence of methane in groundwater is one factor leading to the assessment of the
coalbed methane production potential in Washington. Methane in groundwater occurs in
the basalts, but only in confined aquifers (porous or fractured zones near the top or
bottom of a basalt layer), and is thought to have migrated upward from underlying
coalbeds. Water supply wells and irrigation wells in the Columbia River Basalts and
water wells in numerous different lithologies in the Pacific Coal Region have been
recognized as containing methane. Data demonstrating the co-location of a coal seam
and a USDW were found for Pierce County, where methane gas test well results report
TDS levels far lower than the 10,000 mg/L USDW water quality threshold (Dion, 1984).
11.3 Coalbed Methane Production Activity
Complex stratigraphy and structural deformation creates major obstacles to the
development of gas from the Pacific Coal Region. The coals are known from active and
inactive mines to be gassy, folded, faulted, and commonly steeply inclined. The
difficulties and dangers involved with underground coal mining led to closure of the
mines once the shallow deposits were exhausted. However, their characteristics have
been well documented by the mining operations. Many of these same structural
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -5
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
characteristics have impeded the development of coalbed methane gas. The available
literature indicates that no significant production had been achieved by 1996 (GRI, 1999).
According to the available literature, there were no producing fields in either the Pacific
Coal Region or the Central Coal Region in Washington as of 2000 (GTI, 2001).
However, in northwest Oregon, the Mist gas field was developed in the 1990s.
11.3.1 Pacific Coal Region Production Activity
Between 1986 and 1993, 19 coalbed methane wells were drilled in the northern Pacific
Coal Region (Quarterly Review, 1993). Three tests were conducted near the town of
Black Diamond in the Green River coal area of King County. One of the wells was
hydraulically fractured and the others completed by open-hole cavitation. Steep dips of
the strata led to wellbore deviation during drilling and to caving following the fracturing
operations. One well produced 32,000 to 62,000 cubic feet per day of coalbed methane
gas with no water in an open-hole test. Another was hydraulically fractured with 12/20
mesh sand and nitrogen foam in two zones at depths of 2,228 to 2,442 feet and 2,505 to
2,638 feet, but no test results were released. Caving was so prominent that it interfered
with wellbore cleanup following the hydraulic fracturing operations. According to
available publications, optimal fracturing and completion methods for use in the
structurally difficult Pacific Coal Region are yet to be applied and proven.
11.3.2 Central Coal Region Production Activity
The one commercial gas field (Rattlesnake Hills) in the Central Coal Region was shut
down in 1941. Production from the Cretaceous to Eocene coalbeds that lie below the
basalts may have large potential. Pappajohn and Mitchell (1991) estimated the coalbed
methane potential of the Central Coal Region to be more than 18 Bcf per square mile. It
is unlikely that the whole 63,320 square miles of the region could yield that rate because
the coals are only known to occur below the basalts in the western part of the basin.
Much is not known about the potential coalbed methane production from these obscured
deposits, and development depends on successful exploration.
Although the coals of the Central Coal Region may not be as greatly deformed and
unpredictable as those in the Pacific Coal Region, they are overlain by the Columbia
River Basalt Group, in which individual basalt flows up to 300 feet thick can cover
thousands of square miles. The Rattlesnake Hills gas field operated between 1913 and
1941 in the western part of this region and indicates greater potential for development.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -6
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
11.4 Summary
The geologic structure of the coal-bearing rocks is difficult to interpret in the Pacific and
Central Coal Regions, and methane may be technically difficult to produce in these
regions. A connection exists between the Washington coalbeds and a USDW. However,
there were no producing coalbed methane wells in the Pacific and Central Coal Regions
in Washington as of 2000 (GTI, 2001). In some areas, the Pacific and Central Regions'
coals exist within a potential USDW. In other areas of the basin, there is evidence that
the coalbeds are below a USDW. Hydraulic fracturing has been documented in this
region. Data demonstrating the co-location of a coal seam and a USDW were found for
Pierce County, where methane gas test well results report TDS levels of 1,330 to 1,660
mg/L, far less than the USDW classification limit of 10,000 mg/L (Dion, 1984).
In this region, methane occurs in groundwater flowing through fractured zones in basalts,
although less fractured zone of the basalts appear to act as hydraulic confining layers.
Johnson, et al. (1993) concluded that the greatest volume of this methane has migrated
upward from underlying coalbeds. Water supply wells and irrigation wells in the
Columbia River Basalts and water wells in numerous different lithologies in the Pacific
Coal Region have been recognized as containing methane. Development of coalbed
methane in the Washington Coal Region may have some impact on highly productive
basalt aquifers that meet the requirements of a USDW and are already in use as large
sources of irrigation water for agriculture.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -7
image:
EPA816-R-04-003
Attachment 11
The Pacific and Central Coal Regions
122"
«f
he Pacific Coal Ri*g;on Siiowuig Targ-Dteo Suboa^ins
for Coalbed Wethane Estimates {Choale et al, 1&BO|
figuro ft.11-1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
All-8
image:
EPA816-R-04-003
Attachment 11
The Pacific and Central Coal Regions
in*
Ml U i
Scwaticio
The Central Coal Region (Johnson elal,, 1993)
t-.qur-
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
All-9
image:
EPA816-R-04-003
Attachment 11
The Pacific and Central Coal Regions
-aw m"
'• ^ -- 'iX
*««.t3uvMi __- __• "^ ;.|-^
1 -ILiM,; ' " 'J'"''
1 •, • ' ','- ;•"
Ai«r- ''• S
'• . ! Mr r wui'EOM •
,:. V '#- .; 1
1°** '" " *ha*csri Cs*ssli'-""' -^-si1
/-->--=-- i - :^'~ — 1
' ""''" 1 «, "^ ,*""• *,'" , '~"~~\
\ -•-.. • .. • :-W ~r -MiC" x >
"*•• „ ""•'-.•, "v " - *"" *~ H a ^*T ** "i ;
1 '" '-- '^'fr ^''"---..f ' ' ^_^^^ ^ ' * mtm ~ ^*^i»r tiwily ^
: ^^^^^'•' _y--r; -••s_-,J_.^._ -H
V._. *" |
i i-, *ii..i - _j
".r*. * ;.,-,..
.<"ii«-;fc
_._f--- ^
i
1 ;rA. ! -«=* ^
; ~ i ^
a
-"» .- • "• E
i
- - - - ' ^^^J
! r-.'^^P
"" '" -''HwhUi* .,'htha
i
C '"" , X
'••• \ ...» f
•'>. .• ["
, •' ^ -PLUs^^
,"' \' ^"t
', , il J
-' 'I'"™1, . f
- '"L" I " '•" f\
• « ! •. _ ,_ — . r l
; -',rs^ ^...^r, _,' -->,
i • feiJLJ'1*^ **i**/" , ,, ,., v - v
r;~V---.E^:r l-.'--"^r-f*V *"
% 1 n.iBiiai ^^ j* ~ H fc
»". \ f*1; '" )ar«-r-
, ,, ' n>'*5«-*«ll««» M3r,j|^;,-l
! *l ""*! f , •* s- r ')
', «. - >^« « , \
O "X ^ >Tp^ "' - "
J ' * ^WS¥.^ ' "X. j / /
,',. .1 .- .. • fMifctt,- LMU J
""r 1_. :..,.." - •" "-'Birlm "'
; i . ; .^_» . _ ^ -T--1 ••
j ! *i-.,i...« ! , V •«•••- C«HI n«« ; .;''
"---.! -'"'"X 1 P ,, -*- •••*•
'•r ' v^r^t:. ;J i*-*->. - % ?
IT f £*i J %^' ?^i 1
H - nt e
"" """* - *.
I»H =* if* «w i*>-i »*»«** jfc
~f-,;«^a™= BW?«. w «"-•«*•* fMT
•»*— „ ,t ft.
| j i' «i.hiP,. l»"
1 '-"• _X " ^~
^^^^•^
n K ** at. K
If
I
~ ,tf
\
J
Majdr Cdal-Baanrlg Areas ni West&m Washington (C-haate el a!,, 19flO)
"»f.s-! »*'^' ^i u ' ' , F'liur*? A*^1 "3
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
All-10
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
c f° b
Hi
2. B- C
a i a
-
0 OQ
^ 3
3 §
O C3
-
5 o
OQ g
O O
^+J ft
C/l
•*&,.;* NIC
l tit,
iHQHWANINf
Stratigraphy for Three Coal Districts of the Pacific Coal Region (Choate el al.. I960)
' AIM
B
>
ON
k
6
4^
6
o
U)
The Pacifi
d Ce
o£
O p
f£, o
I* I"
cS S
o' rt
g -
image:
EPA816-R-04-003
Attachment 11
The Pacific and Central Coal Regions
Slfyetyrai of iho Central Columbia Basin aid Yakima Fold Bert
{Johnson et a!.,, 1993)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
All-12
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
REFERENCES
Biekman, H.M., H.D. Gower, and T.A.M. Dana. 1961. Coal Reserves of Washington.
Washington Division of Mines and Geology Bulletin 47, 115 p.
Buckovic. 1979. The Eocene Deltaic System of West-central Washington: in
Armentrout, J. M., Cole, M. R., and TerbestH., eds., Cenozoic paleogeography
of the western United States: Los Angeles, Soc. Econ. Paleont. and Min., Pacific
Section, Pacific Coast Paleogeog. Symposium 3, p. 147-163, as cited by Choate et
al., 1980.
Choate, R., Johnson, D.A., and McCord J.P. 1980. Geologic overview, coal, and
coalbed methane resources of the Western Washington coal region, Lakewood,
Colorado. TRW Energy Systems Group Report for U.S. Department of Energy,
Morgantown Energy Technology Center, Contract DE-AC21-78MC08089, pp.
353-372.
Dion, N. P. 1984. Washington Ground-Water Resources. In National Water Summary,
U.S. Geological Survey Water-Supply Paper No. 2275, pp. 433-438.
Evans, G. W. 1912. The coal fields of King County. Washington Geological Survey
Bulletin 3, 247 pp.
Gas Research Institute. 1999. North American coalbed methane resource map - U.S.
coalbed methane resources.
http://www.gri. org/pub/content/j un/19990614/114314/resources/resources. html.
Gas Technology Institute (GTI) Web site. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org.
Jenkins, Olaf P. 1923. Geological investigation of the coal fields of western Whatcom
County, Washington: Washington Division of Geology Bulletin 28, 135 p., 2
plates.
Jenkins, Olaf. P. 1924. Geological investigation of the coal fields of Skagit County,
Washington: Washington Division of Geology bulletin 29, 63 p.
Johnson, V. G., D. L. Graham, and Reidel, S. P. 1993. Methane in Columbia River
basalt aquifers: isotopic and geohydrologic evidence for a deep coal-bed gas
source in the Columbia Basin. Washington Bulletin of the American Association
of Petroleum Geologists, 77(7): 1192-1207 (July).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -13
image:
EPA 816-R-04-003 Attachment 11
The Pacific and Central Coal Regions
Lindholm, G.F. and Vaccaro, JJ. 1988. Region 2, Columbia Lava Plateau. In The
Geology of North America, Vol. 0-2, Hydrogeology. The Geological Society of
America, Boulder CO, pp. 37-50.
Pappajohn, S. P., and Mitchell, T. E. 1991. Delineation of prospective coalbed methane
trends in western and central Washington State. In Schwochow, S. D. Coalbed
methane of western North America guidebook for the Rocky Mountain
Association of Geologists Fall Conference and Field Trip Sept. 17-20, 1991,
Glenwood Springs, CO, pp. 163-178.
Quarterly Review of Methane from Coal Seams Technology. 1993. Pacific Coal Region.
Methane from Coal Seams Technology, pp. 21 (August).
Reidel, S. P., Fecht, K. R., Hagood, M. C., and Tolan, T. L. 1989. Geologic
development of the central Columbia plateau, in Riedel, S. P. and Hooper, P. R.,
eds., Volcanism and tectonism in the Columbia River Flood-Basalt Province:
Geological society of America Special Paper 239, pp. 247-264.
Snavely, P. D., Jr., Brown, R. D., Jr., Roberts, A. E., and Rau, W. W. 1958. Geology
and coal resources of the Centralia-Chelhalis district, Washington: U.S.
Geological Survey Bulletin 1053, 159 p.
Vine, J. D. 1969. Geology and coal resources of the Cumnberland, Hobart, and Maple
Valley quadrangles, King County, Washington: U.S. Geological Survey
Professional Paper 624, 67 p.
Walker, C. W. 1980. Geology and energy resources of the Roslyn-Cle Elum area,
Kittitas County, Washington: Washington Department of Natural Resources
open-File Report 80-1, 57 p.
Walsh, T. J. and Lingley, W. S., Jr. 1991. Coal maturation and the natural gas potential
of western and central Washington: Washington Division of Geology and Earth
Resources Open-File Report 91-2, 26 p.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A11 -14
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Attachment 2
The Black Warrior Basin
The Black Warrior Basin covers an area of about 23,000 square miles in Alabama and
Mississippi. The basin is approximately 230 miles long from west to east and
approximately 188 miles long from north to south. Coalbed methane production in
Alabama is limited to the bituminous coalfields of west-central Alabama, primarily in
Jefferson and Tuscaloosa Counties.
Coalbed methane production in the Black Warrior Basin is among the highest in the
United States. In 1996, approximately 5,000 coalbed methane wells were permitted in
Alabama. In 2000, this number increased to over 5,800 wells (Alabama Oil and Gas
Board, 2002). Coalbed methane well production rates range from less than 20 to more
than one million cubic feet per day per well (Alabama Oil and Gas Board, 2002).
Between 1980 and 2000, coalbed methane wells in Alabama produced roughly 1.2 trillion
cubic feet of gas. According to the Gas Technology Institute (GTI), annual gas
production was 112 billion cubic feet in 2000 (GTI, 2002).
2.1 Basin Geology
Coalbed methane production in the Black Warrior Basin (Figure A2-1) is contained
within the Upper Pottsville Formation of Pennsylvanian age (300 million years). The
deposit!onal history along the ancient coastline of prehistoric Alabama was characterized
by 8 to 10 "coal deposition cycles" of sea level rising and lowering. Each of these 10
geologic "coal deposition cycles" features mudstone at the base of the cycle (deeper
water) and coalbeds at the top of the cycle (emergence) (Pashin and Hinkle, 1997).
The geologic structure of the Black Warrior Basin is complex. Due to erosion and
structural uplift, not all of the coal zones are present at all locations (Pashin et al., 1991;
Young et al., 1993). In general, however, most coalbed methane wells tap the Black
Creek/Mary Lee/Pratt cycles, at depths that range from 350 to 2,500 feet deep (Hoiditch,
1990).
Alabama coalbeds are typically very thin, ranging from less than 1 inch in thickness to 4
feet (in rare cases they may be up to 8 feet thick in surface mines) (Horsey, 1981; Heckel,
1986; Eble et al., 1991; Carrol et al., 1993; Pashin, 1994) (Figure A2-2). In the area of
coalbed methane development, the Pottsville Formation exists at or near the surface, and
the depth to commercial coalbeds ranges from the surface outcrop to 3,500 feet,
depending on location (Figure A2-3).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-1
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
2.2 Basin Hydrology and USDW Identification
In the location where coalbed methane development is taking place in west-central
Alabama, the Pottsville Formation is an unconfmed aquifer. The matrix permeability of
Pottsville rocks (e.g., mudstone, cemented sandstone) is low, but water is present and
flows within an extensive system of faults, fractures, and joints. Flow patterns within the
Pottsville Formation are strongly controlled by fault- and fold-related isotropic joints and
fractures (Koenig, 1989). The close spacing and systematic pattern of cleats, however,
make coalbeds the most productive aquifers within the Pottsville Formation (Koenig,
1989; Pashin et al., 1991; Pashin and Hinkle, 1997). In the early 1990s, several authors
reported fresh water production from coalbed wells at rates up to 30 gallons per minute
(Ellard et al., 1992; Pashin et al., 1991).
Most of the recharge to the Pottsville aquifer is precipitation that infiltrates from the
surface, but some recharge occurs where streamflow enters the outcrop and moves
laterally into the aquifer along folded anticlinal beds (Pashin and Hinkle, 1997) (Figure
A2-4). Several researchers also propose upwelling of more saline waters from deeper
zones, which takes place along vertical, fault-related, rubble zones (Pashin et al., 1991).
Discharge from the Pottsville aquifer is primarily from the dewatering of coalbeds due
to mining and coalbed methane production (Pashin et al., 1991).
Formation water produced from Alabama coalbed methane wells contains between less
than 50 to over 10,000 milligrams per liter (mg/L) total dissolved solids (TDS) (Koenig,
1989; Pashin et al., 1991; Pashin and Hinkle, 1997). Some portions of the Pottsville
Formation contain waters which meet the quality criterion of less than 10,000 mg/L TDS
for an underground source of drinking water (USDW) (Figure A2-7). According to the
Alabama Oil and Gas Board, some waters in the Pottsville Formation do not meet the
definition of a USDW and have TDS levels which are considerably higher than 10,000
mg/L (Alabama Oil and Gas Board, 2002). Water quality generally decreases with
increasing depth (Figures A2-7 and A2-8), and areally is related to the faulting pattern
(Figure A2-9) (Pashin et al., 1991; Pashin and Hinkle, 1997). Waters exceeding 10,000
mg/L TDS can be found below 3,000 feet in areas near deep vertical faults, suggesting
upwelling from deeper, more saline zones (Pashin and Hinkle, 1997).
2.3 Coalbed Methane Production Activities
Alabama coalbed methane wells are categorized into three distinct types. The first two
types, "gob" wells and horizontal wells, are less common. Gob wells are associated with
mines. The well is drilled to a depth above the mine roof, and when the mine is
abandoned, the roof collapses. Gob wells produce coalbed methane from the fractured
mine debris. A few horizontal wells are drilled from within mines to reduce coalbed
methane concentration in advance of a working face. The third type, which includes 98
percent of all Alabama methane wells, includes vertically drilled wells that utilize
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-2
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
mainstream oilfield technologies (Pashin and Hinkle, 1997). Because neither gob nor
horizontal wells typically are hydraulically fractured, this discussion is limited to vertical
wells.
According to literature, most coalbed methane wells are drilled using water or air rotary
methods or water-based mud, due to lower cost and concerns that mud fluids will invade
the coal. Wells in Alabama are completed with tubing. Water is pumped up the tubing
for disposal, whereas gas is produced up the annulus. Wells are drilled to a depth 10 to
30 feet below the lowest coalbed to create a sump that collects coal fines and allows
water to separate from the coalbed methane (Hoiditch, 1990).
About 95 percent of produced water is disposed by discharge into surface water, via Type
II National Pollution Discharge Elimination System permits (O'Neil et al., 1989; O'Neil
et al., 1993; Pashin and Hinkle, 1997). These permits require some water quality
monitoring and limit instream water quality to 230 mg/L TDS (Pashin and Hinkle, 1997).
Since 1991, about 5 percent of produced water has been injected for disposal into Class II
injection wells (Pashin and Hinkle, 1997). Eight Class II wells are currently active
(Alabama Oil and Gas Board, 2001), disposing coalbed waters into zones between 4,300
and 10,000 feet deep (Ortiz et al., 1993).
Most wells are completed in multiple coal zones using perforations. Some wells are
completed in mudstones immediately below a coal zone, rather than within the coal
("limited entry" completions), and a few wells feature un-cased, open-hole completions.
Each well is hydraulically fractured to allow communication with the thin coal seams
outside of the casing, and most wells are fractured more than once as described below:
• In wells with multiple coal seams present, the hydraulic fracturing process
may involve several or multiple stimulations, using 2 to 5 hydraulic fracture
treatments per well (depending on the number of seams and spacing between
seams); and,
• Many coalbed methane wells are re-fractured at some time after the initial
treatment, in an effort to re-connect the wellbore to the production zones to
overcome plugging or other well problems (remedial fracture-stimulation)
(Holditch, 1990; Saulsberry et al., 1990; Palmer et al., 1991a and 1991b;
Schraufnagel et al., 1991; Holditch, 1993; Palmer et al., 1993b; Spafford et
al., 1993; Schraufnagel et al., 1993) (Figure A2-10).
The geometry of hydraulic fractures in coalbed methane zones usually differs from that
observed in conventional oil and gas scenarios. In conventional hydrocarbon zones, the
gas and/or oil are physically "trapped" by the presence of an impermeable confining
layer, usually shale. Shale formations may present a barrier to upward fracture growth
because of the stress contrast between the coalbed and the higher-stress shale (see
Appendix A). Therefore, for conventional fracturing, the vertical growth of fractures out
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-3
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
of the target zone may be limited by the presence (i.e., stress contrast) of overlying
shales. In conventional gas-well fracture environments, fracture half-length (200-1,600
feet from the well bore) almost always exceeds fracture height (10-200 feet above the
perforations). In the Black Warrior Basin, however, the lithologic properties and stress
fields of the coal cycles typically produce fractures that are higher than they are long
("length" refers to horizontal distance from the well bore) (Morales et al., 1990; Zuber et
al., 1990; Holditch et al., 1989; Palmer and Sparks, 1990; Jones and Schraufnagel, 1991;
Steidl, 1991; Wright, 1992; Palmer et al., 1991b and 1993a).
In the Black Warrior Basin of Alabama, hydraulic fractures created in coalbed methane
deposits are able to grow much higher than some fractures in "conventional" gas
reservoirs. There are three primary reasons for this phenomenon:
• Due to coal's low modulus of elasticity (i.e., brittleness, stiffness) and
complex fracture geometries, high pressures are required to fracture coal
hydraulically (500 to 2,000 pounds per square inch (psi), or 0.7 to 2.0 psi/ft),
and high treatment pressure often causes preferential extension of the fracture
in the vertical dimension (Jones et al., 1987; Reeves et al., 1987; Morales et
al., 1990; Palmer et al., 1991a);
• The economics of coalbed methane production in this basin requires tall
fractures that penetrate several coal seams. The coal seams are typically thin
(1 to 12 inches) and economically viable production requires the drainage of
as many seams as possible. Because coal seams may be vertically separated
by up to hundreds of feet of intervening rocks, operators usually design
fracture treatments to enhance the vertical dimension and might perform
several fracture treatments within a single well (Ely, et al., 1990; Holditch,
1990; Saulsberry etal., 1990; Spafford, 1991; Holditch, 1993); and,
• The other rocks within the Pottsville coal cycles (jointed mudstone and
sandstone) fracture much more easily than coal (Teufel and Clark, 1981;
Saulsberry etal., 1990; Jones and Scraufnagel, 1991; Spafford, 1991).
Because there are no significant barriers to fracture height (Simonson et al.,
1978; Ely et al., 1990; Palmer et al., 1991a), vertical fractures in the Black
Warrior basin typically penetrate several thin coalbeds and hundreds of feet of
intervening rocks (Teufel and Clark, 1981; Hanson et al., 1987; Holditch et
al., 1989; Ely et al., 1990; Palmer et al., 1991c; Schraufnagel et al., 1991;
Spafford, 1991; Palmer et al., 1993b) (Figure A2-11).
Mined-through studies in the Black Warrior Basin identified many instances where thin
(less than 1-foot thick) shales overlying targeted coalbeds were fractured. Penetration
into layers above the coal was observed in more than 80 percent of the fractures
intercepted by mines underground in the Black Warrior Basin (Diamond, 1987b). Some
fractures continued completely through very thin shales (Diamond, 1987a and b). These
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-4
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
studies did not conduct a systematic assessment of the extent of the vertical fractures
through and above the roof rock shales.
Several researchers conclude (based on pressure behavior during fracturing and several
examples where mines penetrated hydraulic fractures) that shallow fractures have a
horizontal component as described below:
• Fractures that are created at shallow depth typically have more of a horizontal
component and less of a vertical component. The vertical component is most
likely due to the presence of vertical natural fractures (cleats and joints) as
pre-existing planes of weakness from which vertical fractures can initiate.
• Fractures created at a greater depth can propagate vertically to shallower
depth, and develop a horizontal component. In these "T-fractures", the
fracture tip may fill with coal fines and/or intercept a zone of stress contrast,
which causes the fracture to "turn" and to develop horizontally.
As noted above, penetration of the layers above the coal was observed in more than 80
percent of the fractures intercepted by mines underground in the Black Warrior Basin
(Diamond, 1987b), but, as coals become shallower, the potential for fracture height
growth decreases. In general, horizontal fractures are most likely to exist at shallow
depths (less than 1,000 feet). As depths increase, it is more likely that a simple vertical
fracture will occur (Gas Research Institute, 1995).
Sand is the most common proppant used in coalbed methane treatments in Alabama. The
amount of sand injected per fracture treatment ranges from 10,000 to 120,000 pounds
(Holditch et al., 1989; Palmer et al., 1991b and 1993a). Fracture widths in the formation
vary from 0.5 inches to closed (i.e., no proppant emplaced), depending on distance from
wellbore and efficiency of the proppant displacement into the length of the fracture
(Palmer and Sparks, 1990; Palmer et al., 1993a; Steidl, 1993).
Fracturing fluid (30,000 to 200,000 gallons per treatment) is injected at a rate of 5 to 50
barrels per minute (which equals 210 to 2,100 gallons per minute) at injection pressures
ranging from 500 to 2,300 psi (Palmer et al., 1989 and 1993b; Holditch et al., 1989;
Pashin and Hinkle, 1997). The most common constituent of fracturing fluid is plain
water. Several researchers conclude that approximately 75 percent of all coalbed
methane wells in Alabama were fractured using cross-linked gel fluids (Palmer et al.,
1993a; Pashin and Hinkle, 1997).
According to service companies, diesel fuel is no longer used as a component of
fracturing fluids in Alabama. In addition, additives that could introduce chemicals
exceeding maximum contaminant levels (MCLs) are no longer used in fracturing fluids in
Alabama.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-5
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Table A2-1 presents some data concerning the general chemical makeup of common
fracturing fluids used in Alabama from literature published prior to the Alabama
hydraulic fracturing regulation (Economides and Nolte, 1989; Penny et al., 1991). In
addition, most gel fluids utilize a breaker compound (usually a borate or persulfate
compound or an enzyme, at 2 lb/1,000 gal) to allow post-treatment thinning and easier
recovery of gels from the fracture. Several researchers conclude that approximately 75
percent of all coalbed methane wells in Alabama were fractured using cross-linked gel
fluids (Palmer et al., 1993a; Pashin and Hinkle, 1997).
According to Hunt and Steele (1992), environmental regulations restrict local disposal of
used fracturing fluids, and fracturing fluids are transported to regulated disposal sites.
Robb and Spafford (1991) reported that acids were used to fracture production zones as
shallow as 400 feet deep.
In fracture treatments of wells in homogeneous formations in conventional gas fields,
injection is temporary and the majority of fracturing fluid is subsequently pumped back
up through the well when production resumes.
There are limited data in the literature concerning the volume of fracturing fluids
subsequently pumped back to the well after stimulation has ceased. Palmer et al. (1991b)
found that only 61 percent of fracturing fluids were recovered during production
sampling of a coalbed well in the Black Warrior Basin of Alabama, and projected that 20
to 30 percent would remain in the formation.
2.4 Summary
Coalbed methane development and hydraulic fracturing in the Black Warrior Basin of
Alabama takes place within a USDW, the Pottsville formation. Some portions of the
Pottsville Formation contain waters which meet the quality criteria of less than 10,000
mg/L TDS for a USDW. Some waters in the Pottsville Formation do not meet the
definition of a USDW and have TDS levels that are considerably higher than 10,000
mg/L (Alabama Oil and Gas Board, 2002).
According to service companies, diesel fuel is no longer used as a component of
fracturing fluids in Alabama. In addition, additives that could introduce chemicals
exceeding MCLs are no longer used in fracturing fluids in Alabama.
In the Pottsville Formation, the lack of a significant vertical barrier can provide for
extensive fracture height growth (Holditch et al., 1989; Lambert et al., 1989; Ely et al.,
1990; Saulsberry et al., 1990; Palmer and Sparks, 1990; Spafford, 1991; Palmer et al.,
1991a and 1993a; Spafford et al., 1993; Gas Research Institute, 1995). Mined-through
studies in the Black Warrior Basin identified many instances where thin (less than 1-foot
thick) shales overlying targeted coalbeds were fractured. Penetration into layers above
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-6
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
the coal, which are typically shale, was observed in more than 80 percent of the fractures
intercepted by mines underground in the Black Warrior Basin (Diamond, 1987b).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-7
image:
EPA816-R-04-003
Attachment 2
The Black Warrior Basin
Table A2-1. Chemical Components Previously Used in Typical
Fracturing/Stimulation Fluids for Alabama Coalbed Methane Wells
Type of
Stimulation Fluid
Fluids
Hydrochloric acid
"Slick" water
Diesel oil
Composition
15% HC1 water solution
water-soluble solvent as
viscosity reducer (% unknown)
NA
pH
NA
NA
Gels1
R-F
Pfizer Flocon 4800
Marathon MARCH
DuPont LuDox SM
CPAM crosslinked with
Pfizer Floperm 500
Drilling Specialties
HE-100 Crosslinked
3% resorcinol, 3% formaldehyde,
0.5% KC1, 0.4% NaHCO3
3+
0.4% xanthan, 154 ppm Cr:
(asCrC!3), 0.5%KC1
1.4% polyacrylamide (HPAM), 636 ppm
Cr3+(as acetate), !%NaCl
10% colloidal silica, 0.7% NaCl
0.4% cationic polyacrylamide (CPAM),
1520 ppm glyoxal, 2% KC1
0.3% HP AM-AMPS, 100 ppm Cr
(as acetate), 2% KC1
3+
6.5
4.0
6.0
8.2
7.3
5.0
1 Gels are typically mixed at a ratio of 40 Ibs. per 1000 gal. water; compositions shown
are "as mixed".
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
image:
EL a EL
C? 3. G
ISi
R ??5
o p. d
s'il
"'aq
I |
3. cn
OQ g
O O
en
1
K»
O
O
r •
i
Bocrti* Crank —_£j
. _j, ___.„,_,
; Wolf Cr*ek
S 8 5 10 15fcUomM<*
Coalbed Methane Fields of Alabama (Pashin and Hinkle, 1997)
oo
K^
ON
O
4^
6
o
H
Cd
I
^>
Cd
fa
FflU-eA2-1
image:
EPA816-R-04-003
Attachment 2
The Black Warrior Basin
Cat
Uttofegy
if _ ' «*
Curry 12-10-37D
C**w Co**
E*
z
?
^~
«.»»fcl.7
'—3 f<r
Cfww
4J'
JTjorTipson
i^ ~—IK~J < :j-
B'lr
«.<r
•
TOomiit -
«.r
taeo
Rock lypM
bi-Z-Z-I-rf HrfMong
»
(lilharanilo)
Coal
J,.
GaffiM H;iy
t MlS**f "-4 --[: :njl
MickCm* Js "2 -
ejf«1* C
* Ertr«cl*st»
Coal Cycles of the Pottsville Formation in the Black Warrior Basin (Pashin and Hinkle, 1997)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A2-10
image:
EL a EL
C? 3. G
r-K ^
&T^
o & C
s'il
R'oq
l|
3. en
OQ g
O O
C/l
> i
V o
i—1 O
A
Southwest
Moll li**
half
4ML4JLJI
Symbols
Cydf bwmdary
Well
(arrow
eyelti
amt
§w - Gnwn cyde
cb- Cobbeyd*
p. Pfat
c - Curfy cyete
g -
mi - toe
obc - Creek
Itsc > Or eefc and
Indti
Cross-Section of the Pottsvflle Formation in the Deerlick Creek Field (Pashin and Hinkfe, 1997)
K
oo
K^
ON
O
4^
6
o
H
Cd
I
^>
Cd
fa
image:
EPA816-R-04-003
Attachment 2
The Black Warrior Basin
i— i —=i
VCBTICtt EM£GEnMK»l IS M
ex PLAN* now
377.1 -C
CO AS. BE I) METHANE WILL
T
Y/A FAULT
C SURFACE
E TOS> OF WATER CONTAIN! NQ
»ftd, TOS
WATEfl
FLOW OF SU3SURFACE WATER
OAl'tM: MEAN SEA LEVEE.
B TOP Of CVCtS
U TOP OF UTiE* CYCLE
6 TOP OP <5WH CYCLE
C TOP OP C-OiS C VCli.
P TOP OF PFWTT CYCiE
®c TOP OF on tffSPYicunny cvcLf
ML TOP OF MARY L EE CYCLE
1C TOP Of CYCLE
Hydrogedogic Cross-Section of the Pottsville Formation in the Brookwood Field
Figure 2-1 for Field Location) (Pashin and Hinkle; 1997)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A2-12
image:
EL a EL
C? 3. G
r-K ^
&T^
2. B- C
a i a
fj- fO
0 OQ
^ 3
3 §
o id
-
5 o
OQ g
O O
«5 a>
VI
^ I
i O
i—1 O
%»-ptf
EXPLANHTIOM \*
FWiT, TICK
ON fflOE
I
A. i
i. I
^ '
0. H<XT ' I
E-
F. OEERLICK
O.
II
i.
a
I I S I I
10
IDS concentration of Pottsville Aquifer, B lack Warrior Basrn, Alabama (Pashrn and Hinkle, 1997}
>
oo
K^
ON
O
o
H
Cd
I
^>
Cd
fa
7ri4ni kit
image:
EL a EL
C? 3.
r-K ^
&T^
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
3.
3
OQ
o
i
TOS CdflKwt (m»l)
tJntfsrgrebrrf oaaj tWM
IDS Concentration of the Mary Lee Coal Seam of the Pottsvllle Aquifer in the
Ea^ern Part of the Stock Warrior Basin, Alabama (Pashin et al., 1991)
oo
K^
ON
O
4^
6
o
H
Cd
I
<?>
td
fa
7/14O1 Ilk
image:
EL a EL
C? 3. G
r-K ^
&T^
2. B- C
a i a
fj- fO
0 OQ
^ 3
3 §
O C3
-
5 o
OQ g
o o
«5 ft
VI
1
K»
O
o
UJ
II.
oc
u
m
Q
o
a.
tu
Q
UJ
y.
10
30
I
*
*
** •
0
2 4 § 8
10
in TDS with in the
et 3!., 1991)
rjar F^ws A2»7
oo
K^
ON
O
4^
6
o
H
Cd
I
<?>
Cd
fa
image:
EL a EL
C? 3.
r-K ^
tf^
2. B- C
a i a
ti- o
0 OQ
^ 3
3 §
o id
-
3.
s
OQ
o
ON
1
K<
O
O
Na+fC 1 O
so3 Surface
SO4
(150 1
a Oak fieW {1,104 ft)
OaJc field (1,188 ft)
Grove field ft)
ct
Blye field ft)
ft}
field ft)
Cedar ffelcf (3,115 ft)
'Stiff Diagram of Water Salinity (Pashln et al.. 1991}. The diagrarr indicates that it'.ost %ater samples contain than 10.000 rrg/L
IDS (except near a fcwfaulls), and the salinity generally increases with depth in the Pottsvilb Aquifer, Alabanra. Note that as n'g/L of NaCI.
the ' Equivalent per Liter (eprr.f units of the table correspond to 233? rrg/L {40 ejHrJ, 4675 rrg/L (BO eprr), 7012 rrfl/L (120 eptr),
9350 rrrg? L (180 eprr), and 11.888 rrg/L (160 eprr). NaCIis, by far, the dorrinant species af rrost Pottsvilb water sarrtples,
and rp^/L as NaCI dasely approxirrates ft:g/L TDSs
oo
K^
ON
O
4^
6
o
H
Cd
I
<?>
Cd
fa
image:
EL a EL
C? 3. G
s.R-1.
p p? t1
rt ft" nS
I a"
^3 v?^ O
o r** • r~^
s'il
"'cro
r?|
3. en
OQ g
O O
^+J ft
C/l
> i
V o
i—1 O
ft 8 W.
R. 6W.
ft. S W,
DATA powr
v
•
SWClJ«i
MO»CUN£
FAULT; res MARX
ON DQWffFFKJWS S*E_;
CWC
tNTERVAL a »,,&Qfl J%U
Afeu£C8Eta<AmtcLwe L,JS
8. AilPWANT ruaottcww C0|*rf !
C, " «
ft SBOiWaiffiMfClW
Relationship of Water Salinity to Structural Features in the Oak Grove Ftefdj
Black Warrior Basin, Alabama (Pashin and Hinkle, 1997),
In general, water containing more than 3,000 rrsg/L TDS is related to upweing of
saline water near a few geologic faults
oo
K^
ON
O
4^
6
o
H
Cd
I
<?>
Cd
fa
Fig we A2-9
image:
EL a EL
C? 3. G
ro B ?5
P. R- £••
'
o p. c;
s'il
"'cro
l|
3. en
OQ f=
O O
^+J ft
VI
^ I
I O
i—1 O
Field*
Oak Grove
Deerlick Creek
( Lambert et al,
1987/1990)
Cedar Cove
(Sparks and
Richardson. 1991)
Mound ville
(Elv ct al,
Productive Coal
Number of
Producing
Groups
2 or 3
"?
4
Up to
h or 7*
Seams and th
Black Warrior
Depth Number
Coal Range of Separate
Groyp (ft) Stimulations
Pratt SOO--2500 2 yr 3
Klarv Lee**
BLick Creek
UlOO-oOOO 3
Cobb 2000-3500 4
Pratt
Marv Let-**
Black Creek
Brookwood 3000-5000 3-6
Utlev
Gwjn
Cobb
Pratt
Marv Lee**
Creek
e Typical Number of Stimulations Per Well as of 1993,
Basin, Alabama (Palmer et al., 1993)
Tt-mm m m2Hrf-mfn.-o,m? FlpliB A2-10
oo
K^
ON
O
4^
6
o
H
Cd
I
<?>
Cd
fa
image:
EL a EL
C? 3.
r-K ^
&T^
2. B- C
a i a
ti- o
0 OQ
^ 3
3 §
O C3
3.
s
OQ
o
Directbn of
Fracture Propgation
oo
K^
ON
O
4^
6
o
H
Cd
I
1
K»
O
O
Sketch of a Vertical, Unconfined Fracture Typical of the Black Warrior Basin (Palmer et al,, 1991)
Cd
fa
Figure A2-11
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
REFERENCES
AAPG = American Association of Petroleum Geologists
SPE = Society of Petroleum Engineers
Alabama Oil and Gas Board, 2001. Alabama OGB staff, personal communication.
Alabama Oil and Gas Board, 2002. Public Comment OW-2002-0002-0029 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002
Carrol, R.E., Pashin, J.C., and Kugler, R.L. 1993. Burial history and source rock
characteristics of Mississippian and Pennsylvanian strata, Black Warrior Basin,
Alabama. Alabama Geological Society Guidebook, pp. 79-88.
Diamond, W.P. 1987a. Underground observations of mined-through stimulation
treatments of coalbeds. Quarterly Review of Methane from Coal Seams
Technology, v. 4, n. 4 (June 1987), pp. 19-29.
Diamond, W.P., 1987b, Characterization of fracture geometry and roof penetrations
associated with stimulation treatments in coalbeds; Proceedings of the 1987
Coalbed Methane Symposium, University of Alabama (Tuscaloosa), p. 243.
Diamond, W.P. and D.C. Oyler. 1987. Effects of stimulation treatments on coalbeds and
surrounding strata, evidence from underground observations. US Department of
Interior, RI9083, USBM, pp. 1-47.
Eble, C.F., Gastaldo, R.A., Demko, T.M., Liu, Y. 1991. Coal compositional changes
along a swamp interior to swamp margin transect in the Mary Lee coalbed,
Warrior Basin, Alabama USA. Proceedings of the Eighth Annual Meeting of the
Society for Organic Petrology, pp. 3-8.
Ellard, J.S., Roark, R.P., and Ayers, W.B. 1992. Geologic controls on coalbed methane
production: an example from the Pottsville formation, Black Warrior Basin,
Alabama USA. Symposium on Coalbed Methane Research and Development in
Australia. James Cook University, p. 45-61.
Ely, J.W., Zubitowski, R.L., and Zuber, M.D. 1990. How to develop a coalbed methane
prospect: a case study of an exploratory five-spot well pattern in the Warrior
Basin, Alabama. Proceedings, 1990 Society of Petroleum Engineers Annual
Technical Conference and Exhibition (Production Operations and Engineering),
pp. 487-496.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-20
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Gas Research Institute (GRI), 1995, Fracturing Experience at the Rock Creek Multiple
Coal Seams Project; Topical Report, prepared by S.W. Lambert, J.L. Saulsberry,
P.P. Steidl, M.W. Conway, and S.D. Spafford, July 1995.
Gas Technology Institute (GTI) Web site, 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Hanson, M.E., Neilsen, P.E., Sorrels, G.G., Boyer, C.M., and Schraufnagel, R.A. 1987.
Design, execution, and analysis of a stimulation to produce gas from thin multiple
coal seams. SPE Paper No. 16860, Proceedings, 1987 Society of Petroleum
Engineers Annual Technical Conference and Exhibition.
Heckel, P.H. 1986. Sea-level curve for Pennsylvanian eustatic transgressive-regressive
depositional cycles along mid-continent outcrop belt, North America. Geology,
14:330-334.
Hill, David. 2001. Gas Technology Institute (GTI). Expert peer-review panelist, personal
communication.
Holditch, S.A., Ely, J.W., Semmelbeck, M.E., Carter, R.H., Hinkle, 1, and Jeffrey, R.G.
1989. Enhanced recovery of coalbed methane through hydraulic fracturing. SPE
Paper No. 18250, Proceedings 1988 SPE Annual Technical Conference and
Exhibition (Production Operations and Engineering), p. 689.
Holditch, S.A 1990. Completion methods in coal seam reservoirs. SPE Paper No.
20670. Proceedings 1990 SPE Annual Technical Conference and Exhibition
(Production Operations and Engineering), pp. 533-542.
Holditch, S.A., 1993. Completion methods in coal-seam reservoirs. Journal of Petroleum
Technology, March 1993.
Horsey, C.A. 1981. Depositional environments of the Pennsylvanian Pottsville
formation in the Black Warrior Basin of Alabama. Journal of Sedimentary
Petrology, 51:799-806.
Hunt, A.M. and Steele, D.J. 1992. Coalbed methane technology development in the
Appalachian Basin. Quarterly Review of Methane from Coal Seams Technology,
pp. 15-17 (April).
Jones, A.H., Bell, G.J., and Morales, R.H. 1987. The influence of coal fines/chips on the
behavior of hydraulic fracture stimulation treatments. Proceedings of 1987
Coalbed Methane Symposium, University of Alabama (Tuscaloosa), Tuscaloosa,
pp. 93-102.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-21
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Jones, A.H., and Schraufnagel, R.A. 1991. In-situ stress variations in the Black Warrior
Basin. Proceedings of 1991 Coalbed Methane Symposium, University of
Alabama (Tuscaloosa), Tuscaloosa, p. 75.
Koenig, R.A. 1989. Hydrologic characterization of coal seams for optimum dewatering
and methane drainage. Quarterly Review of Methane from Coal Seams
Technology, 7:30-33.
Lambert, S.W., Graves, S.L., and Jones, A.H. 1989. Warrior basin drilling and
stimulation. Oil and Gas Journal, p. 19 (October 9, 1989).
Morales, R, H, McLennan, J.D., Jones, A.H., and Schraufnagel, R.A. 1990.
Classification of treating pressures in coal fracturing. Proceedings of the 31st U.S.
Symposium on Rock Mechanics, 31, pp. 687-694.
Naceur, K.B. and Touboul, E., 1990, Mechanisms controlling fracture height growth in
layered media; SPE Production Engineering, v.5 n.2 (May 1990), pp. 142-150.
O'Neil, P.E., Harris, S.C., Drottar, K.R., Mount, D.R, Fillo, J.P., and Mettee, M.F., 1989,
Biomonitoring of a produced water discharge from the Cedar Cove degasification
field, Alabama; Alabama Geological Society, Bulletin 135, 195 pp.
O'Neil, P.E., Harris, S.C., Mettee, M.F., Shepard, I.E., and McGregor, S.W., 1993,
Surface discharge of produced waters from the production of methane from coal
seams in Alabama; Alabama Geological Society, Bulletin 155, 259 pp.
Ortiz, I, Weller, T.F., Anthony, R.V., Frank, J., andNakles, D., 1993, Disposal of
produced waters: underground injection option in the Black Warrior basin;
Proceedings 1993 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), pp. 339-364.
Palmer, ID., Davids, M.W., and Jeu, S.F. 1989. Analysis of unconventional behavior
observed during coalbed fracturing treatments. Proceedings 1989 Coalbed
Methane Symposium, University of Alabama (Tuscaloosa).
Palmer, ID. and Sparks, D.P. 1990. Measurement of induced fractures by downhole TV
camera in coalbeds of the Black Warrior Basin. Society of Petroleum Engineers
Paper No. 20660, Proceedings, 1990 Society of Petroleum Engineers Annual
Technical Conference and Exhibition, pp. 445-458.
Palmer, ID., King, N.S., and Sparks, D.P. 1991a. The character of coal fracture
treatments in Oak Grove field. Black Warrior Basin, SPE Paper No. 22914,
Proceedings, 1991 Society of Petroleum Engineers Annual Technical Conference
and Exhibition, pp.277-286.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-22
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Palmer, ID., Fryar, R.T., Tumino, K.A., and Puri, R. 1991b. Comparison between gel-
fracture and water-fracture stimulations in the Black Warrior Basin. Proceedings
1991 Coalbed Methane Symposium, University of Alabama (Tuscaloosa), pp.
233-242.
Palmer, ID., Fryar, R.T., Tumino, K.A., and Puri, R. 1991c. Water fracs outperform gel
fracs in coalbed pilot. Oil and Gas Journal, pp.71-76 (August 12, 1991).
Palmer, ID., King, N.S., and Sparks, D.P. 1993a. The character of coal fracture
treatments in the Oak Grove field, Black Warrior Basin. In Situ, Journal of Coal
Research, 17(3):273-309.
Palmer, ID., Lambert, S.W., and Spitler, J.L. 1993b. Coalbed methane well completions
and stimulations. Chapter 14 of AAPG Studies in Geology 38, pp. 303-341.
Pashin, J.C., 1994, Coal body geometry and syn-sedimentary detachment folding in the
Oak Grove coalbed methane field, Black Warrior basin, Alabama; AAPG
Bulletin, v. 78, pp. 960-980.
Pashin, J.C., Ward, W.E., Winston, R.B., Chandler, R.V., Bolin, D.E., Richter, K.E.,
Osborne, W.E, and Sarnecki, J.C. 1991. Regional analysis of the Black Creek-
Cobb coalbed methane target interval, Black Warrior Basin, Alabama. Alabama
Geological Survey Bulletin 145, 127pp.
Pashin, J.C. and Hinkle, F. 1997. Coalbed Methane in Alabama. Geological Survey of
Alabama Circular 192, 71pp.
Reeves, S.R., Wallaca, J.A., and Beavers, W.M., 1987, The influences of reservoir
properties and geologic setting on coal bed fracturing and production. SPE Paper
16423, Proceedings, SPE/DOE Symposium on Gas Production from Low
Permeability Reservoirs.
Robb, J.C. and Spafford, S.D., 1991, Stimulations without proppant and resulting
production from four wells completed in the Pratt coal group near depths of 500
feet; Proceedings 1991 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa) (Tuscaloosa), pp. 381-389.
Saulsberry, J.L., Schraufnagel, R.A., and Jones, A.H. 1990. Fracture height growth and
production from multiple reservoirs. SPE Paper No. 20659, Proceedings, 1990
Society of Petroleum Engineers Annual Technical Conference and Exhibition, pp.
433-443.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-23
image:
EPA 816-R-04-003 Attachment 2
The Black Warrior Basin
Schraufnagel, R.A., Spafford, S.D., and Saulsberry, J.L. 1991. Multiple seam
completion and production experience at Rock Creek, Alabama. Proceedings
1991 Coalbed Methane Symposium, University of Alabama (Tuscaloosa), pp.
211-221.
Schraufnagel, R.A., Spafford, S.D., and Conway, M.W. 1993. Restimulation techniques
to improve fracture geometry and overcome damage. Proceedings, Gas
Technology Symposium, pp. 611-625.
Simonson, E.R., Abou-Sayed, A.S., Clifton, R.J., 1978, Containment of massive
hydraulic fractures; SPE Journal, February 1978, pp. 27-32.
Spafford, S.D. 1991. Stimulating multiple coal seams at Rock Creek with access
restricted to a single seam. Proceedings 1991 Coalbed Methane Symposium,
University of Alabama (Tuscaloosa), p. 243.
Spafford, S.D, Saulsberry, J.L., and Scraufnagel, R.A. 1993. Field verification of
fracture height growth associated with a restricted-access completion.
Proceedings 1993 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), pp. 139-144.
Steidl, P.F., 1991, Inspection of induced fractures intercepted by miningin the Warrior
basin, Alabama; Proceedings 1991 Coalbed Methane Symposium, University of
Alabama (Tuscaloosa), pp. 181-191.
Teufel, L.W. and Clark, J.A. 1981. Hydraulic fracture propagation in layered rock:
experimental studies of fracture containment. SPE Paper No. 9878, Proceedings
of the DOE/SPE Symposium: Gas Production from Low Permeability Reservoirs.
Wright, C.A. 1992. Effective design, real-data analysis, and post-job evaluation of
hydraulic fracturing treatments. Methane from Coal Seams Technology Journal,
pp. 29-32 (July).
Young, G.B., Paul, G.W., Saulsberry, J.L., and Schraufnagel, R.A. 1993.
Characterization of coalbed reservoirs at the Rock Creek project site, Alabama.
Proceedings 1993 Coalbed Methane Symposium, University of Alabama
(Tuscaloosa), pp. 705-714.
Zuber, M.D., Kuuskraa, V.A., and Sawyer, W.K. 1990. Optimizing well spacing and
hydraulic fracture design for economic recovery of coalbed methane. SPE
Formation Evaluation, 5(1): 98-102.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A2-24
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
Attachment 3
The Piceance Basin
The Piceance Coal Basin is entirely within the northwest corner of Colorado. (Figure A3-
1). The coalbed methane reservoirs are found in the Upper Cretaceous Mesaverde
Group, which covers about 7,225 square miles and ranges in thickness from about 2,000
feet on the west to about 6,500 feet on the east side of the basin (Johnson, 1989). It is
estimated that 80 trillion to 136 trillion cubic feet (Tcf) of gas are contained in coalbeds
within the basin (Tyler et al., 1998). Total coalbed methane production was 1.2 billion
cubic feet in 2000 (GTI, 2002).
3.1 Basin Geology
The Piceance is a northwest trending asymmetrical, Laramide-age basin in the Rocky
Mountain foreland with gently dipping western and southwestern flanks and a sharply
upturned eastern flank (Figure A3-1) (Tremain and Tyler, 1997). The Douglas Creek
Arch bounds the basin on the northwest, and separates it from the Uinta Coal Basin,
which lies almost entirely in Utah. The Mesaverde Group is sharply upturned to near
vertical along the Grand Hogback, which forms the eastern boundary of the basin and
separates the basin from the White River uplift to the east. Most of the Piceance Basin's
coal deposits are contained in the lies and Williams Fork Formations of the Late
Cretaceous Age Mesaverde Group, which are approximately 100 to 65 million years in
age (McFall et al., 1986). These formations composed of sandstone and shale, were
deposited in a series of regressive marine environments (McFall et al., 1986; Johnson,
1989). It is believed that the coals were deposited in marine transitional, brackish,
interdistributary marshes and freshwater deltaic swamps (Collins, 1976 in McFall et al.,
1986). Figure A3-2 presents a stratigraphic section shown with a gamma ray-induction
log from the Barrett 1-27 Arco Deep well (Reinecke et al., 1991). The Mesaverde Group
is underlain by the marine Mancos Shale and overlain by the lower Tertiary Age Fort
Union and Wasatch Formations, which consist of fluvial sandstones and shales. The
Mancos Shale, Fort Union, and Wasatch Formations are essentially barren of coals
(McFall et al., 1986). Depths to the coal-bearing sediments vary from outcrops around
the margins of the basin (Figure A3-1) to more than 12,000 feet in the deepest part of the
basin (Tyler et al., 1996).
The major fold structure of the Piceance Basin is the Grand Hogback Monocline, formed
as the White River Uplift was uplifted and thrust westward during the Laramide Orogeny
in Late Cretaceous through Eocene time (McFall et al., 1986). Broad folds, such as the
Crystal Creek and Rangley Syncline, trend northwest to southeast, and generally parallel
to the axis of the basin (Figure A3-1). Intrusions occur throughout the southeast part of
the basin, locally elevating coal ranks to as high as anthracite grade. A buried laccolith
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-1
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
intrusion is thought to be present under a coal basin anticline along the southeast margin
of the basin (Figure A3-1) where high quality coking coal was mined since the 1800s
(Collins, 1976).
Coalbed methane reservoirs occur exclusively in the Upper Cretaceous Mesaverde Group
(Figure A3-2), which covers an area of approximately 7,255 square miles (Tremain and
Tyler, 1997). Depths to the Mesaverde Group range from outcrop to greater than 12,000
feet along the axis of the basin (Tyler et al., 1996; Tremain and Tyler, 1997). Two-thirds
of the coalbed methane occurs in coals deeper than 5,000 feet, making the Piceance Basin
one of the deepest coalbed methane areas in the United States (Quarterly Review, August
1993).
The major coalbed methane target, the Cameo-Wheeler-Fairfield coal zone (Figure A3-
3), is contained within the Williams Fork Formation of the Mesaverde Group and holds
approximately 80 to 136 Tcf of coalbed methane (Tyler et al., 1998). This coal zone
ranges in thickness from 300 to 600 feet, and lies more than 6,000 feet below the ground
surface over a large portion of the basin (Tyler et al., 1998). Individual coal seams of up
to 20 to 35 feet thick can be found within the group, with net coal thickness of the
Williams Fork Formation averaging 80 to 150 feet thick. In 1991, at the Grand Valley
field (Figure A3-4), there were 23 coalbed methane wells and 18 conventional gas wells
(Reinecke et al., 1991). However, in 1984, most wells at the Rulison field (Figure A3-4)
were conventional gas wells.
Initially, it was anticipated that coalbed methane wells in the sandstones and coals of the
Cameo Zone would have high production rates of water. However, testing later showed
that they produced very little water (Reinecke et al., 1991). Both the sandstones and
coalbeds are tight, poorly permeable, and are generally saturated with gas rather than
water or a mixture of water and gas. The dynamic flow of a hydrologic system enhances
the collection of gas in traps, but in much of the Piceance Basin that flow is not present
because of the over-pressuring and saturation with gas.
Consequently, the conventional models for coalbed methane accumulation developed for
other basins do not apply well for exploration and development in the Piceance Basin.
Tyler et al. (1996) concluded, "very low permeability and extensive hydrocarbon
overpressure indicate that meteoric recharge, and, hence, hydropressure, is limited to the
basin margins and that long-distance migration of groundwater is controlled by fault
systems." Recharge is limited along the eastern and northeastern margins of the basin
because of offsetting faults, but zones of transition between hydropressure and
hydrocarbon overpressure in the western part of the basin and on the flanks of the Divide
Creek Anticline in the southeastern part of the basin may possess better coalbed methane
potential, as indicated by the exploration targets delineated in Tyler et al. (1998) (Figure
A3-5).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-2
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
3.2 Basin Hydrology and USDW Identification
The Piceance Basin contains both alluvial and bedrock aquifers. Unconsolidated alluvial
aquifers are the most productive aquifers in the Piceance Basin. These alluvial deposits
are narrow, and thin deposits of sand and gravel formed primarily along stream courses.
The City of Meeker, Colorado is supplied by wells tapping these deposits where they are
over 100 feet thick in the White River Valley (Taylor, 1987).
The most important bedrock aquifers are known as the upper and lower Piceance Basin
aquifer systems. These consolidated rock aquifers are lower Tertiary Eocene in age and
occur within and above the large oil shale reserves. The upper and lower aquifers are
separated by the Mahogany Zone of the Parachute Creek Member (Figure A3-6). The
Mahogany Zone is a poorly permeable oil shale, which retards water movement but does
not stop it. Both bedrock aquifers overlie the older Cretaceous Mesaverde Group where
the coal and coalbed methane are located.
The upper aquifer system is about 700 feet thick and consists of several permeable zones
in the Eocene Uinta Formation and the upper part of the Parachute Creek Member of the
Eocene Green River Formation. Sub-aquifers of the Uinta Formations are silty sandstone
and siltstone, while those of the Parachute Creek Member of the Green River Formation
are fractured dolomite marlstone. There is some primary porosity (i.e., the porosity
preserved from during or shortly after sediment deposition, such as the spaces between
grains) in the sandstone and the permeability of the sub-aquifers has been enhanced by
natural fracturing that occurred during post-deposition deformation. Layers between the
individual sub-aquifers are less permeable than the sub-aquifers themselves, but they do
not prevent water movement between the sub-aquifers.
The lower aquifer system is about 900 feet thick and consists of a fractured dolomitic
marlstone of part of the lower Parachute Creek Member of the Green River Formation. It
is semi-confined below the Mahogany Zone and above the Garden Gulch Member of the
Green River Formation and a high resistivity zone just above it (USGS, 1984 and Taylor,
1987) (Figure A3-6). Fracturing during deformation of the rocks and subsequent solution
enlargement owing to dissolution of soluble evaporite minerals has increased
permeability of this lower aquifer system.
Groundwater is recharged from snowmelt on high ground from where it travels down
through the upper aquifer system, the Mahogany Zone, and into the lower aquifer system.
The groundwater then moves laterally and/or upward discharging from both the upper
and lower aquifer systems into streams (Figure A3-7). The minerals nahcolite
(NaHCOs), dawsonite (NaAl(OH)2CO3) and halite (NaCl) are present in the groundwater,
and the circulation of the groundwater (with these minerals in solution) has caused
enlargement of the natural fractures (Taylor, 1987). Water in the lower aquifer is
reported to contain several hundred milligrams per liter (mg/L) of chloride (Taylor,
1987).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-3
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
Wells in these two bedrock aquifer systems, the upper and lower Piceance Basin aquifers,
typically range in depth from 500 to 2,000 feet and commonly produce between 2 to 500
gallons per minute of water (USGS, 1984). These Tertiary bedrock aquifers are
strati graphically separated from the base of the Cameo Coal Zone in the Cretaceous
Mesaverde Group by from less than 1,500 feet of strata along the Douglas Creek Arch to
more than 11,000 feet along the basin trough just west of the Grand Hogback (Johnson
and Nuccio, 1986) (Figure A3-2).
Aquifer maps do not exist for the Piceance Basin, but water quality in the Piceance Basin
is poor owing to nahcolite (sodium bicarbonate) deposits and salt beds within the basin
(Graham, 2001). Only very shallow waters such as those from the surficial Green River
Formation are used for drinking water (Graham, CDWR, 2001). In general, the potable
water wells in the Piceance Basin extend no further than 200 feet in depth, based on well
records maintained by the Colorado Division of Water Resources (CDWR). At least two
wells in the area are approximately 1,000 feet in depth, but they are used for stock
watering. A composite water quality sample taken from 4,637 to 5,430 feet deep within
the Cameo Coal Group in the Williams Fork Formation exhibited a total dissolved solid
(TDS) level of 15,500 mg/L, which is above the 10,000 TDS water quality criterion for a
underground source of water (USDW) (Graham, CDWR, 2001). The produced water
from coalbed methane extraction in the Piceance Basin is of such low quality that it must
be disposed of in evaporation ponds or re-injected into the formation from which it came
or at even greater depths (Tessin, 2001).
It is unlikely that any USDWs and coals targeted for methane production would coincide
in this basin. These targeted coals are generally located at great depth, of at least 4,000
feet. The thousands of feet of stratigraphic separation between the coal gas bearing
Cameo Zone and the lower aquifer system in the Green River Formation should prevent
any of the effects from the hydrofracturing of gas-bearing strata from reaching either the
upper or the lower bedrock aquifers.
Permeability of the coal and the surrounding sandstone and shale is generally quite low
except near outcrop, creating little potential for these rocks to contain a USDW.
Researchers (Reinecke et al., 1991) report that the permeability of gas-bearing coal and
sandstone of the Cameo Zone is so low that the gas is over-pressured and has forced
groundwater out of the zone, a condition that tends to disfavor the entrapment of
methane. Tyler et al. (1998) state that high coalbed methane gas productivity requires
geologic and hydrologic conditions, and that these conditions are not optimal throughout
much of the Piceance Basin because of the absence of dynamic groundwater flow and the
low permeability of the host rocks.
The above conditions prevail in the central part of the basin, previously favored as a
coalbed methane development fairway, and heavily targeted for exploration (Nowak,
1991). However, analyses by Tyler et al. (1998) suggest that a transitional zone, between
the deeply buried coal and the outcrops at the boundaries of the basin, where groundwater
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-4
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
circulation may be sufficient to create more favorable trapping conditions (Figure A3-5),
may be a better target area for coalbed methane production exploration. These
exploration target zones could possibly have sufficient meteoric groundwater circulation
to meet the water quality criterion of USDWs. However, Figure A3-3 shows that the
depths to coals in the targeted methane producing zones (Figure A3-5) are greater than
4,000 feet below ground surface and therefore, are not likely to contain water that would
meet the USDW quality criterion of less than 10,000 mg/L TDS. Currently, test-drilling
information is insufficient to determine if this is the case. Nevertheless, due to the very
low permeability, great depth, and expected poor water quality of the targeted coalbed
methane producing zones, conflicts with USDWs are considered to be of very low
probability.
3.3 Coalbed Methane Production Activity
Measurements of coal permeabilities in the Piceance Basin have shown that the deep
coals typical of the basin are much less permeable than coals in top-producing coalbed
methane basins such as the San Juan Basin in Colorado (Quarterly Review, 1993).
Consequently, operators rely on large hydraulic fractures to produce coalbed methane
from the deep, low permeability coals (Quarterly Review, 1993).
Exploration for coalbed methane began in the basin during the early 1980s, but viable
commercial production did not begin until 1989 (Quarterly Review, 1993). The first well
to commercially produce coalbed methane from the Piceance Basin, Exxon's Vega No. 2
well in Mesa County, went off-line in 1983 (Quarterly Review, 1993). Amoco
Production Company attempted multi-well coalbed methane development in the late
1980s, and finally ceased activity in 1989. Commercial production was finally achieved
in 1989 in the Parachute fields operated by Barrett Resources. Barrett Resources drilled
68 wells in 1990 and had planned for 22 more in 1991 (Western Oil World, 1991). The
wells targeted both coals and sandstone within the Cameo Coal Zone and the Mesaverde
sandstones, just above the Cameo coals. Other operators soon followed suit, including
Fuelco at White River Dome field in the northern part of the basin (Figure A3-1),
Conquest Oil Company near Barretts Resource's production in the central part of the
basin, Chevron USA Inc., and many others. However, not all operators were successful
in locating or producing coalbed gas. Ultimately, Barrett found the sandstones to be far
more productive than the coalbeds, and attempts to complete wells in the coalbeds were
largely abandoned.
According to the Colorado Geological Survey (2002), some operators are having success
in their pilot coalbed methane production program in White River Dome Field northwest
of Meeker. Their success is attributed to the extensive natural fracturing found in the
coal seams at White River Dome. Fracturing may be particularly extensive as a result of
the formation of the White River anticline and the proximity to the large Danforth Hills
Mesaverde outcrop. As a result, operators are taking another look at coalbed methane
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-5
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
development in the Piceance Basin. In addition, one of the operators is drilling (but not
fracturing) horizontal wells in the coal seams to take advantage of the anomalous natural
fracturing found at White River Dome field. In some areas of coalbed methane potential,
horizontal well technology may replace hydraulic fracturing as a method to enhance
coalbed methane well performance.
Within the Cameo Coal Zone, Barrett Resources typically used 3,000 to 3,500 barrels of
gelled 2% potassium chloride water with 273,000 to 437,000 pounds of sand over a
maximum 450 feet of the Cameo Coal Zone to stimulate coalbed methane wells
(Quarterly Review, 1993). It was shown that these hydraulic stimulations created short
(100-foot), multiple fractures around the wells (Quarterly Review, August 1993). Fuel
Resources Development Company used 3,000 to 10,000 barrels of gelled water and
200,000 to 1,300,000 pounds of sand to fracture their wells in the White River Dome
Field (Quarterly Review, 1993). All but one of Conquest Oil Company's wells was
hydraulically fractured with 1,500 barrels of water or cross-linked gel and 31,000 to
230,000 pounds of regular or resin-coated sand (Quarterly Review, 1993).
3.4 Summary
The Piceance Basin shows promise as a source for coalbed methane production based on
the estimated 80 to 136 Tcf of gas contained within the Cameo-Wheeler-Fairfield coal
zone (Tyler et al., 1998). However, overall low permeabilities as well as great depths to
coalbeds appear to have slowed coalbed methane development in the basin.
Nevertheless, a pilot program in White River Dome Field has had success in coalbed
methane production, attributable primarily to the extensive natural fracturing in the area.
As a result, operators are taking another look at coalbed methane development in this
basin.
Hydraulic fracturing is the common method used to extract coalbed methane. Drilling of
horizontal wells in the coal seams is a method that is being evaluated in the White River
Dome Field pilot project as an alternative to hydraulic fracturing. In some areas of
coalbed methane potential, horizontal well technology may replace hydraulic fracturing
as a method to enhance coalbed methane performance.
The fluids used for fracturing vary from water with sand proppant to gelled water and
sand. Between 1,500 to more than 11,000 feet of strata separate the coals from the
shallow USDWs, indicating that the potential for water quality contamination from
hydraulic fracturing techniques is minimal. The only hydraulic fracturing fluid
contamination pathway to the USDWs might be through faults or fractures extending
between the deep coal layers and the shallow aquifers. The occurrence of these fractures
and faults has not been substantiated in any of the literature examined for this
investigation.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-6
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
Research suggests that exploration may target areas where groundwater circulation may
enhance gas accumulation in the coal and associated sandstones (Tyler et al., 1998).
Under these exploration and development conditions, a USDW located in shallower
Cretaceous rocks near the margins of the basin, could be affected by hydraulic fracturing.
The depth to methane-bearing coals (about 6,000 feet) seems to indicate that, in the
Piceance Basin, the chances of contaminating any overlying, shallower USDWs (no
deeper than 1,000 feet) from injection of hydraulic fracturing fluids and subsequent
subsurface fluid transport are minimal. Potable wells in the Piceance Basin generally
extend no further than 200 feet in depth. The coalbed methane producing Cameo Zone
and the deepest known aquifer, the lower bedrock aquifer, have a stratigraphic separation
of over 6,000 feet.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-7
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
mr I
«^
r-:.Tv.'..r.i:Tr*«KjiV>
- --' < -f j^'St^V
U 30 K?T.
MW> EiSo R»J Rsd
jlcrop
7— A'tl.;hr,8 -4- Mw^slipi
-)g '""'I, z- ^t^f t" .-
.v?i .> 9« O, j^\ --
J|JJ.PC£4^
Me»/J,l:r!d I mm -3rtts 1 !SB3i, JOWSdft O3*?'-
•isw jnti Haisn J19TSI. ind Sisr* i'l-*'
f fcKruf thms!
Tectonic of the Piceance Basin (Tremain and Tyler, 1997)
*
A3-1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
1-27 -J- 5
U 1 H T A
f ^ I __ f___m
I
| —
"i
(* _=?
i* ==--?J3||5|Sp--
;•, , ' .—
J'- *~L3^
|. -J
f '' ""
? "-•"
/
-?"*
i ,
'(* '•
| -
•
63
ii
._
S
8
IU
Q
<
cc
s
| GULCH
!»
K
} ;
\ .'
:-•-- 1 „.--< • j
•*-= f
"*" ' i
-' = i
f \
!-, . |
1"
t •
t"~
'^
i s t
-1 .' i
.C-
1
DC
OJ
S
UJ
cc
—
I
rt
-^$
^ I
u. ;
CE
UJ
>
CE
i *^
!L_
; LLI
_'*.,
"3-.
~i-
f
^
•J
*.*=
1
;;
|
i i
ad
;O
•S
f
f...
'*:: .
d( :'
<f
:! —
-<"
j-.
%
>,
t
V
N
**
I- 1 'l
LL|
I
0
H
<
«
5
DC
.J
€-v?
~i
1
I
LU i' -v:
If '5
^ i $
jl_
.f
f
j
i >
~ J
• '.*
I %.
! • n •.,;,^«i_
*
i" ' EC*
.CO
j s
; |5
f- : 13
j a
f :_J
: JUJ
: !^
; ._v,,>^,v^.
t_
•T"
*•>
>,:
.«* *
i««.
. ™.
" =.„«„
^"
r :
v;
'! -- a , ,
^ 11 !
V:-_ ^
" -
'
\
^CORCORAN":
i
^
sSS
iiU.
"o
>.
DC
3
i __
Is" ^>
<* j
<n
^
(p*6^^
-eg
OC
UJ
t---"
^C/>
i i
*"|0s.
y,{
s
:<:
o
=**-, E
^ i
QC i v'°'*
01
t/>
S:
4
B™
J
_J
yj
Q
o|
cl<c
!
UJ:I
? i
>t, i
<_ t1"!
5[ j
« yy
S
tf)
1
UJ
s
-~l
:
cc
o
Stratigraphic Section of the Piceance Basin (Reineeke et a!.. 1991 )
f fl 4,*} 1 «K 10ffl ?
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3-9
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
Generalized Depth to
of Coal -
Cameo Group
»f 000**" Qspth Contoyr
Centour lns«f»»t 200O Fetn
Pieeance Basin, Colorado I
JOBJW M1OtW KliW *»KW BUSW
to of Inc. 1986)
t* Figure A3»3
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3-10
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
I or
108''
40"
•= >-&^$&. c> ;/,„.,, RfO
; fv-::2«3r'" '"""" '
h ,.-> '' "s'-T "
40';
tPLIFF
'\.'jo~r GARFIELO
_
o
;O
••9»,<fc..
109° i......
39°
MESA
I"
r,f;l>.
Nf:;
39''
DELTA
j
109"
Locations of Gas Fields (Retnecke etaL, 1991)
Figure A3-4
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3-11
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
i-i"
"
Exploration Target Areas, Piceanoe Basin (Tyler etal., 1998)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3-12
image:
EPA816-R-04-003
Attachment 3
The Piceance Basin
' '-•• ^ %»,: :.-. '.--.
' -" •
Diagrammatic East-West Sections of Hydrologie System (Taylor. 1987)
* A3-6
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3-13
image:
EPA 8 16-R-04-003
Attachment 3
The Piceance Basin
l'. vJAI r iVATTR RKO; RO"': x AU. AttF iUMFR
,a«w ,---
"""" " '"
-="—.s,4'^j|r-'."•
Dominant Chemical Constituents in the Two Major Bedrock Aquifers (Taylor, 1987)
i Fiour© A3-T
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A3 - 14
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
REFERENCES
Collins, B.A. 1976. Coal deposits of the Carbondale, Grand Hogback, and Southern
Danforth Hills coal fields, Eastern Piceance Basin, Colorado. Colorado School of
Mines Quarterly, 71(1).
Colorado Geological Survey. 2002. Public Comment OW-2001-0002-0086 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Gas Technology Institute (GTI) Web site. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Graham, G. 2001. Colorado Division of Water Resources. Personal communication.
Johnson, R. C., 1989. Geologic history and hydrocarbon potential of Late Cretaceous-
age, low-permeability reservoirs, Piceance Basin western Colorado: U. S.
Geological Survey Bulletin 1787-E, 51 p.
Johnson, R. C., and Nuccio, V. F. 1986. Structural and Thermal history of the Piceance
Creek Basin, western Colorado, in relation to hydrocarbon occurrence in the
Mesaverde Group: in Spencer, C. W., and Mast, R. F., eds., Geology of Tight
Gas Reservoirs: American Association of Petroleum Geologists Studies in
Geology No. 24, p. 165-205.
McFall, K.S., Wicks, D.E., Kruuskraa, V.A., and Sedwick, K.B. 1986. A geologic
assessment of natural gas from coal seams in the Piceance Basin, Colorado. Gas
Research Institute, Topical Report. GRI-87/0060 (September 1985-September
1986), 76 p.
Nowak, Henry C. 1991. Deposit!onal environments and stratigraphy of Mesaverde
Formation, Southeastern Piceance Basin, Colorado-Implications for coalbed
methane exploration. Guidebook for the Rocky Mountain Association of
Geologists Fall Conference and Field Trip September, 17-20, 1991, Rocky
Mountain Association of Geologists Denver, Colorado, pp. 1-20.
Quarterly Review. 1993. Coalbed Methane - State of the Industry. Methane from Coal
Seams Technology, August.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3-15
image:
EPA 816-R-04-003 Attachment 3
The Piceance Basin
Reinecke, Kurt M., Rice, D. D. and Johnson, R. C. 1991. Characteristics and
development of fluvial sandstone and coalbed reservoirs of Upper Cretaceous
Mesaverde Group, Grand Valley field, Colorado. Guidebook for the Rocky
Mountain Association of Geologists Fall Conference and Field Trip September,
17-20, 1991, Rocky Mountain Association of Geologists Denver, Colorado, pp.
209-225.
Taylor, O. J. 1987. Oil shale, water resources, and valuable minerals of the Piceance
basin, Colorado: The challenges and choices of development. U. S. Geological
Survey Professional Paper 1310, 143 p.
Tessin, Robert. 2001. Colorado Oil and Gas Conservation Commission. Personal
communication.
Tremain, Carol M. and Tyler, R. 1997. Cleat, fracture, and stress patterns in the
Piceance Basin, Colorado: Controls on coalbed methane producibility. Rocky
Mountain Association of Geologists, Fractured Reservoirs: Characterizations and
Modeling Guidebook.
Tyler, R, Scott, A. R., Kaiser, W.R., Nance, H. S., McMurry, R. G., Tremain, C. M., and
Mavor, M. J. 1996. Geologic and hydrologic controls critical to coalbed methane
producibility and resource assessment: Williams Fork Formation, Piceance Basin,
Northwest Colorado. The University of Texas at Austin, Bureau of Economic
Geology, topical report prepared for the Gas Research Institute, GRI-95/0532,
398 p.
Tyler, R., Scott, A. R. and Kaiser, W. R. 1998. Defining coalbed methane exploration
fairways: An example from the Piceance Basin, Rocky Mountain Foreland.
Western United States; in Mastalerz, M., and Glikson, M. eds., Coalbed Methane,
Scientific, Environmental and Economic Evaluation: International Conference on
Coal Seam Gas and Oil, Brisbane, Queensland, Australia, pp. 67-87.
USGS-National Water Summary. 1984. Hydrologic events, selected water-quality
trends, and ground-water resources. United States Geological Survey Water-
Supply Paper No. 2275.
Western Oil World. December 1991. Colorado-Utah coalbed methane projects multiply.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A3 -16
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
Attachment 4
The Uinta Basin
The Uinta Coal Basin is located mostly within eastern Utah; a very small portion of the
basin is in northwestern Colorado (Figure A4-1). The basin covers approximately 14,450
square miles (Quarterly Review, 1993) and is structurally separated from the Piceance
Basin by the Douglas Creek Arch (Figure A4-1), an up-warp near the Utah - Colorado
state line. Coalbeds are present within Cretaceous strata throughout much of the Uinta
Basin. However, coalbed methane exploration, to date, has targeted coalbeds in the
Ferron Sandstone Member of the Mancos Shale and coalbeds in the Blackhawk
Formation of the Mesaverde Group. The total, in-place, coalbed gas resources in the
Wasatch Plateau, Emory, Book Cliffs and Sago coal fields have been estimated at 8
trillion cubic feet (Tcf) to more than 10 Tcf by the Utah Geological Survey (Gloyn and
Sommer, 1993). This estimate is based on extrapolation of known coal resources to a
depth of 9,000 feet and an average projected gas content of 330 cubic feet per ton and
does not include the Tabby Mountain or Vernal coalfields, or the Sevier-Sanpete coal
region. Total production stood at 75.7 billion cubic feet (Bcf) of coalbed methane in
2000 (GTI, 2002).
4.1 Basin Geology
Much of the Rocky Mountain region, including the Uinta Basin was covered by an
epicontinental sea. Deposition in the sea lasted from the Albian (about 100 million years
ago) through the Cenonmanian (about 83 million years ago), with the deposition of the
upper part of the Mesaverde Group generally marking the end of marine deposition in the
basin (Howells et al., 1987).
The Uinta Basin formed as a result of uplift and deformation that began in the Late
Cretaceous. The Cretaceous sediments outcrop along the perimeter of the basin. The
basin is asymmetrical in shape with strata on the northern flank of the basin dipping
steeply toward the basin axis, while strata on the southern flank dip gently toward the
basin axis. The stratigraphic units of the coal bearing Cretaceous rocks of the Uinta
Basin are shown in Figure A4-2.
Two Cretaceous stratigraphic units have been targeted for coalbed methane exploration:
the Ferron Sandstone Member of the Mancos Shale and the Blackhawk Formation of the
Mesaverde Group (Figure A4-2). The Ferron Sandstone Member was deposited in the
Last Chance delta, a fluvial-deltaic environment (Garrison et al., 1997). The coalbeds
and interbedded sandstone units form a wedge of clastic sediment 150 to 750 feet thick
stratigraphically above the Tunuck Shale Member of the Mancos Shale and below the
Lower Blue Gate Shale Member of the Mancos Shale (Figure A4-2). Both of these shale
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-1
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
units have a very low permeability and constitute confining units for water and gas in the
Perron Sandstone Member. The coal-bearing rocks are thickest to the west and south
margins of the basin, nearer to the upland sources of sediment. Coalmines producing
from the Perron Sandstone Member are located along the eastern boundary of the
Wasatch Plateau south of Castle Dale, Utah (Figure A4-1). Depths to coal in the Perron
Sandstone Member range from 1,000 to over 7,000 feet (Garrison et al., 1997). Primary
coalbed methane activity from the Perron Sandstone takes place in the Drunkard's Wash
Unit. Total coal thickness in this area ranges from 4 to 48 feet (averaging 24 feet) from
depths of 1,200 to 3,400 feet (Lamarre and Burns, 1996).
The Blackhawk Formation consists of coal interbedded with sandstone and a combination
of shale and siltstone. The Blackhawk Formation is underlain by the Star Point
Sandstone and overlain by the Castlegate Sandstone (Figure A4-2). The Castlegate
Project in the Book Cliffs coalfield initially targeted coals in the Blackhawk Formation at
depths ranging from 4,200 to 4,400 feet (Gloyn and Sommer, 1993).
4.2 Basin Hydrology and USDW Identification
Groundwater hydrology of the Uinta Basin is controlled primarily by the geologic
structure of the region (Howells et al., 1987). Variations of aquifer and aquitard
permeability owing to differences of lithology and facies changes also play an important
role in the hydrology, as does widespread faulting and fracturing of the rocks (Howells et
al., 1987). Because of the basin's structure, the area may be a groundwater basin with
internal drainage. If there were a deep groundwater outlet for the basin, it would be along
or near the axis of the Uinta Basin at its western edge. The general pattern of
groundwater flow is centripetal, with water flowing inward from recharge areas at
exposures of permeable strata at the margins of the basin. Recharge is greatest near the
northern edge of the basin. Other recharge areas include Eocene and Oligocene
Formations in the basin interior.
Most of the sandstone formations in the Mesozoic rocks in the Upper Colorado River
Basin are identified as aquifers by the United States Geological Survey (Freethey and
Cordy, 1991). Freethey and Cordy stated that in the Uinta Basin, the older and deeper
aquifers in strata below the Ferron Sandstone Member, (for example, the Navajo-Nugget
Aquifer, Entrada-Preuss Aquifer, Morrison Aquifer, and the Dakota Aquifer) generally
contain very saline to briny water, with total dissolved solids (TDS) values greater than
10,000 milligrams per liter (mg/L). The water quality component of the underground
source of drinking water (USDW) definition specifies that a USDW contain less than
10,000 mg/L of TDS. The Ferron Sandstone Member (Figure A4-3) is designated as a
producing aquifer in east-central Utah (Freethey and Cordy, 1991). In regard to the
Mesaverde Group Aquifer, which includes the Star Point Sandstone, the Blackhawk
Formation, the Castlegate Sandstone and the Price River Formation, (Figure A4-3)
Freethey and Cordy (1991), stated that, "water in these aquifers is more likely to be
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-2
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
developed where the saturated thickness is large and the depth to the aquifer is less than
2,000 ft." They further stated that the margins of the Uinta Basin where these rocks are
near the surface or outcrop is a possible location for development of groundwater with
low enough IDS to be used for drinking water.
Wells in the Perron Sandstone Member at the Drunkard's Wash coalbed methane field
typically penetrate to depths ranging from 1,200 to 3,400 feet (Lamarre and Burns, 1996).
An average water quality value of 13,120 mg/L IDS (Gwynn, 1998) for production
waters that have been retained in catchment ponds suggests that these wells are not within
a USDW. Gwynn (1998) however, does state that due to the ponding of the produced
water in evaporation lagoons, the concentration of salts in these waters has probably
increased from their original levels. This implies that these water quality data may not be
useful in the confirmation of USDW qualifications. Quarterly Review (1993) reported
that three wells producing gas and water from the Perron Sandstone Member coalbeds in
the Drunkard's Wash field yielded over 49,000 gallons of water per day with a TDS level
of about 5,000 mg/L (sodium bicarbonate) during the first 2 to 3 months of operation.
The Perron Sandstone is hydrologically confined above and below by shale members of
the Mancos Shale formation. Water produced from the Perron Sandstone is thought to be
connate water that was trapped in the sediment during coalification (Gloyn and Sommer,
1993). Hunt (Utah Division of Oil, Gas, and Mining, 2001) noted that there were no
USDWs located immediately above the Perron Sandstone Member due to the thick
tongues of Mancos Shale that encapsulate the coal-bearing interval (Figure A4-2).
Beds targeted for methane gas exploration and production within the Blackhawk
Formation are approximately 4,200 to 4,400 feet below the ground surface (Gloyn and
Sommer, 1993). Coalbed gas production in the Castlegate Field accounted for less than
10 percent of the coalbed methane production in the Uinta Basin (Petzet, 1996). The
average gas well producing from the coalbeds in the Blackhawk Formation (Castlegate
field) yielded 318 barrels of water per day, and TDS levels of 5,489 mg/L have been
measured in the produced waters (Gloyn and Sommer, 1993).
According to the State of Utah Department of Natural Resources (DNR), Division of Oil,
Gas and Mining, the water quality in the Perron and Blackhawk varies greatly with
location, each having some TDS levels below and some above 10,000 mg/L (Utah DNR,
2002). In general, the quality of Blackhawk water is higher than that of Perron water.
The most recent Underground Injection Control application received for the Drunkard's
Wash field (Perron) showed a composite quality of input water to be about 31,000 mg/L
TDS, and for the Castlegate field (Blackhawk) 9,286 mg/L TDS. At some locations,
either formation member would not qualify as a USDW.
In the western part of the Uinta Basin, the Castlegate Sandstone, an aquifer, is separated
from the Black Hawk Formation coalbeds by approximately 300 feet of alternating shale
and sandstone (Utah DNR 2002). The Star Point Sandstone is located below
approximately 400 feet of alternating sandstone and shale that underlies the bottom coal
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-3
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
of the Black Hawk Formation. In some areas, the shale and sandstone underlying the
Black Hawk coals are highly faulted. There is some potential that hydraulic fracturing
fluids could be transported through natural fracture networks in these areas and reach the
Star Point Sandstone. The relatively impermeable upper Blue Gate Shale Member of the
Mancos Shale Further would prevent further downward migration.
In reference to the quality of water produced by the coalbed gas wells in both the Ferron
Sandstone Member of the Mancos Shale and the Blackhawk Formation, Quarterly
Review (1993) states: "Disposal of produced water does not appear to present a major
environmental problem in the Uinta basin, unlike the San Juan and some other western
basins. Rates are moderate, 200 to 300 barrels per day per well during early stages of
production and TDS levels are not high (about 5,000 mg/L)." Because these TDS values
are less than the 10,000 mg/L limit, both the Ferron Sandstone Member of the Mancos
Shale and the Blackhawk Formation may qualify as USDWs.
Tabet (2001) suggests that coalbed methane extraction wells are not located in
"producing" aquifers and that most of the potable water in the sparsely populated area is
supplied by surface water and shallow alluvial aquifers.
4.3 Coalbed Methane Production Activity
Full-scale exploration in the Uinta Basin began in the 1990s (Quarterly Review, 1993).
The most active operators at that time were PG&E Resources Company, the River Gas
Corporation, Cockrell Oil Corporation, and Anadarko Petroleum Corporation. PG&E
acquired the Castlegate Field, from Cockrell Oil (Gloyn and Sommer, 1993). Gas was
produced from coalbeds in the Blackhawk Formation. The five wells initially drilled in
the Castlegate Field were hydraulically fractured with 80,000 to 143,000 pounds of sand
and unreported volumes of fluid. Other wells were to be fractured with a low-residue gel
system to ensure breakdown within the reservoir (Quarterly Review, 1993).
The Castlegate field was off-line due to production water disposal problems (Tabet,
2001; and Hunt, Utah Division of Oil, Gas, and Mining, 2001). According, to the State
of Utah DNR, Division of Oil, Gas and Mining, the field is now on production (Utah
DNR, 2002).
The River Gas Corporation operates the Drunkard's Wash Unit, producing methane gas
from coals within the Ferron Sandstone Member. The company reported that high
fracture gradients hampered hydraulic fracturing stimulations using cross-linked borate
gel with 250,000 pounds of proppant (Quarterly Review, 1993). Excessive proppant
flowback resulted in one well where nitrogen foam was used for the fracturing. The
Buzzard Bench Field, also producing gas from the Ferron Sandstone Member, was
initially operated by Chandler & Associates, Inc. (Petzet, 1996) and is currently being
managed by Texaco (Garrison et al., 1997).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-4
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
A query of a database covering the Uinta Basin revealed that there are about 1,255
coalbed methane wells in production in the basin (Osborne, 2002). Gas Technology
Institute (GTI) places the annual coalbed methane production in the Uinta Basin at 75.7
Bcf in 2000 (GTI, 2002).
4.4 Summary
Waters from coalbed methane production in the Perron Sandstone Member of the Mancos
Shale in the Drunkard's Wash Unit are conflictingly reported to have IDS values of
about 13,000 mg/L according to one source of information or to have levels of TDS of
about 5,000 mg/L from another. However, the higher values were derived from water
samples taken from evaporation lagoons and these high values might represent elevated
concentrations of salts owing to evaporation. Consequently, if the more moderate TDS
levels were correct, then the Perron Sandstone would qualify as a USDW.
According to the State of Utah DNR, Division of Oil, Gas and Mining, the water quality
in the Perron and Blackhawk varies greatly with location, each having TDS levels below
and above 10,000 mg/L (Utah DNR, 2002). In general, the quality of Blackhawk water is
fresher than Perron water. The most recent Underground Injection Control application
received for the Drunkard's Wash field (Perron) showed a composite quality of input
water to be about 31,000 mg/L TDS, and for the Castlegate field (Blackhawk) 9,286
mg/L TDS. At some locations, neither formation member would qualify as a USDW.
The Drunkard's Wash and Castlegate coalbed methane extraction fields are located in a
sparsely populated section of Utah. Tabet (Utah Geological Survey, 2001) suggests that
coalbed gas extraction wells are not located in "producing" aquifers and that most of the
potable water in the sparsely populated area is supplied by surface water and shallow
alluvial aquifers.
The Blackhawk Formation is underlain by 300 feet of shale and sandstone that separate it
from the Castlegate Sandstone aquifer. It is underlain by similar geologic strata, which
separate it for the Star Point Sandstone. Only in highly faulted areas is there a reasonable
possibility that hydraulic fracturing fluids could migrate down to the Star Point
Sandstone.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-5
image:
EPA816-R-04-003
Attachment 4
The Uinta Basin
TAB8V Ducheana
MOUNTAIN
COAL FICLO
EMERY CO
SAN
RAFAEL
SWELL
SEVIERCO, \SP t~ U
Index Map of Coal Fields in Uinta Basin, Utah (Quarterly Review, 1993)
Figure A4-1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A4-6
image:
EPA816-R-04-003
Attachment 4
The Uinta Basin
Age Nomenclature Thickness
2-
1
Cretaceous
I
|
i
Campanian
Santonian
c
8
C
3
i
a
North Horn Formation
CL Price River
•2 Formatton
^ Castlegate
.§ Formatton
^ Blackhawk
«5 Formation
5
Star Point
Sandstone
Mancos Shale
Upper Blue Gate
Shale Member
Emery Sandstone
Member
Lower Blue Gate
Shale Member
Member
Tunuck Shale
Member
Dakota Sandstone
Cedar
Mountain
Formation
Morrison
Formation
Adaj
I1!'!'!'!'!! Carbonate
Upper Member
Brushy Basin
Member
Salt Wash
Member
>ted (rom Hintze (1
II Si
500-2500
600-1000
150-500
700-1500
100-1000
700-1000
800-50
1600-2400
400-650
0-60 "^
100-200
0-50 ^
570
510
Littiology
i«Mr3
m^si^
;;;;;;;;;;)
|
:::::::::::]
^P^liliiin^ t t i i SonnaCM^^m,
' '„*,;•* S&^ml^ng^..^
-^ ^
»Jtnrr^-^~-
••• in J -^ uppwG*n»y Canyon
- - , i ' L&mtn G*rte* Canvon
VV-VV-j Memb«(NJ
"v;-;-^i _ ^
p^-' • * 1
988}
ale and Siltstone B§S2S9 Conglomerate
ind stone
Stratigraphic Column for the Cretaceous of the Castle \
/alley (Garrison etal., 1997)
71 1 4/01 kk 1 027-«F-Uima-2 o)r Fi Q U r6 A4-2
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A4-7
image:
Q2 C^ ^
a- 3. g"
S^ 8,
£j g. g"
ft ft >73
2 £• C
5' 3 s
E. &
R'oq
3 §
0 C3
c? ^*
3. in
OQ g
o o
VI
p
ft
^ 0
i O
OO .£*
$
^
i A
r{
\
1* *
5 (
"*. 1 *
, ' ,J
y ?/
"/•-,
:• '" 1
Clf
:f *
. 1
,™
i
i'"^
Nf
1
/ s"~-i
/'/A •
1 pjX
i re \
\\t
\
i i
y
f
ii
i
i •
»
i i
if?
rjl
•*
r:
T
1,* • " '
_
9Kf*f fc '
f" * iiT
'• '""/'
. f
^ -..
i * "
— —
1 1
i
\
4
r>7
R 1 |C I
i'j^i i
IT?
JJ;
1 . ^
f 1
i "?•> "'
1 ! «-
11 .-
',
-
"**" '^
^
Coaibed
_.',
'l-*. ;•"
"r i
' -IX' -II j
Ml1'1'
f^^V i'
i ^J*V<; j]
^ " ^"V_^^ • \
i".- . - '_;
* * ' ".">
1
I ,
i .'• • •
i i *
. a
' 1
I
i
i=r .-
- ,•
i^i
/ •_-_-:-•.
=5BF
^" ' ^_ — ^ ^' "
eo.
UT*«CCJDLL-=ENCCO i 1
• 1 ^^^^^^^^^^^H ' '"" n» "%i-
' ^^^^^^^H ' O f> i J ' Ar^dar^ .'^^ i
I _H^M|H«W^M^ PiHfAium -'. ^^
f. i; •"" "B- -j [ / i
S Of li, ^ C 1 ^ 3 F S j : (4 j j > » j
, V -— — tt» . 7 / j j^. ^__ 1 . •• »* '
--" \ r,>'\;' "" 7 « i "" -- ». r?'** ^
'"""^>-' * ,-.'•-'.',',, ._ ---!>x _. . .x*- v «*
-, « F i * i"-.
.. S-'*'yi _ " -"
fr* i ! ^ i -v,-^-t
/ l '- f f '
* • ' f .; .'
" "" ' ' " 3- ' />
r_. i-'uA^U! -^ )»X-"..-,MA-|ON-.\P'JC-
R-ie1 Oai Coup . =ft© ' [ ] -• cm-.-, .... ; t;
^ r «fcl^~ i i *"
j ' •'"v ' .^"!
, Hvi-Mardj .yjsfi L>T? ^ * "J1-"" / ~s'' '>"
i , -1 •' • / ,//
1 r' !. .. vv. ., ' ,/ /f/f ?
' r - --• ' • L'^ / // ; • 4- T1«s
1 / * .'
/ rl ' ^
/ • r 1 i p
'd C4P.3CN :•:• / ' . • :"» ,-
R-JF R10E ^iiE H1ZE '
rtj i.Q»*cknC'4«fc:FjunZ!ird j (J - • s it y, las
Methane Oevelap«ne«t in Carbon County. Ulah. (Mettvane Coal Seams Technology, 19S3)
Areas shewn may not represeKnl aclual fisld. unit or tease boundaries.
>--{•' ~*, ^iru1^ ^i^-"*i
^
OO
K^
ON
0
^
O
sl
i. "
MB
?D
p; M
££. r4~
3 •**•
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
REFERENCES
Freethey and Cordy, 1991. Geohydrology of Mesozoic Rocks in the Upper Colorado
River Basin in Arizona, Colorado, New Mexico, Utah, and Wyoming. United
States Geological Survey Professional Paper 1411-C
Garrison, James R., Jr. (non-survey author), van den Bergh, T.C.V. (non-survey author),
Barker, Charles E., Tabet, David E. (non-survey author). 1997. Depositional
sequence stratigraphy and architecture of the Cretaceous Perron Sandstone;
implications for coal and coalbed methane resources; a field excursion. Link,
Paul Karl (non-survey editor), Kowallis, Bart J. (non-survey editor), Mesozoic to
Recent geology of Utah, Geology Studies, 42(2): 155-202.
Gloyn, Robert W. and Sommer, Steven N. 1993. Exploration for coalbed methane gains
momentum in Uinta Basin. Utah Geological Survey, Oil & Gas Journal,
Exploration, pp. 73-76, May 31, 1993.
Gas Technology Institute (GTI) website. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Gwynn, J. Wallace. November 1998. Potential mineral precipitation and water
compatibilities related to the Drunkards Wash Project, Carbon County, Utah.
Report of Investigation 241, Utah Geological Survey.
Howells, Lewis, Longson, Mark S., and Hunt, Gilbert L. 1987. Base of moderately
saline groundwater in the Uinta Basin, Utah, with an introductory section
describing the methods used in determining its position. State of Utah,
Department of Natural Resources, Technical Publication No. 92, U.S. Geological
Survey Open-File Report 87-394, 1987.
Hunt, Gill. January 2001. Utah Division of Oil, Gas and Mining, Salt Water Disposal
Unit. Personal communication.
Lamarre, Robert A. and Burns, Terry D. 1996. Drunkard's Wash Unit: Coalbed
methane production from ferron coals in East-Central, Utah. GSA Abstracts with
Programs, p. A-58.
Osborne, Paul. 2002. USEPA Region VIIIUIC Program. Personal communication.
Petzet, G. Alan. 1996. Utah coalbed gas exploration poised for growth. Oil & Gas
Journal, Exploration, p. 54, August 5, 1996.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-9
image:
EPA 816-R-04-003 Attachment 4
The Uinta Basin
Quarterly Review. 1993. Coalbed methane - state of the industry. Quarterly Review,
1993.
Tabet, D. January 2001. Utah Geological Survey, Personal communication.
Utah Department of Natural Resources. 2002. Public Comment OW-2001-0002-0090 to
"Draft Evaluation of Impacts to Underground Sources of Drinking Water by
Hydraulic Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol.
63, No. 185. p. 33992, September 24, 2002.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A4-10
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
Attachment 5
The Powder River Basin
The Powder River Basin is located in northeastern Wyoming and southern Montana. The
basin covers an area of approximately 25,800 square miles (Larsen, 1989), approximately
75 percent of which is within Wyoming (Figure A5-1). Fifty percent of the basin (Figure
A5-2) is believed to have the potential for production of coalbed methane (Powder River
Coalbed Methane Information Council, 2000). Much of the coalbed methane-related
activity has been north and south of Gillette in northeastern Wyoming (Figure A5-2).
The majority of the potentially productive coal zones range from about 450 feet to over
6,500 feet below ground surface (Montgomery, 1999). In addition to being an important
resource for coalbed methane, the basin has also produced coal, petroleum, conventional
natural gas, and uranium oxide (Law et al., 1991; Randall, 1991). Recent estimates of
coalbed methane reserves in the Powder River Basin have been as much as 40 trillion
cubic feet (Tcf) (PRCMIC, 2000) but more conservative estimates range from 7 to 12 Tcf
(Montgomery, 1999). Annual production volume was estimated at 147 billion cubic feet
(Bcf) in 2000 (GTI, 2002). In 2002, wells in the Powder River Basin produced about 823
million cubic feet (Mcf) per day of coalbed methane (DOE, 2002).
The information available indicates that hydraulic fracturing currently is not widely used
in this region due to concerns about the potential for increased groundwater flow into the
coalbed methane production wells and collapse of open hole wells in coal upon
dewatering. According to the available literature, where hydraulic fracturing has been
used in this basin, it has not been an effective method for extracting methane.
5.1 Basin Geology
The Powder River Basin is a thick sequence of sedimentary rock formed in a large
downwarp within the Precambrian basement. The basin is bounded on the east by the
Black Hills uplift, on the west by the Big Horn uplift and Casper Arch, on the south by
the Laramie and Hartville uplifts and, on the north, it is separated from the Williston
Basin by the Miles City Arch and the Cedar Creek Anticline (Larsen, 1989) (Figure A5-
1). The long axis of the basin is aligned in a generally southeast to northwest direction,
and it is as much as 18,000 feet deep (Randall, 1991) (Figures A5-1 and A5-3).
Sediments range from Paleozoic at the bottom through Mesozoic to Tertiary at the top
(DeBruin et al., 2000). The basin is a large asymmetrical syncline with its axis (deepest
part) near the west side of the basin (Figure A5-3). From outcrops along the eastern edge
of the basin, the sediments slope gently (1.5°, about 100 feet per mile) downward to the
southwest and then bend steeply upward (10 to 45°) to outcrop in a monocline along the
western edge of the basin.
Several periods of deposition by marine and fluvial-deltaic processes have occurred
within the basin during the Cretaceous and Tertiary periods. These Cretaceous and lower
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-1
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
Tertiary rocks have a total thickness of up to 15,000 feet (Montgomery, 1999). Coal is
found in the Paleocene Fort Union and Eocene Wasatch Formations (Figure A5-4). The
Wasatch Formation occurs at land surface in the central part of the basin and is covered
by alluvium or White River Formation in some places (Figure A5-4). Most of the
coalbeds in the Wasatch Formation are continuous and thin (six feet or less) although,
locally, thicker deposits have been found (DeBruin et al., 2000). The Fort Union
Formation lies directly below the Wasatch Formation and can be as much as 6,200 feet
thick (Law et al. 1991). The Fort Union Formation outcrops at the ground surface on the
eastern side of the basin, east of the City of Gillette and on the western side of the basin,
north and south of Buffalo. The coalbeds in this formation are typically most abundant in
the upper Tongue River Member (Figure A5-4). This member is typically 1,500 to 1,800
feet thick, of which up to a composite total of 350 feet of coal can be found in various
beds. The thickest of the individual coalbeds is over 200 feet (Flores and Bader, 1999).
The coalbeds are interspersed with sandstone, conglomerate, siltstone, mudstone and
limestone (Montgomery, 1999).
Most coalbed methane wells in the Powder River Basin are in the Tongue River Member
of the Fort Union Formation, in the Wyodak-Anderson coal zone, which contains up to
32 different coalbeds according to some authors (Ayers, 1986), including the Big George
in the central part of the basin (Flores and Bader, 1999). The Wyodak is one of the thick
coalbeds that are targeted for coalbed methane development. This coalbed is also called
the Wyodak-Anderson or the Anderson, and it can be subdivided further into several
other coalbeds. These coalbeds are the Canyon, Monarch, and Cook. All of these
coalbeds are coalbed methane targets. Most coalbeds are found within 2,500 feet of the
ground surface.
The Wyodak or Wyodak-Anderson coalbed in the Wyodak-Anderson coal zone is
prominent in the eastern portion of the Powder River Basin near the City of Gillette
(Figures A5-3, A5-5 and A5-6). The Wyodak has been identified as the largest single
coalbed in the country (Montgomery, 1999). The coal is close to the ground surface and
mining of the coal is common. The Wyodak coalbed gets progressively deeper and
thicker toward the west. This bed ranges from 42 to 184 feet thick. Most of the coalbed
methane wells in the Powder River Basin are within the Wyodak coal zone near the City
of Gillette.
The Big George Coalbed is located in the central and western portion of the Powder
River Basin (Figure A5-7). Although the Big George is stratigraphically higher than the
Wyodak, owing to the structure of the basin, the Big George, in the center portion of the
basin, is deeper than the Wyodak at the eastern margin of the basin (Tyler, et al., 1995).
To date, the Big George has not been developed for coalbed methane production to the
same extent as the Wyodak-Anderson coal zone. This is due to a combination of factors
including greater depth to coal, more groundwater, and longer distances to available
transmission pipelines. However, as of December 2001, there were about 850 coalbed
methane wells drilled into the Big George with a large number of wells planned for the
future (Osborne, 2002).
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-2
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
A third significant coal zone, the Lake De Smet coal zone in the Wasatch Formation, is
up to 200 feet thick and is located in the Lake De Smet area (Figure A5-8), 55 miles
southwest of Recluse on the western side of the basin (Larsen, 1989). It has not yet been
widely used for coalbed methane production.
Most of the coal in the Powder River Basin is subbituminous in rank, which is indicative
of a low level of maturity. Some lignite, lower in rank, has also been identified. The
thermal content of the coals found in the Powder River Basin is typically 8,300 British
thermal units per pound (Randall, 1991). Coal in the Powder River Basin was formed at
relatively shallow depths and relatively low temperatures. Most of the methane
generated under these conditions is biogenic, which means that it was formed by bacterial
decomposition of organic matter. Thermogenic formation (formed under high
temperature) was not significant in most locations within the Powder River Basin.
Consequently, coal in the Powder River Basin contains less methane per unit volume than
many other coal deposits in other parts of the country. Coal in the Powder River Basin
has been found to contain 30 to 40 standard cubic feet of methane per ton of coal
compared to 350 standard cubic feet of methane per ton in other areas (DeBruin et al.,
2000). The gas is typically more than 95 percent methane, the remainder being mostly
nitrogen and carbon dioxide. This resource was overlooked for many years because it
was thought to be too shallow for the production of significant amounts of methane
(Petzet, 1997). However, the relatively low gas content of Powder River Basin coal is
compensated by the thickness of the coal deposits. Because of the thickness of the
deposits and their accessibility, commercial development of the coalbed methane has
been found to be economical.
The Powder River Basin contains approximately 60 percent of the coalbed methane
reserves in the State of Wyoming (DeBruin et al., 2000). Recent estimates of coalbed
methane reserves in the Powder River Basin have been as much as 40 Tcf (PRCMIC,
2000) but more conservative estimates range from 7 to 12 Tcf (Montgomery, 1999). As
of December 1999, monthly production exceeded 7 Bcf from 1,657 wells (DeBruin et al.,
2000). Wells typically produce 160,000 cubic feet of gas per day (DeBruin et al., 2000).
Annual production volume was estimated at 147 Bcf in 2000 (GTI, 2002). In 2002, wells
in the Powder River Basin produced about 823 Mcf per day of coalbed methane (DOE,
2002). Coalbed methane has been developed along both the east and west flanks of the
basin where the coalbeds are buried but relatively shallow. Many existing wells are
awaiting connection to the distribution system and still more wells are being drilled. The
estimated lifetime production from these wells is 300 to 400 Mcf per well (Petzet, 1997).
The amount of coalbed methane produced from each well is highly variable, and the
volume of gas depends on the quality and thickness of the coal, the frequency of natural
cleats in the coal, and the amount of water present. Other factors, such as well
completion techniques and well stimulation techniques, also control the amount of gas
produced from a well. Maximum coalbed methane flow from a well is typically achieved
after one to six months of dewatering (Montgomery, 1999). Stable production is usually
experienced for one to two years before production begins to decline (Montgomery,
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-3
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
1999). Production often declines at a rate of 20 percent per year until the well is no
longer economically useful (Montgomery, 1999). Several options exist at that point,
including re-fracturing the well, completing the well in a deeper coal formation,
converting the well to a water supply well, or abandoning the well.
5.2 Basin Hydrology and USDW Identification
A report prepared by the United States Geological Survey (USGS) showed that samples
of water co-produced from 47 coalbed methane wells in the Powder River Basin all had
total dissolved solids (TDS) levels of less than 10,000 milligrams per liter (mg/L) (Rice
et al., 2000). Based on the water quality component of the underground source of
drinking water (USDW) definition, which specifies that a USDW contain less than
10,000 mg/L of TDS, the Fort Union Formation coalbeds are within a USDW. The water
produced by coalbed methane wells in the Powder River Coal Field commonly meets
drinking water standards, and production waters such as these have been proposed as a
separate or supplemental source for municipal drinking water in some areas (DeBruin et
al., 2000). Sandstones in the sediments both above and below the coalbeds are also
aquifers.
In 1990, Wyoming withdrew an average of 384 million gallons per day of groundwater
for a variety of purposes, the majority of which was agriculture. Approximately 13
percent was used for potable water supplies. Approximately 22 percent was withdrawn
by industry and mining (Brooks, 2001). The proportion of this 22 percent attributable to
coalbed methane production is increasing rapidly, and a concern exists that such good
quality water in a semiarid region should be conserved (Quarterly Review, 1993). In
1990, before the rapid expansion of coalbed methane extraction in the region, Campbell
County was identified by the USGS as an area of major groundwater withdrawal.
Approximately 80 percent of Wyoming residents rely on groundwater as their drinking
water source (Powder River Basin Resource Council, 2001). Few public water supply
systems exist in the Powder River Basin due to relatively low population densities. The
City of Gillette, the largest in the major coalbed methane development area (Figure A5-
2), uses groundwater from two sources identified as "in-town wells", and the "Madison
Well Field". The city has experienced considerable drawdown and reduced production
from their in-town wells that are completed in the Fort Union and Lance/Fox Hills
aquifers (Brooks, 2001). It is unclear how much of the drawdown is attributable to
withdrawals for water supply as a consequence of population growth and how much is
attributable to nearby coalbed methane production. Between 1995 and 1998, the city
restored and/or replaced several of its wells. The Madison Well Field produces water
from the Madison Formation and is approximately 60 miles east of the city. There are no
coalbed methane wells in the vicinity of the Madison Well Field (Brooks, 2001).
Regional groundwater flow in the basin is reported to be toward the northwest (Martin et
al., 1988 in Law, 1991), with recharge occurring in the east along the Rochelle Hills.
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-4
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
Cleats and other fractures within the coalbeds create high hydraulic conductivities and
facilitate the flow of groundwater and high water production within the coalbeds
(Montgomery, 1999). The coalbeds are largely hydraulically confined by underlying
shale and by basinward pinch-out. Surficial water and rainwater can enter the Fort Union
coals from land surface at the eastern edge of the basin and at the Black Hills uplift. This
flow inward from outcrop areas at higher elevations on the edge of the basin may have
created artesian conditions in the deeper central portions of the basin. However, this
view may not be entirely correct. For example, coalbed research (Law et al., 1991)
hypothesizes that the sodium bicarbonate water in the Fort Union coal near the central
part of the basin may not be derived from meteoric recharge, but rather from interstitial
waters of the original peat deposits. Furthermore, Martin et al. (1988, as cited in Law et
al., 1991) concluded on the basis of isotopic composition of water samples that only part
of the water near outcrops was of meteoric origin. Although artesian pressure in the
center of the basin has been thought to be evidence that the center of the basin is fed from
meteoric recharge at the basin margins, the apparent artesian pressure (flowing wells)
could be explained by the airlift effect of methane coming out of solution within the
rising well water column.
Because the coalbeds are productive aquifers, they also require more dewatering of
coalbed methane wells for methane production. Groundwater production, in terms of
volume of water produced, was a major factor considered in the selection of sites for
early coalbed methane wells and may still guide development of sites in some parts of the
Powder River Basin. Wells in the eastern portion of the basin have been found to contain
less water due to their location above the water table within the eastern anticlinal updip of
the formation and, in some areas, due to the presence of nearby mines that dewater the
aquifer. Drawdowns of up to 80 feet have been measured in wells near active mines;
however, water levels have been reported to be unaffected at distances of more than three
miles from mines (Randall, 1991). The Bureau of Land Management in conjunction with
the State Engineer's Office has been conducting ongoing research on the effects of
coalbed methane production on drawdown (Wyoming Geological Association, 1999).
5.3 Coalbed Methane Production Activity
Coalbed methane activity in Wyoming occurs predominantly in Campbell, Sheridan and
Johnson Counties (DeBruin, 2001). Wells are spaced from 40 to 80 acres per well, as
determined by the State. Permits are required under both state water well regulations and
state gas well regulations before drilling can commence. A discharge permit from the
Wyoming Department of Environmental Quality is also required for the water that is
removed from the well. Coalbed methane production wells in the Powder River Basin
are typically 400 to 1,500 feet deep and can be as shallow as 150 feet (PRCMIC, 2000).
By comparison, conventional gas and oil wells installed in the area are typically
4,000 to 12,000 feet deep (PRCMIC, 2000). Plans for construction of approximately
4,000 new coalbed methane production wells in the Montana portion of the Powder River
Basin await completion of an in-depth environmental study (DeBruin, 2001).
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-5
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
Commercial development of methane directly from the coal seams began approximately
in 1986. There were only 18 wells producing coalbed methane in the Powder River
Basin by 1989. The number grew slowly through the early 1990s with 171 wells
producing approximately 8 Bcf of gas per year. The rate of development of the resource
accelerated greatly from 1997 to 1999. In 1999, there were 1,657 coalbed methane wells
operating in the Powder River Basin, producing approximately 58 Bcf per year (Figure
A5-9) of coalbed methane. As of November 2000, there were about 4,270 wells in
Wyoming producing 15 Bcf of coalbed methane in that month alone (Osborne, 2002).
By November 2001, monthly coalbed methane production had climbed to 23.5 Bcf from
7,870 producing wells in Wyoming (Osborne, 2002). In Montana, 246 active wells
produced 872,008 Mcf of coalbed methane in December, 2001 (Osborne, 2002). The
Powder River Basin has become the most active coalbed methane exploration and
production area in the country (DeBruin et al., 2000). Despite all of the activity, less than
5 percent of the land underlain by coal in the Powder River Basin had been explored for
the presence of coalbed methane as of the year 2000 (PRCMIC, 2000).
During the early years of coalbed methane development in the Powder River Basin
(1980s to early 1990s), gas exploration and development companies completed wells
with and without hydraulic fracture techniques. Larsen (1989) indicated that early wells
were completed without fracturing treatments, particularly wells targeting gas reserves in
coals interspersed between sandstone layers. However, the Quarterly Review (1989)
reported that in one well, Rawhide 15-17, located north of Gillette, Wyoming, an "open
frac" hydraulic fracturing was performed using 13,000 Ibs of 12/20-mesh sand in 3,500
gallons of gelled water. Several wells installed in the early 1990s by Betop, Inc. were
fractured using 4,000 to 15,000 gallons of a solution with 2 percent potassium chloride
(KC1) in water. Sand was used to prop the fractures open in five of these wells (Quarterly
Review, 1993). However, hydraulic fracturing experienced little success in this basin.
Fractured wells produced poorly because the permeable, shallow subbituminous coals
collapsed under the pressure of the overburden after they were dewatered (Lyman, 2001).
The Powder River Basin contains coals of high permeability. The permeability is so high
in many areas that drilling fluid (typically water) is lost when drilling the coalbeds.
Many times drilling mud is substituted to prevent loss of circulation (DeBruin, 2001).
Because of this high permeability, most coalbed wells in the Fort Union Formation can
be drilled and completed without the use of hydraulic fracturing (DeBruin, 2001;
Quarterly Review, 1993). This has been confirmed by USGS officials in Wyoming
(Brooks, 2001). Hydraulic fracturing is also avoided to prevent fracturing of
impermeable formations adjacent to the coal, such as shales, that prevent the migration of
groundwater. It is thought that fracturing the shale would increase the amount of water
flowing into the wells. When fracturing has been done, it has been with water or
sand/water mixtures. Unspecified "modest" improvements in coalbed methane gas flow
have been observed (Quarterly Review, 1993).
In the Powder River Basin, two different coalbed methane sources are commonly
developed: (1) gas extraction from methane-charged dry sand layers overlying or
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-6
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
interbedded with the coals, and (2) conventional methane extraction from the water
saturated coal seams. In the eastern (up dip) portion of the basin, the coals in the
Wyodak-Anderson seam are relatively shallow and interbedded with sands (Montgomery,
1999) (Figure A5-6). In up dip areas above the water table, wells require minimal
dewatering for coalbed methane production because there is little to no water in the sands
(Quarterly Review, 1989; Montgomery, 1999). Coal mining operations near Gillette
have lowered the water table in the vicinity of the mines, thereby dewatering nearby
coalbeds and allowing desorption of methane gas from the coal. The sands are penetrated
using open-hole techniques, generally without any fracture treatments (Quarterly Review,
1989). Further west, down dip (Figure A5-6), the coalbed methane producing sands and
coals of the Fort Union Formation are separated from the overlying Wasatch Formation
by a poorly permeable shale of limited areal extent (Quarterly Review, 1989; Quarterly
Review, 1993). Further west, down dip (Figure A5-6) in this more water-saturated part
of the basin, coalbed methane wells are also completed as open-hole wells.
The practice of open-hole drilling is commonly used in this region. In this practice, a
portion of the borehole in the coal is drilled without any casing or well screen. Most
other regions of the country where coalbed methane is recovered use a perforated casing
throughout the target coal interval. The open coal zone is then cleaned out with water,
and the surrounding coal formation is sometimes fractured to improve recovery of the
methane. A submersible pump is set at the bottom of the target zone with tubing to the
ground surface to remove groundwater from the well. The methane gas travels up the
space between the water tubing and the casing. The well is capped to control the flow of
methane gas. Wells are often dewatered for several months before producing optimal
quantities of methane gas.
Side jetting has also been performed with some success; however, dynamic open-hole
cavitation had not been attempted as of 1993. Side jetting is the process by which water
and air are injected at high pressure to enlarge the boring in the coal seam. The cavitation
process uses dynamic pressure changes to break apart the coal and to widen the boring
within in the coal seam (Quarterly Review, 1993).
Production of coalbed methane from water-saturated coalbeds below the water table first
requires partial dewatering of the coal to allow desorption of methane from the coal.
Production from water-bearing coal seams can yield significant volumes of water;
enough to make it difficult or infeasible to dewater the formation sufficiently to initiate
coalbed methane flow (Montgomery, 1999). Tests on 11 wells reported by Crockett
(2000) indicate that coalbed methane is desorbed from coal as a consequence of
decreased hydrostatic pressure caused by pumping groundwater. One well started
desorbing at 92 percent of the original reservoir pressure. "Most drilling to date has
attempted to remain near or above the existing water table to minimize water production"
(Montgomery, 1999). Modifications to well spacing and pumping configuration have
been cited by Montgomery (1999) as showing some promise for allowing greater
production from the water-saturated coal seams in the future. Because the water in the
deeper coal seams may be original interstitial water, and recharge from meteoric water
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-7
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
might not be an important factor (Montgomery, 1999), dewatering of these coals for the
purposes of coalbed methane production might become economically feasible.
Disposal of water produced by coalbed methane wells is an issue at many well locations.
Coalbed methane wells are generally pumped constantly, removing as much as 168,000
gallons per day of water from deeper formations (Randall, 1991). Averages of 17,000
gallons per day per well are more common (Powder River Basin Resource Council,
2001). Water produced during the dewatering of coalbed wells is generally discharged to
stock ponds, water impoundments (reservoirs), drainages with ephemeral and intermittent
streams, and surface waters. A National Pollution Discharge Elimination System permit
is required for surface discharge of production water. The water is generally of potable
quality in the center of the basin, becoming more saline to the north and south. It is
sometimes used for irrigation and watering livestock (DeBruin, 2001). IDS levels are
typically less than 5,000 parts per million. The water's salt content is primarily sodium
bicarbonate (Quarterly Review, 1993). Average analytical results from 47 USGS water
quality analyses of untreated, co-produced water from coalbed methane wells in the
Powder River Basin are displayed in Table A5-1 below.
Table A5-1. Average Water Quality Results from Produced Waters (Rice et al.,
2000)
Parameter
PH
temperature
specific conductance
IDS
fluoride
chloride
sulfate
bromide
alkalinity (as HCO3)
ammonium
calcium
potassium
magnesium
sodium
barium
iron
Result
7.3
19.6
1,300
850
0.92
13.0
2.4
0.12
950
2.4
32
8.4
16
300
0.62
0.8
Units
N/A
°C
microsiemens
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
mg/L
As a result of the rapid growth in the coalbed methane industry, the Wyoming State
Engineer's Office (SEO) requested funding for drilling, equipping, and monitoring of
observation wells, and the installation of surface water measuring devices to be located in
coalbed methane production areas. These monitoring facilities would become part of the
SEO statewide observation well network to monitor changes in groundwater levels and
stream flow over time. As of 1999, work was underway, but no report of results had yet
been made available.
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
5.4 Summary
Based on the information for the Powder River Basin, the coalbeds that are being
developed, or which may be developed, for coalbed methane in the Powder River Basin
are also USDWs. Coalbeds in this basin are interspersed with sandstone and shale at
varying depths. The Fort Union Formation that supplies municipal water to the City of
Gillette is the same formation that contains the coals that are developed for coalbed
methane. The coalbeds contain and transmit more water than the sandstones. The
sandstones and coalbeds have been used for both the production of water and the
production of coalbed methane. TDS levels in the water produced from coalbeds meet
the water quality criteria for USDWs.
The information available indicates that currently hydraulic fracturing is not widely used
in this region due to concerns about the potential for increased groundwater flow into the
coalbed methane production wells and the consequent collapse of open hole wells in coal
upon dewatering. According to the available literature, where hydraulic fracturing has
been used in this basin, it has not been an effective method for extracting methane.
Hydraulic fracturing has been conducted primarily with water, or gelled water and sand,
although the recorded use of a solution of KC1 was identified in the literature.
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-9
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
CEDAR
ANTICLINE
PIG SNOWY
ANTICLINORIUM
WILLIS TON
BULL MTNS
BASIN
etc
HORN
UPLIFT
BLACK
HILLS
UPLIFT
OWL CREEK
UPLIFT
HARTVILLE
UPLIFT
WIND RIVER
BASiN
SWEETWATEH
UPLIFT
Location of the Powder River Basin of Wyoming and Montana
flarsen. 1989 and Rice at at. 2000}
ir ,*"-V 1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
A5-10
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
Iv - Current C8M Play
J
Q
O
Map o* the development of Coal bed Methane in the Powder R ver Basin
(Montgomery, 1999)
[ Figyra A5-2
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
A5-11
image:
a EL
C 3.
ts s
§* K •
o .
5'S
a
fj- ft
0 OQ
II
3. c0
OQ g
o o
VI
> M
L/i ^°
i O
i—- O
K< 4^
WEST
EAST
Gil HttP
Sundance
47651
Juraswc & -
Crctncrous Roct-s
Tnaurjc Rrjckn
Upper DrrlaccouE RixkE
LnwEir CrdacnnLW Recks
Linen
.-•,•1-1 .--I-
Conceptual Cross Section of the Powder River Basin (West-Easi) (DeBruin et al.. 2000)
u Figure AS>3
W
oo
*^
ON
o
U)
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
in
H
IT
DC
IT
UJ
h-
(/>
UJ
E
UJ
C/>
o
o
gv
6
c
SI
o
o
LU
m
c
u
O
Jj
'co
Upper
Cretaceous
(part)
STRATIGRAPHIC UNIT
While River Formation
Wasaich
Formation
Fort Union
Formation
Tongue Rivec Member
Lebo Shale Member
Tullock Member
Lance
Formation (part)
Strasigtaphie Diagram of Geology n the Po*der Rivc« Bas n
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
A5-13
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
•^ ,. .,, ^
Detailed Crass Section of the Wasatdi and Fort Union Formations in the Powder
R ver Basin (DeBruin et al.. 2000)
Figure
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
A5-14
image:
a EL
C 3.
ts s
§* K •
ft
o .
5'S
a
-
0 OQ
^ 3
3 §
o p
-
OQ
O
> M
L/i ^°
I O
i—- O
Ui 4^
Coal Bed Methane Wells
Open Pd
Coal Mine
Tongue River Member
N
LED-O and Tultock Members
w
oo
K^
ON
Conceptual Cross Secton (West-East) of tha Fort Union Formation en ihe Eastern Powder River Basin
Mean Gillette. Wyoming <De6a4in eS al.. 2000)
td 1
image:
a EL
C 3.
ts s
§* K •
ft
o .
5'S
a
-
0 OQ
^ 3
3 §
o p
-
OQ
O
> M
L/i ^°
I O
i—- O
ON 4^
W
oo
K^
ON
Numerous Thin
Coal Beds
Powder River
Wasatcn Fonrnalion
Fort union Formation
Crass Kfr-r:t nr Nfi.^r tie Gpntar o1 tie Powder River Ras r fRfirm.in
Figure A5-7
image:
a EL
C 3.
ts s
§* K •
^5 HH Q^
o r*^ • c~!
S'||
"'cro
I |
3. cn
OQ g
O O
en
Interstate 90
Late DeSmet
Ws sa tcPKForma ti on
Fort Union Formalian
(no con I;.
Conceptual Cross Section in ttie Western Powder River Basin Near Lake DeSrnet (DeBruin el al.r 2000)
t 'A1;1 Hf- PR.J,;*
Figure A5-8
w
oo
K^
ON
o
U)
image:
EPA816-R-04-003
Attachment 5
The Powder River Basin
i
CJ:
ifl
:C
[2
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
Junes 2004
A5-18
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
REFERENCES
Ayers, W.B. Jr. 1986. Coal Resources of the Tongue River Member, Fort Union
Formation, Powder River basin, Wyoming. Geological Survey of Wyoming
Report of Investigations 35, 21 p.
Brooks, Myron. 2001. (Water Resources Division, Wyoming District Office Chief)
United States Geologic Survey, Wyoming. Personal communication.
Crockett, FJ. 2000. Interim Drainage Report on Coalbed Methane Development in T.
43-52 N., R. 70-75 W., Campbell County Wyoming. Wyoming Reservoir
Management Group.
DeBruin, Rodney H., Oil and Gas Geologist. 2001. Wyoming State Geological Survey.
Personal communication.
DeBruin, Rodney H., Lyman, Robert M., Jones, Richard W., and Cook, Lance W. 2000.
Information Pamphlet 7. Wyoming State Geological Survey.
Department of Energy. 2002. Public Comment OW-2001-0002-0141 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185.
p. 33992, September 24, 2002.
Flores, R.M. and Bader, L.R. 1999. Fort Union Coal in the Powder River Basin,
Wyoming and Montana: A synthesis. U.S. Geological Survey Professional Paper
1625-A, Chapter PS, 49 p., on CD-ROM.
Gas Technology Institute (GTI) website. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Larsen, Very E. 1989. Preliminary evaluation of coalbed methane geology and activity
in the Recluse Area, Powder River Basin, Wyoming, Quarterly Review of
Methane from Coal Seams Technology, June
Law, Ben E., Rice, Dudley D., and Flores, Romeo M. 1991. Coalbed gas accumulations
in the Paleocene Fort Union Formation, Powder River Basin, Wyoming. Rocky
Mountain Association of Geologists, Coalbed Methane.
Lyman, Bob. 2001. (Coalbed Methane Expert) Wyoming State Geological Survey.
Personal communication.
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-19
image:
EPA 816-R-04-003 Attachment 5
The Powder River Basin
Martin, L.J., Naftz, D.L., Lowham, H.W., and Rankl, J.C. 1988. Cumulative potential
hydrologic impacts of surface mining in the eastern Powder River structural
basin, northeastern Wyoming. U.S. Geological Survey Water-Resources
Investigations Report 88-4046, 201 p.
Montgomery, Scott L. 1999. Powder River Basin, Wyoming: An expanding coalbed
methane (CBM) play. American Association of Petroleum Geologists Bulletin,
(August).
Osborne, Paul. 2002. USEPA Region VIIIUIC Program. Personal communication.
Petzet, G. Alan. 1997. Powder River coalbed methane output growing fast. Oil and Gas
Journal, March 10, 1997.
Powder River Basin Resource Council, February 15, 2001,
http://www.PowderRiverBasin.org
Powder River Coalbed Methane Information Council (PRCMIC). 2000. Coalbed
Methane Development Information, Sheridan Wyoming.
Quarterly Review, Methane from Coal Seams Technology. 1989. Larsen, Very E.,
Preliminary evaluation of coalbed methane geology and activity in the Recluse
Area, Powder River Basin, Wyoming. Methane from Coal Seams Technology,
June.
Quarterly Review, Methane from Coal Seams Technology. 1993. Powder River Basin
Wyoming and Montana. Methane from Coal Seams Technology, August.
Randall, A.G. 1991. Shallow tertiary gas production, Powder River Basin, Wyoming.
The Coalbed Methane, May 13-16, 1991.
Rice, C.A., Ellis, M.S., and Bullock, J.H., Jr. 2000. Water co-produced with coalbed
methane in the Powder River Basin, Wyoming: preliminary compositional data.
U.S. Geological Survey Open-File Report 00-372.
Tyler, R., W.A. Ambrose, A.R. Scott, and W.R. Kaiser. 1995. Geologic and hydrologic
assessment of natural gas from coal: Greater Green River, Piceance, Powder
River, and Raton basins. University of Texas at Austin, Bureau of Economic
Geology, Report of Investigations 228, 219 p.
Wyoming Geological Association. 1999. Coalbed Methane & Tertiary Geology, Powder
River Basin 50th Field Conference Guidebook.
Evaluation of Impacts to Underground Sources Junes 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A5-20
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
Attachment 6
The Central Appalachian Coal Basin
The Central Appalachian Coal Basin is the middle basin of three basins that comprise the
Appalachian Coal Region of the eastern United States. It includes parts of Kentucky, Tennessee,
Virginia, and West Virginia (Figure A6-1). It covers approximately 23,000 square miles, contains
six major Pennsylvanian age coal seams, and contains an estimated 5 trillion cubic feet (Tcf) of
coalbed methane (Zebrowitz et al., 1991; Zuber, 1998). These coal seams typically contain
multiple coalbeds that are widely distributed (Zuber, 1998). The coals seams, from oldest to
youngest (West Virginia/Virginia name), are the Pocahontas No. 3, Pocahontas No. 4, Fire
Creek/Lower Horsepen, Beckley/War Creek, Sewell/Lower Seaboard, and lager/Jawbone
(Kelafant et al., 1988). The Pocahontas coal seams include the Squire Jim and Nos. 1 to 7 and
Nos. 3 and 4 are the thickest and most areally extensive. The majority of the coalbed methane
(2.7 Tcf) occurs in the Pocahontas seams (Kelafant et al., 1988). The highest potential for
methane development is in a small, 3,000 square mile area in southwest Virginia and south central
West Virginia, where target coal seams achieve their greatest thickness and occur at depths of
about 1,000 to 2,000 feet (Kelafant et al., 1988). The Gas Technology Institute (GTI) reported
that the entire basin's annual production was 52.9 billion cubic feet (Bcf) of gas in 2000 (GTI,
2002).
6.1 Basin Geology
The Central Appalachian Basin is characterized structurally by broad, open, northeast-south west
trending folds that typically dip less than five degrees (Kelafant et al., 1988) (Figure A6-2). The
only documented exception to this is the Pine Mountain Overthrust Block in the southeast
portion of the basin (Kelafant et al., 1988). Faults and folds associated with this 25 mile-wide
and 125 mile-long structural feature are more intense as evidenced by overturned beds and even
brecciated zones in some locations (Kelafant et al., 1988). The overthrust block is believed to
have been transported about five miles from the southeast to the northwest (Kelafant et al.,
1988). The two dominant joint patterns within the coals are most likely due to the basin having
undergone two distinct patterns of structural deformation. These deformations include the
Appalachian Orogeny and the tectonic event associated with development of the Pine Mountain
overthrust (Kelafant et al., 1988).
The regional dip of coal-bearing Pennsylvanian strata is to the northwest at a rate of 75 feet per
mile (Kelafant et al., 1988). Sedimentation within the Central Appalachian Basin was influenced
somewhat by the Rome Trough, an Early Cambrian graben structure. Sediment deposition
during early Pennsylvanian time (about 320 million years ago) occurred to the southeast of the
Rome Trough in a rapidly but intermittently subsiding basin (Kelafant et al., 1988). As this
tectonic activity began to abate in the Central Appalachian Basin, subsidence to the northeast of
the Rome Trough began to form the Northern Appalachian Basin. However, subsidence rates in
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-1
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
the Northern Appalachian Basin were comparatively slower, enabling the formation of more
regionally extensive coalbeds (Kelafant et al., 1988).
There are three coal-bearing formations in the Central Appalachian Basin (Kelafant et al., 1988).
From deepest to shallowest, they are the Pocahontas Formation, the New River/Lee Formation,
and the Kanawha/Norton Formation. Each formation [Pennsylvanian in age (approximately 320
to 290 million years old)] is part of the Pottsville Group, and has varying nomenclature from
state to state (Kelafant et al., 1988).
The Pocahontas Formation directly overlies the Mississippian Bluestone Formation, and was
deposited in an unstable basin that was rapidly subsiding to the southeast (Kelafant et al., 1988).
This is reflected in the thickness of the formation, which is thickest in the southeast and thins to
the northwest. It also thins to the south and west due to erosion caused by the basal sandstone
member of the overlying New River/Lee Formation (Kelafant et al., 1988). The Pocahontas
Formation reaches its maximum thickness of 750 feet near Pocahontas, Virginia (Kelafant et al.,
1988). The formation consists mostly of massively bedded, medium-grained subgraywacke,
which can be locally conglomeratic (Kelafant, 1988). Gray siltstones and shales are interbedded
within the sandstone (subgraywacke) unit, and coal seams comprise about two percent of the
total thickness of the Pocahontas Formation (Kelafant et al., 1988).
The New River/Lee Formation conformably overlies the Pocahontas Formation in the
northeastern portions of the basin (i.e., there are no time gaps in the depositional record), but
there is an unconformity in the east-central portion of the basin (Kelafant et al., 1988). In the
southern portion of the basin, the New River/Lee Formation unconformably overlies the
Bluestone Formation. It is difficult to correlate this formation across state boundaries as
nomenclature varies (Kelafant et al., 1988). The overall thickness of the formation decreases
from east to west, with the thickest portion (1,000 feet) in parts of Virginia and West Virginia,
lessening to fewer than 100 feet along the Ohio River in Kentucky (Kelafant et al., 1988).
Coalbeds encountered in the New River/Lee Formation include the Fire Creek/Lower Horsepen,
Beckley/War Creek, Sewell/Lower Seaboard, and the lager/Jawbone (Kelafant et al., 1988).
These coalbeds thin and pinch-out towards the south and west; therefore, there are no equivalent
coalbeds in Kentucky and Tennessee (Kelafant et al., 1988).
The Kanawha/Norton Formation varies from a maximum thickness of 2,000 feet in West
Virginia to less than 600 feet in portions of Dickenson and Wise Counties, Virginia (Kelafant et
al., 1988). The formation is composed of irregular, thin- to massively-bedded subgraywackes
interbedded with shale. Several thin carbonate units also occur within the formation as well as
over 40 multi-bedded coalbeds.
All coal seams within the basin occur within the Pennsylvanian Pottsville Group (Figure A6-3).
Specific stratigraphic nomenclature varies from state to state within the basin. (Names used in
this summary are consistent with the West Virginia/Virginia nomenclature).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-2
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
The Pocahontas No. 3 coal seam ranges in depth from outcrop along the northeastern edge of the
basin to about 2,500 feet, with a thickness ranging up to seven feet (Kelafant et al., 1988).
Depths to the Pocahontas No. 4 coal seam are somewhat similar to those for the Pocahontas No.
3 coal seam, as the No. 4 seam overlies the No. 3 seam by roughly 30 to 100 feet. The thickness
of the No. 3 coal seam varies, with a maximum of approximately seven feet (Kelafant et al.,
1988). The Fire Creek/Lower Horsepen coalbed ranges in depth from roughly 500 feet over half
of its area, to a maximum depth of approximately 1,500 feet, with a maximum thickness of
roughly six feet (Kelafant et al., 1988). The Beckley/War Creek coalbed is approximately two to
five feet thick, and reaches to a maximum depth of about 2,000 feet (Kelafant et al., 1988). The
Sewell/Lower Seaboard coalbed is fairly shallow, less than 500 feet in depth over half the area it
covers, reaching to a depth over 1,000 feet in one small area. While this coal ranges in thickness
from two to six feet, it averages about two feet in West Virginia and one foot in Virginia
(Kelafant et al., 1988). The youngest targeted coal seam, the laeger/Jawbone, is generally less
than 500 feet in depth, reaching its maximum depth of over 1,000 feet in two Virginia Counties.
The thickness of the laeger/Jawbone coal ranges from two to six feet (Kelafant et al., 1988).
Figures A6-4 through A6-9 are isopach maps for the six major coal groups of the Appalachian
Coal Basin (adapted from Kelafant, et al., 1988).
6.2 Basin Hydrology and USDW Identification
The primary aquifer in the Kentucky portion of the Central Appalachian Basin is a
Pennsylvanian sandstone aquifer underlain by limestone aquifers (National Water Summary,
1984). Water wells are typically 75 to 100 feet deep in the Pennsylvanian aquifer and commonly
produce one to five gallons per minute of water (National Water Summary, 1984). The basin is
located in a portion of the Cumberland Plateau physiographic province in Tennessee (National
Water Summary, 1984). The primary aquifer in this area is a Pennsylvanian sandstone aquifer,
comprising water-bearing sandstone and conglomerate subunits with interbedded shale and coal
(National Water Summary, 1984). Water wells are typically 100 to 200 feet deep and usually
produce 5 to 50 gallons per minute of water (National Water Summary, 1984). In Virginia, the
basin is located in a portion of the Appalachian Plateau physiographic province. The primary
aquifer in this region is the Appalachian Plateau Aquifer, a consolidated sedimentary aquifer
consisting of sandstone, shale, siltstone, and coal (National Water Summary, 1984). Water wells
are typically 50 to 200 feet deep, and commonly produce one to 50 gallons per minute of water
(National Water Summary, 1984). In West Virginia, the basin is in a portion of the Appalachian
Plateaus physiographic province of that state. The primary aquifers in this area are Lower
Pennsylvanian aquifers, which include the Pottsville Group (National Water Summary, 1984).
Wells are commonly 50 to 300 feet deep and typically produce one to 100 gallons per minute of
water (National Water Summary, 1984).
Produced water volumes from coal seams within the Central Appalachian Basin are relatively
small, typically only several barrels or less per day per well, with high total dissolved solid
(TDS) levels, usually greater than 30,000 milligrams per liter (mg/L) (Quarterly Review, 1993).
Half the states (Kentucky and Ohio) within the Central Appalachian Basin have maps to locate
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-3
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
the undulating interface between saline and freshwater aquifers. The remaining states
(Tennessee and Virginia) have no maps defining this interface. Mike Burton (2001), a geologist
with the Oil and Gas Office of the Tennessee Geology Division (TGD), reports that the state has
no data relating to coalbed methane, which suggests that little or no coalbed methane extraction
occurs inside Tennessee's borders (Burton, 2001). Luke Ewing (Ewing, 2001) of the TGD
reported that the state had no aquifer maps. Scotty Sorles (Sorles, 2001) of Tennessee's
Underground Injection Control Program mentioned that within the state, produced water disposal
methods vary on a site-by-site basis. Depending on site characteristics, all injected waters must
either be returned to the formation from which they came, or be treated to drinking water levels
prior to injection elsewhere (Sorles, 2001).
Robert Wilson, Director of the Virginia's Division of Gas & Oil, stated that there is no mapping
program for underground sources of drinking water (USDWs) or for the fresh/saline groundwater
interface in Virginia. He reported that the most potable water is found far above the coal zones
used for coalbed methane extraction, with fresh water typically found at less than 300 feet deep.
He believes most drinking water in southwestern Virginia comes from wells in fractured bedrock
aquifers or shallow coal aquifers, or, in some areas, directly from springs. Mr. Wilson also
stated that some coalbed methane exploration has moved to shallower coal seams. The
Commonwealth of Virginia has instituted a voluntary program concerning depths at which
hydraulic fracturing may be performed (Virginia Division of Oil and Gas, 2002). This program
involves an operator's determination of the elevations of the lowest topographic point and the
deepest water well within a 1,500-foot radius of any proposed extraction well (Wilson, 2001).
Hydraulic fracturing should occur at least 500 feet deeper than the lower of these two points
(Wilson, 2001).
According to Mr. Tony Scales of the Virginia Department of Mines, Minerals and Energy, coal
seams are the most permeable layers in the geologic subsurface in Virginia. For this reason,
many private wells in the coalbed methane-producing counties are finished within the coalbeds.
Mr. Scales stated that impacts to water supplies have occurred if the coal seams have been
punctured by coalbed methane well drilling. The puncture hole acts as a conduit for the flow of
water out of the coals and into lower formations. The puncture hole also allows methane to rise
up to the surface (Virginia Department of Mines, Minerals, and Energy, 2002).
The following table contains information concerning the relative locations of the base of the zone
of fresh water and potential methane-bearing coalbeds in the Central Appalachian Coal Basin.
The table provides useful information that can help in determining whether coalbeds being used
or slated for methane development lie within USDWs. Note that the 10,000 mg/L level of TDS
in groundwater is the water quality criterion for a USDW. The depth to the USDW will thus lie
well below the fresh water/ saline water interface. The area of focus for coalbed methane
exploration in the basin only covers parts of Virginia and West Virginia (Figure A6-1). In
Virginia, the depth to the base of fresh water is approximately 300 feet, whereas the depths to the
bases of USDWs are greater. Thus, as can be seen in Table A6-1, methane-producing coalbeds
could lie within USDWs in Virginia. West Virginia's interface between fresh and saline water
(Foster, 1980) is based on a qualitative assessment, and is estimated at 280 to 730 feet. Again,
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-4
image:
EPA816-R-04-003
Attachment 6
The Central Appalachian Basin
the depths to USDWs are greater, and thus the coalbeds of interest could lie within potential
USDWs in West Virginia. Finally, in Kentucky the interface between fresh water and saline
water is based on a TDS level of 1,000 mg/L (Hopkins, 1966). Although the depths to methane-
producing coalbeds in Kentucky are not listed in the Table A6-1, it is possible that, as in Virginia
and West Virginia, such depths could be lower than the base of USDWs in Kentucky.
Table A6-1. Relative Locations of USDWs and Methane-Bearing Coalbeds
C \'[ Ir jl
A ;i|i jl/'.'iu L ("ii.il
lt:i-. Ill Stlltr* i.'l..::
lifgf F'.'Ju'M'Iniif
Se-wt li'i.»w*r
Utrfc.1* yftt if Cr»f It
Fh r < i <rtb I.HH tt
j'^l-llbMBllis Nit J
?«..»,., *...!
Te*SfM«
Dt-fiili tii ifrfi
nfO.lJll
in-
v.v
S A
S-A
'N A
K,,
S A
Wj««r
• hi
N ,%
Virginia
TiC|"lll fll I4|i
il .;"II».F
jlh
0 I" iA'iji'i
iljn •„• |i.ii.n'i
5 iln vi 211 Hi.1
iljri !«,! \-,w>
111" Vi Jl'liH.'
5i.ir< vi Iiiiii'1
flfl-Ms.ll
TJ* alf1"
• !ii
•- .KXi
Wtst Virgil in
ol Cu^l'
0 (H ' iJllH.i
•s,,,-^;
• SllfHii
lin.'if!
MI. i :.
I'M c
11 i) ii •
il 1 |,
' ^1 •!
;r:::;
.,,„-,.
kf hill', k1,
rT!' !',,•"''
<•; \
v .',
^ A
S-A
V ,'«
S A
flf l-M-sh
• Hi
• " II. • i ji M
Nrtlc Tin- I>MC" f.f •lliiili 'A'jla" |> n»l ' .r -i_,e iil'llie I.:KDU I j IV-.IMIH m*.'l Ol II'S fill . . Ilia- - '-il.l .Sefil'it- llif l-.i.= --I f i I'SIl*'
A'llt- is «.'ilIsE ii« f.'Sljty mil I!K I"- -f -I ft; o n,iwr Udbc-.e I he l'«e of Iht I
KcUUni ti »l . I'IKS
* ftiltull. BOO, [l«.pHil'-:.MIIllilHM. illijli .'"ill
"'Hul 4'. jiLi
6.3 Coalbed Methane Production Activity
Coalbed methane operators in the Central Appalachian Basin include Equitable Resources,
CONSOL (Consolidation Coal Company), and Pocahontas Gas Partnership, all located in
Virginia (Zuber, 1998). GTI reported that the entire basin's annual production was 52.9 Bcf of
gas in 2000 (GTI, 2002).
The Nora Field in southwestern Virginia is one of the better known coalbed methane production
fields. Equitable Resources operates the Nora Field in southwestern Virginia. According to the
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A6-5
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
Virginia Division of Gas and Oil, over 700 coalbed methane wells were drilled in the Nora Field
in 2002 and more than 1,800 coalbed methane wells were drilled in southwestern Virginia's
Buchanan County (VA Division of Gas and Oil, 2002). Foam or water is used as the fracturing
fluid and about 70,000 to 100,000 pounds of sand per well serve as proppant (Zuber, 1998).
CONSOL and Pocahontas Gas Partnership produce coal methane from coal mine developments
in Buchanan County, in southwestern Virginia (Zuber, 1998).
Many other smaller test projects were carried out in the basin in the 1970s, including the New
River Coal Company/Lick Run Mine Project, Department of Energy (DOE)/Clinchfield Coal
Company Project, U.S. Bureau of Mines (USBM)/Occidental Research/Island Creek Coal
Company Project, Gas Research Institute/Wyoming County Co-op Project, USBM Federal No. 1
Project, and the Consolidation Coal Company/ Kepler Mine Project (Hunt and Steele, 1991).
These projects were very small (five wells or fewer) and achieved limited success in terms of
production. During development of some wells in the DOE/Clinchfield Coal Company project
and the USBM Federal Project No. 1, fracture treatments "screened out" (i.e., the proppant
placement failed), affecting those coalbed methane wells' production viability.
No coalbed methane production occurred in Tennessee between 1995 and 1997 (Lyons, 1997).
Three coalbed methane wells produced gas from 1957 to 1980 in Harlan County, Kentucky, and
only one test well was in production in the early 1990s in eastern Kentucky (Lyons, 1997). The
Kentucky Department of Mines and Minerals website (2002) indicated that 1,338 gas wells were
in operation in Kentucky at the end of 2000, but no indication was given whether these were
coalbed methane wells or conventional gas wells.
In August 2001, EPA attended a hydraulic fracturing field visit in the Central Appalachian coal
basin in Virginia. Pocahontas Oil & Gas, a subsidiary of Consol Energy, Inc., invited EPA
personnel to a well location where a hydraulic fracturing treatment was being performed by
Halliburton Energy Services, Inc. This treatment employed a variety of fluids and additives to
create fractures in select coal seams at various depths. The main fracturing fluid was nitrogen
foam (70% nitrogen / 30% water mixture). Prior to injection of the foam, 6 barrels of 15 percent
hydrochloric acid were introduced into the well to dissolve the grout surrounding the injection
perforations. Once the fracture was propagated to its maximum extent, 16/30 sand suspended in
a 10-pound linear gel was injected to prop the fracture open. All the fluids and additives used
were produced by Halliburton, including a scale inhibitor and a microbicide additive.
Halliburton staff stated that typical fractures range in length from 300 to 600 feet from the well
bore in either direction, but that fractures have been known to extend from as few as 150 feet to
as many as 1,500 feet in length. According to the fracturing engineer on-site, fracture widths
range from one eighth of an inch to almost one and a half inches (Virginia Site Visit, 2001).
Once a well is drilled and fractured in Virginia, several weeks might elapse before fracturing
fluid flowback is initiated because a pipeline system must be constructed to transport the
produced coalbed methane away from the well. Flowback fracturing fluids are collected in lined
pits and tanks and transported off-site for disposal. The State of Virginia does not regulate the
use of any drilling or fracturing fluids (Wilson, 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-6
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
6.4 Summary
The area with the highest potential for coalbed methane production in the Central Appalachian
Coal Basin is southwestern Virginia (Dickenson and Buchanan Counties) and southern West
Virginia (Wyoming and McDowell Counties) (Figure A6-1). The coal seams achieve their
greatest thickness in these regions and occur at depths of approximately 1,000 to 2,000 feet.
Based on Table A6-1, methane-producing coal may lie within a USDW, providing the potential
for impact of water supplies.
Hydraulic fracturing is common practice in this region. Foam and water are the fracturing fluids
of choice and sand serves as the proppant. Because most of the coal strata dip, a coalbed
methane well's location within the basin may determine if hydraulic fracturing during the well's
development will likely affect water quality within the surrounding USDW. For instance, on the
northeastern side of the basin, the depth to the Pocahontas No. 3 coalbed is less than 500 feet.
This depth increases to over 2,000 feet in the western portion of the basin, in the direction of the
coal seam dip. Therefore, a well tapping this coal seam in the western portion of the basin may
be below the base of a USDW but a well tapping this coal seam in the eastern portion of the
basin may be within a USDW. Additionally, the base of the freshwater is not a flat surface, but
rather an undulating one. These factors indicate that the relationship between a coalbed and a
USDW must be determined on a site-specific basis.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-7
image:
O o W
o ^ •<
E, O £i
cr 3. c
en ^ en
2 K g-
< VJ °
2. &C
a g a
.
a o
OQ
1
K<
O
O
West Virginia
Area of Highest Methane Concentration (Keiafart el al., 1968)
W
hd
>
oo
ON
O
4=.
O
O
f
p
i.
I
ft
B ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
en ^ en
2 K g-
< VJ °
2. &C
a g a
.
a o
OQ
1
K<
O
O
,i.
Arttellne
tHry«
Slructural Features (Keiafar^t el al.« 1988}
w
hd
>
oo
ON
O
4=.
O
O
f
O
ft
u
-
ft
B ON
image:
EPA816-R-04-003
Attachment 6
The Central Appalachian Basin
TW ft **
t *
fe
m
3
c
19
C
C
d
I
I
J.
i'l
*
5'
*
;[ 1
i $!
i
t"i
„"* *
i'l
i
il
IAHfiEL£.QALfiED&
i Virginia na-Tift'Vwglnli namt)
Representative Stratigrapnlc Column of Pennsylvanian Age Formations
(Ketafanteial., 1988)
hgirn
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A6-10
image:
O o W
o ^ •<
E, O £i
cr 3. c
S B' B
ST CTQ -^
CD Q
a
5 2
OQ C
> i
£ I
Isopach Map: Thickness of Cover Over the Pocahonlas f4d. 3 Coated (Kelafanlei a , 1988)
w
hd
>
oo
ON
?
O
-^
O
o
L.J
f
O
i.
I
3
B ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
S P B
ST CTQ _
CD Q
a
5 2
OQ C
> i
£ I
K» 4^
Map Thickness of Cover Over trie Poca^iioonld& No 4 Coalbed {Kelafant el at., 1988}
w
hd
>
oo
ON
%
O
-^
O
o
U)
f
O
I,
I
3
B ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
S P B
ST CTQ _
CD Q
a
5 2
OQ C
> i
£ I
OJ 4^
feopach Map: Thickness of Cover Ove-r Ihc r ire Crcu-^ - Lcwar horsep&n Coalbed
cl al,,
w
hd
>
oo
ON
O
o
figure Afl-6
f
O
I,
t
pT >
O pf-
E3"1 p;
td I
g
B' ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
S B' B
ST CTQ -^
CD Q
a
5 2
OQ C
> i
£ I
LM »!
f «* Hi-i
Isop^ch Map: Thickrics*. of Cover Ovc? Bccklc-y - War Cruok Coa bed (Kcfafent ct al.f 198S)
w
hd
>
oo
ON
?
O
-^
O
o
L.J
f
O
i.
I
•A» '« F i g Lre A 6 •• 7
3
B ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
S P B
ST CTQ _
CD Q
a
5 2
OQ C
> i
£ I
(Ji 4^
r1,. /
V- <V V.
r "~~ *» *%
Central Appalachian
off Cover war
Coal
Map: Thickness of Cover Over ihe Sewell/Lower Seaboard Coalbed
(KeEafantelal., 198fi)
w
hd
>
oo
ON
^
O
-^
O
o
U)
f
O
I,
I
An B
ft
B ON
image:
O o W
o ^ •<
E, O £i
cr 3. c
S P B
ST CTQ _
CD Q
a
5 2
OQ C
> i
£ I
ON 4^
\
VIRGINIA
y \
V
2^
Central Appalachian Basin
• Of Caver Over tno
Coat
L*g*fl* -^ fa,, Lsfi* in
1 1 AJ«*I Krt«Tt«! ftad
Isopach Map: Thckness of Cover Over me Leagen'Jawbor-e Coaoed (Ke afsnt al al.,
Figure AtJ-f
W
hd
>
oo
ON
%
O
-^
O
o
U)
f
p
i.
I
3
B ON
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
REFERENCES
Burton, Mike. January, 2001. Tennessee Geology Division. Personal communication.
Ewing, Luke. January, 2001. Tennessee Geology Division. Personal communication.
Foster, James B. 1980. Fresh and saline ground-water map of West Virginia. U. S. Geological
Survey, West Virginia Geological and Economic Survey, Map WV-12.
Gas Technology Institute (GTI) website, 2002. Drilling and Production Statistics for Major US
Coalbed Methane and Gas Shale Reservoirs.
http: //www. gastechnol ogy. org
Hopkins, Herbert T. 1966. Fresh-saline water interface map of Kentucky. U. S. Geological
Survey, Kentucky Geological Survey, Series X.
Hunt, A. M., and Steele, D. J. 1991. Coalbed methane development in the Northern and Central
Appalachian Basins - past, present and future. The 1991 Coalbed Methane Symposium,
The University of Alabama/Tuscaloosa, May 13-16, 1991.
Kelafant, J. R., Wicks, D. E., Kuuskraa, V. A. March, 1988. A geologic assessment of natural
gas from coal seams in the Northern Appalachian Coal Basin. Topical Report - Final
Geologic Report (September 1986 - September 1987).
Kentucky Department of Mines and Minerals website, 2002. 2000 Annual Report.
http ://www. caer .uky. edu/kdmm/arOO. htm
Lyons, Paul C. 1997. Central-Northern Appalachian Coalbed Methane Flow Grows. Oil & Gas
July 7, 1997, pp. 76-79.
Quarterly Review. 1993. Coalbed methane - state of the industry. Quarterly Review, August,
1993.
Scales, Tony. 2001. Virginia Division of Mines, Minerals and Energy. Personal
communication.
Sorles, Scott. February, 2001. Tennessee Underground Injection Control Program. Personal
communication.
United States Geological Survey. 1973. State of Kentucky, 1:500,000 topographic map.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-17
image:
EPA 816-R-04-003 Attachment 6
The Central Appalachian Basin
United Stated Geological Survey. 1984. National Water Summary. Hydrologic events, selected
water-quality trends, and ground-water resources. United States Geological Survey
Water-Supply Paper No. 2275.
Virginia Site Visit. 2001. EPA observed hydraulic fracturing performed by Halliburton,
Inc. for Consol Energy (VA), August 9, 2001.
Virginia Division of Oil Gas & Oil. 2002. Public Comment OW-2001-0002-0084 to "Draft
Evaluation of Impacts to Underground Sources of Drinking Water by Hydraulic
Fracturing of Coalbed Methane Reservoirs." Federal Register. Vol. 63, No. 185. p.
33992, September 24, 2002.
Wilson, Robert. February, 2001. Director, Virginia Division of Gas & Oil, Department of
Mines, Minerals, and Energy. Personal communication.
Zebrowitz, M. J., Kelafant, J. R., and Boyer, C. M. 1991. Reservoir characterization and
production potential of the coal seams in Northern and Central Appalachian Basins.
Proceedings of the 1991 Coalbed Methane Symposium, The University of
Alabama/Tuscaloosa, May 13-16, 1991.
Zuber, Michael D. 1998. Production characteristics and reservoir analysis of coalbed methane
reservoirs. Lyons, Paul C. (editor). Appalachian coalbed methane. International Journal
of Coal Geology, 38 (l-2):27-45. Meeting: Appalachian coalbed methane, Lexington,
KY, United States, Sept. 27-30, 1997.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A6-18
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
Attachment 7
The Northern Appalachian Coal Basin
The Northern Appalachian Coal Basin is the northernmost of the three basins comprising
the Appalachian Coal Region of the eastern United States, and includes parts of the States
of Pennsylvania, West Virginia, Ohio, Kentucky, and Maryland (Figure A7-1). The
basin trends northeast-southwest and the Rome Trough, a major graben structure, forms
the southeastern and southern structural boundaries (Kelafant et al., 1988). The basin is
bounded on the northeast, north, and west by outcropping Pennsylvanian-aged sediments
(Kelafant et al., 1988). The basin lies completely within the Appalachian Plateau
geomorphic province, covering an area of approximately 43,700 square miles (Adams et
al., 1984 as cited in Pennsylvania Department of Conservation and Natural Resources,
2002). It consists of six Pennsylvanian age coal units, and contains an estimated 61
trillion cubic feet of coalbed methane (Kelafant et al., 1988). Coal seam depths range
from surface outcrops to up to 2,000 feet below ground surface, with most coal occurring
at depths shallower than 1,000 feet (Quarterly Review, 1993). Annual coalbed methane
production stood at 1.41 billion cubic feet in 2000 (GTI, 2002).
7.1 Basin Geology
The six Pennsylvanian aged coal zones located within the Northern Appalachian Coal
Basin are the Brookville-Clarion, Kittanning, Freeport, Pittsburgh, Sewickley, and the
Waynesburg. These coal units are contained within the Pottsville, Allegheny, and the
Monongahela Groups (Figure A7-3) (Kelafant et al., 1988).
In the Northern Appalachian Basin, the Pottsville stratigraphic group is generally 200 to
300 feet thick, and thins to the north and west into Ohio and Pennsylvania (Kelafant et
al., 1988). The coals in this group are interbedded with fluvial and deltaic sands and
shales and are capped by marine limestones and shales (Kelafant et al., 1988).
Deposition of this group took place on irregular Mississippian terrain, forming thin and
erratic coals (Kelafant et al., 1988).
The Allegheny Group reaches a maximum thickness of 200 to 300 feet in western
Maryland and thins westward to about 150 to 200 feet in Ohio. Deposition of this group
occurred as cyclothem-type sedimentation, resulting in a complex sequence of lenticular,
thin- to massive-bedded subgraywacke, shale, and mudstone interbedded with clays and
coal (Kelafant et al., 1988). Due to their alluvial and delta plain depositional
environments, Allegheny coals, which include the Brookville/Clarion, Kittanning, and
the Freeport, are 2 to 6 feet thick and aerially extensive (Kelafant et al., 1988). The
coalbeds decrease in number from the eastern to the western edge of the basin (Kelafant
etal., 1988).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-1
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
The Monongahela Group was deposited primarily in lacustrine and swamp environments.
Some of the most important economic coals of the basin were deposited in large lakes,
such as the Pittsburgh and Sewickley coals (Kelafant et al., 1988). The Monongahela
Group is thickest along the Monongahela River at 400 feet and thins to 250 feet in the
southwest along the Ohio River. Shales, mudstones, and freshwater limestones are the
major rock types of the Group (Kelafant et al., 1988). The Waynesburg coals are also
contained within the Monongahela Group. In general, the coalbeds of the Monongahela
Group are laterally extensive.
The total thickness of the Pennsylvanian-aged coal system averages 25 feet; however,
better-developed seams within the coal zones can increase in thickness by up to twice the
average (Quarterly Review, 1993). Within the Pennsylvanian Coal System, the deepest
coal, the Brookville-Clarion, ranges in depth from surface exposures in anticlines to
2,000 feet below ground surface. The Kittanning Group reaches a maximum depth of
2,000 feet and is approximately 800 feet deep in more than half the area in which the
group occurs. The distance between the Upper and Lower Kittanning is approximately
100 feet (Kelafant et al., 1988). Freeport coals are at a maximum depth of 1,800 feet in
the central portion of the Northern Appalachian Coal Basin. The Upper and Lower
Freeport are separated vertically by a distance of 40 to 60 feet (Kelafant et al., 1988).
The Pittsburgh coals achieve a maximum depth of 1,200 feet and roughly half of the
coals can be found at depths greater than 400 feet (Kelafant et al., 1988). Sewickley
coals are deeper than 400 feet with the deepest coals located at 1,200 feet below ground
surface (Kelafant et al., 1988). The final and youngest group discussed here, the
Waynesburg group, is the shallowest, reaching a maximum depth of 800 feet in the center
of the basin. Figures A7-4 through A7-9 (adapted from Kelafant et al., 1988) are isopach
maps of sediment cover for the six major coal zones of the Appalachian Coal Basin.
7.2 Basin Hydrology and USDW Identification
The Northern Appalachian Basin is situated in the Appalachian Plateaus physiographic
province of the region. The primary aquifer in this area is a Pennsylvanian sandstone
aquifer underlain by limestone aquifers (National Water Summary, 1984). Water wells
are typically 75 to 100 feet in depth in the Pennsylvanian aquifer and commonly produce
one to five gallons per minute of water (National Water Summary, 1984). The primary
aquifers in the Maryland portion of the basin are Appalachian sedimentary aquifers,
which are mostly sandstones, shales, and siltstones with some limestone, dolomite, and
coal. Water wells here are typically 30 to 400 feet in depth and usually produce 10 to
100 gallons per minute of water (National Water Summary, 1984).
In Ohio, the primary aquifers are sandstone aquifers, shaly sandstone and carbonate
aquifers, and coarse-grained aquifers (comprised of alluvium and glacial outwash)
associated with river valleys (National Water Summary, 1984). Water wells within these
aquifers typically range from 25 to 300 feet in depth, and common water production rates
vary between 1 and 500 gallons per minute (National Water Summary, 1984).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-2
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
In Pennsylvania, the primary aquifers are sandstone and shale aquifers, with smaller
unconsolidated sand and gravel aquifers surrounding river courses (National Water
Summary, 1984). Well depths in the sandstone and shale aquifers in Pennsylvania are
usually 80 to 200 feet in depth, and the wells typically produce 5 to 60 gallons per minute
of water (National Water Summary, 1984).
In West Virginia, the primary aquifer is an Upper Pennsylvanian-aged aquifer consisting
of the Dunkard, Monongahela, and Conemaugh Groups (National Water Summary,
1984). This aquifer consists of nearly horizontal beds of shale, sandstone, siltstone, coal,
and limestone (National Water Summary, 1984). Water wells typically extend from 50 to
300 feet in depth in this area of West Virginia, and commonly produce 1 to 30 gallons per
minute of water (National Water Summary, 1984).
Individual states containing portions of the basin have developed various maps and
documents locating underground sources of drinking water (USDWs) and aquifers within
their state boundaries, mostly as a part of their respective Underground Injection Control
(UIC) Programs. EPA's Regional Office also has information concerning the location of
these resources, as not all states within the Northern Appalachian Coal Basin have
primacy over their UIC Program. Water quality data from eight historic Northern
Appalachian Coal Basin projects show that estimated total dissolved solids (IDS) levels
ranged from 2,000 to 5,000 milligrams per liter (mg/L) at depths ranging from 500 to
1,025 feet below ground surface (Zebrowitz et al., 1991), well within EPA's water quality
criterion for a USDW of less than 10,000 mg/L of IDS (40 CFR §144.3).
Most states within the Northern Appalachian Basin, including Kentucky, Ohio, and West
Virginia have mapped the interface between saline and freshwater aquifers. For
Maryland and Pennsylvania, no maps have been identified that define the interface
between saline and freshwater aquifers. In Maryland, a deep well drilled in southern
Garrett County encountered the fresh/saltwater interface at a depth of 940 feet (Duigon
and Smigaj, 1985). Groundwater in Pennsylvania deeper than 450 feet is not considered
to be a USDW (Platt, 2001) because of the existence of non-water producing shale from
450 to 1000 feet, and IDS levels in water below this shale that are typically greater than
100,000 mg/L. The following table contains information concerning the relative location
of potential USDWs and potential methane-bearing coalbeds in the Northern Appalachian
Coal Basin.
As shown in Table A7-1, coalbeds with methane production potential in the Northern
Appalachian Basin do occur within USDWs, indicating the potential for impact. West
Virginia's interface line between fresh and saline water (Foster, 1980) is based on a
qualitative assessment, Ohio's interface line is based on a TDS level of 3,000 mg/L
(Sedam and Stein, 1970), and Kentucky's interface line is based on a TDS level of 1,000
mg/L (Hopkins, 1966). In Maryland, the fresh water distinction was probably made
based on a TDS level of 1,000 mg/L, as the reference refers to sodium and chloride
concentrations of 1,800 mg/L and 2,900 mg/L as "high levels" (Duigon and Smigaj,
1985).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-3
image:
E. D EL
I
5'I a
ti- o
° ^
11
3. en
OQ g
O O
C/l
1
K»
O
O
Northern
Apf alaeUan < '»ul
Burin, St*lw nid
( II jl ( |HHL]>*
WajKitalnurg
SifwitUy
PiimlMiiigh
Fnqpriutt
Kitt«nmnjg^
Bru*ik>iLl€'.'t1ariu'ii
P*iin*>1vaMigi
L1-: *:. '• • "
' ' ' t
ttto 8"1'1
'1 W 12''H'I
1 Mo 1 2M '
Oto IfrJO
1 1 t<t ' i' In
* livic
= .-•*,:-,,
- 4?n
We« Virgaia
j "
i) to 8" » i
o 1* L'Wi-
«i to L2M
Mlo
(I |9 ] I'HH)
••••ll.Hllo
Jlllil«
f T* f
'i, " !
»p ,
- l5<Mo
54««i'
OU«
.
* i • * * •
II f-i 4 HI
M|,T \lll '
II t-l b^l't , .
C? ii> - IJ i » • 500
I' •••• ' 'Ml''
• IMC |u
1.!""
Kentucky
: T- ' ••:
,i
n tn 4«-t
n to -- H-«
n m • 4«"
11 ti; • 4f>i
inn • I<M
• S»i.«'
f. . n
l a •
~r»\ V, i-r
ifV>
- luuto
500
Mirykand
1'^.\ •}. '- :• I
•1 " ..
n 10 ';*»'
n (0 •: 41111
ii |p «:4ni'i
l"| (o •:4l)l)
II (ij •:4nu
"SOU
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
Therefore, in these states, the depth to the 10,000 mg/L level of TDS in groundwater is
potentially and likely deeper than the depths presented above (Table A7-1). This
assumption is confirmed by a structure elevation map (Figure A7-6) of the Upper and
Lower Freeport Sandstones of the Upper Allegheny Group (Figure A7-3) in Ohio. With
the increasing depth of these strati graphic units toward the basin center, much of the
formation waters in these units south of the easternmost counties in Ohio contain TDS
levels in excess of 10,000 mg/L (Vogel, 1982). Likewise, the Pittsburgh Group Coals in
Pennsylvania range in depth from outcrop to 1,200 feet below ground surface (Figure A7-
7). Over this length of "dip", it is likely that the coals intersect drinking water aquifers
before they reach depths where TDS levels exceed the 10,000 mg/L TDS water quality
criterion of a USDW.
7.3 Coalbed Methane Production Activity
Coalbed methane has been produced in commercial quantities from the Pittsburgh
coalbed of the Northern Appalachian Coal Basin since 1932 (Lyons, 1997), after the
1905 discovery of the Big Run Field in Wetzel County, West Virginia (Hunt and Steele,
1991). Coalbed methane production development in the Northern Appalachian Basin has
lagged, however, due to insufficient reservoir knowledge, inadequate well completion
techniques, and coalbed gas ownership issues revolving around whether the gas is owned
by the mineral owner or the oil and gas owner (Zebrowitz et al., 1991). Annual coalbed
methane production stood at 1.41 billion cubic feet in 2000 (GTI, 2002). As of October
2002, 185 coalbed methane wells were producing coalbed methane in Pennsylvania
(Pennsylvania Department of Conservation and Natural Resources, 2002). Discharge of
produced waters has also proven to be problematic (Lyons, 1997) for coalbed methane
field operators in the Northern Appalachian Coal Basin.
Some operators in the Northern Appalachian Coal Basin and several test projects are
discussed below. As of 1993, O'Brien Methane Production, Inc. had at least 20 wells in
southern Indiana County, Pennsylvania (Quarterly Review, 1993). They received a water
treatment and discharge permit that allowed O'Brien to discharge produced water into
Blacklick Creek. The wells in O'Brien's field were hydraulically fractured with water
and sand. Nitrogen was being contemplated for future fracturing. O'Brien's operations
have since been assumed by Belden and Blake. BTI Energy, Inc. also had a few coalbed
wells in northern Fayette County, Pennsylvania. Two were completed in 1993 and the
firm held permits for eight additional wells.
Other projects in the basin included the Lykes/Emerald Mines Project of the United
States Bureau of Mines (USBM) and the Penn State University/Carnegie Natural
Gas/U.S. Steel Wells Project, both in Greene County, Pennsylvania. Depths to the top of
the Pittsburgh coals in Greene County range from 800 to 1,200 feet below ground surface
(Kelafant et al., 1988). Hydraulic fracturing fluids included water and sand, and nitrogen
foam and sand (Hunt and Steele, 1991). The Christopher Coal Company/Spindler Wells
Project, which took place from 1952 to 1959, fractured one well with 12 quarts of
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-5
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
nitroglycerin (Hunt and Steele, 1991). In the Vesta Mines Project of Washington County,
Pennsylvania, the USBM used gelled water and sand to complete five wells in the
Pittsburgh Seam (Hunt and Steele, 1991).
Within the State of Pennsylvania, there have been complaints of methane migrating into
water supplies (Markowski, 2001). According to the Pennsylvania Department of
Conservation and Natural Resources (2002), none of these complaints were linked
specifically to hydraulic fracturing of coalbed methane wells. During a telephone
interview (Markowski, 2001), Ms. Markowski stated methane contamination is due to the
fact that many coalbed methane wells in southwestern Pennsylvania are completed in
abandoned mine shafts. A puncture in the roof of the mineshaft provides a migration
pathway for methane into overlying groundwater. These wells are known as gob wells,
and are not usually hydraulically fractured or stimulated.
7.4 Summary
Based on available information, coal seams with methane production potential are located
within USDWs throughout the Northern Appalachian Coal Basin, and hydraulic
fracturing takes place in this basin. Because most of the coal strata dip, a well's location
within the basin determines whether it is within a USDW, and whether the potential for
impact exists. For example, in the Pittsburgh Coal Zone in Pennsylvania, the depth to the
top of this coal zone varies from outcrop to about 1,200 feet in the very southwestern
corner of the state. The approximate depth to the bottom of the USDW is 450 feet.
Therefore, production wells operating down to approximately 500 feet could potentially
be hydraulically connected to the USDW. However, those wells operating at depths
greater than 900 feet would probably not be hydraulically connected to the USDW,
unless a fracture extending beyond the coal layers to the shallower aquifer was to occur.
Milici (2002) indicated that the Pittsburgh Coal in Pennsylvania is mined out along its
outcrop and the remaining coal resources are deeper (> 450 feet) in the basin. While this
situation would greatly minimize the possibility of water quality impacts for this coal
zone in Pennsylvania, the potential for contamination from the Pittsburgh coalbeds in
other states within the basin still exists (see Table A7-1).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-6
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
""'- Sfy
<—•""* i i* .1
*" r\
•\ /
f \_
Coytity etal.,
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-7
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
Map at al.,
A7-2
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
E
£
«
>>
en
Permian
i
Fennsylvanian
Group
Dunkard
Monongahela
Conemaugh
Allegheny
Pottsville
Coal
Group
Waynes burg
Sewickley
Pittsburgh
Freeport
Kitttnning
Brookvi lie Clarion
Generalized Stratigraphic Column of the Northern Appalachian Coal Basin
(KelafantetaL, 1988)
iv^4iji fe iC'j?^F.r-feAo^-i^a Fipur© A7^3
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-9
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
*% Ltn* IB
w^Sfl ol Mo Ceal
, WEST VIRGINIA •-•"'
* V \
0 1C K f)
mg^^tySSSi
SCAUi ;)' ».(,E5
of Cover to the Brookville/Clarion
(Keiafantetaf.,
Ik
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-10
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
in
i^fs^ *-reas ol No
. MARYLAND'S,,*.
• / ,-_*„ •'"' -
Depth of Cover to the Kltanning
kk A7-5
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-11
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
*"***%__ In
of Mo
. ^
/***,— -"*,
>v
/
>X'\- '
/ .-*'-; u
of Cover to the
A7"6
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-12
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
In
to the
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-13
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
Line in
/ .X % «• —^ .'' 4
r^/,- >.;
of to the
ik A7-B
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-14
image:
EPA816-R-04-003
Attachment 7
The Northern Appalachian Basin
*% in
A / —"
-i /v ;
Depth of Cover to the Waynesburg
(Kelafant et al., 1988)
A7-i
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A7-15
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
REFERENCES
Adams et al., 1984, as cited by Pennsylvania Department of Conservation and Natural
Resources, 2002) Adams, M.A., Eddy, G.E., Hewitt, J.L., and others. 1982.
Northern Appalachian coal basin report - A study of Carboniferous geology, coal,
and the potential coalbed methane resources of the Northern Appalachian coal
basin in Pennsylvania, Ohio, Maryland, West Virginia and Kentucky: McLean,
VA, TRW Inc., TRW Coalbed Methane Program, Report to DOE/METC,
Contract No. DE-AC21-81 MC 14900, 179 p.
Duigon, Mark T. and Smigaj, Michael J. 1985. First report on the hydrologic effects of
underground coal mining in Southern Garrett County, Maryland, U.S. Geological
Survey Report of Investigations No. 41.
Foster, James B. 1980. Fresh and saline ground-water map of West Virginia. U. S.
Geological Survey, West Virginia Geological and Economic Survey, Map WV-
12.
Gas Technology Institute (GTI) Web site. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Hopkins, Herbert T. 1966. Fresh-saline water interface map of Kentucky. U.S.
Geological Survey, Kentucky Geological Survey, Series X.
Hunt, A. M., and Steele, D. J. 1991. Coalbed methane development in the Northern and
Central Appalachian Basins - past, present and future. Proceedings of the 1991
Coalbed Methane Symposium, The University of Alabama/Tuscaloosa, May 13-
16, 1991.
Kelafant, J. R., Wicks, D. E., and Kuuskraa, V. A. March, 1988. A Geologic
Assessment of Natural Gas from Coal Seams in the Northern Appalachian Coal
Basin. Topical Report - Final Geologic Report to the Gas Research Institute
(September 1986 - September 1987).
Lyons, Paul C. 1997. Central-Northern Appalachian Coalbed Methane Flow Grows. Oil
& Gas July 7, 1997, pp. 76-79.
Markowski, Toni. January, 2001. Pennsylvania Geological Survey. Personal
communication.
Milici, R.C. February, 2002. U. S. Geological Survey. Personal communication.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-16
image:
EPA 816-R-04-003 Attachment 7
The Northern Appalachian Basin
National Water Summary. 1984. Hydrologic events, selected water-quality trends, and
ground-water resources. United States Geological Survey Water-Supply Paper
No. 2275.
Pennsylvania Department of Conservation and Natural Resources. 2002. Public
Comment OW-2001-0002-0089 to "Draft Evaluation of Impacts to Underground
Sources of Drinking Water by Hydraulic Fracturing of Coalbed Methane
Reservoirs." Federal Register. Vol. 63, No. 185. p. 33992, September 24, 2002.
Platt, Steve. January, 2001. U.S. EPA Region 3. Personal communication.
Quarterly Review. 1993. Coalbed Methane - State of the Industry. Methane From Coal
Seams Technology, August 1993.
Sedam, A. C., and Stein, R. B. 1970. Saline ground-water resources of Ohio. Hydrologic
Investigations Atlas HA-366, Department of the Interior, U. S. Geological
Survey.
Vogel, Donald A. 1982. Final Report, U.I.C. Program, Salt/Fresh Water Interface
Ground-Water Mapping Project. Ohio Dept. of Natural Resources, Division of
Water, Columbus.
Zebrowitz, M. J., Kelafant, J. R., and Boyer, C. M. 1991. Reservoir characterization and
production potential of the coal seams in Northern and Central Appalachian
Basins. Proceedings of the 1991 Coalbed Methane Symposium, The University of
Alabama/Tuscaloosa, May 13-16, 1991.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A7-17
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Attachment 8
The Western Interior Coal Region
The Western Interior Coal Region comprises three coal basins, the Arkoma, the
Cherokee, and the Forest City Basins, and encompasses portions of six states: Arkansas,
Oklahoma, Kansas, Missouri, Nebraska, and Iowa (Figure A8-1). The Arkoma Basin
covers about 13,500 square miles in Arkansas and Oklahoma, with an estimated 1.58 to
3.55 trillion cubic feet (Tcf) of gas reserves, primarily in the Hartshorne coals (Quarterly
Review, 1993).
The Cherokee Basin is part of the Cherokee Platform Province, which covers
approximately 26,500 square miles (Charpentier, 1995) in Oklahoma, Kansas, and
Missouri. The basin contains an estimated 1.38 million cubic feet of gas per square mile
(Stoeckenger, 1990) in the targeted Mulky, Weir-Pittsburg, and Riverton coal seams of
the Cherokee Group (Quarterly Review, 1993). In total, the basin contains approximately
36.6 billion cubic feet (Bcf) of gas. However, the Petroleum Technology Transfer
Council (1999) indicates that there are nearly 10 Tcf of gas in eastern Kansas alone. The
Forest City Basin covers about 47,000 square miles (Quarterly Review, 1993) in Iowa,
Kansas, Missouri, and Nebraska, and contains an estimated 1 Tcf of gas (Nelson, 1999).
For the entire region, coalbed methane production was 6.5 Bcf in 2000 (Gas Technology
Institute (GTI), 2002).
8.1 Basin Coals
The Arkoma Basin is the southernmost of the three basins comprising the Western
Interior Coal Region, and is bounded structurally by the Ozark Dome to the north, the
Central Oklahoma Platform and Seminole Uplift on the west, and the Ouachita
Overthrust Belt to the south (Quarterly Review, 1993). Middle Pennsylvanian coalbeds
occur within the Hartshorne and McAlester Formations (Figure A8-2), as well as the
Savanna and Boggy Formations (Quarterly Review, 1993).
The Cherokee Basin is the central basin of the Western Interior Coal Region, and is
bounded on the east and southeast by the Ozark Dome, on the west by the Nehama Uplift,
and on the north by the Bourbon Arch (Quarterly Review, 1993). Principal coals occur in
the Krebs and Cabaniss Formations of the middle Pennsylvanian Cherokee Group (Figure
A8-3).
The Forest City Basin (Figure A8-4), the northernmost basin of the Western Interior Coal
Region, is a shallow cratonic depression bounded by the Nemaha Ridge to the west, the
Thurman-Redfield structural zone to the north, the Mississippi River Arch to the east, and
the Bourbon Arch to the south (Bostic et al., 1993). Methane-bearing coals occur in the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-1
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
middle Pennsylvanian Cherokee and Marmaton Groups, with the Cherokee Group being
of primary interest (Tedesco, 1992).
8.1.1 Arkoma Basin Coals
The Hartshorne coals of the Hartshorne Formation are the most important for coalbed
methane production in the Arkoma Basin. Their depths range from 600 to 2,300 feet in
two productive areas in southeastern Oklahoma (Quarterly Review, 1993). lannacchione
and Puglio (1979) estimated that 58 percent of the coalbed methane in the Hartshorne
coals in southeastern Oklahoma occurs at 500- to 1,000-foot depths. These coals can
reach depths of greater than 5,000 feet, and are three to nine feet thick (Quarterly Review,
1993). Depths to the top of the Hartshorne coal in southeastern Oklahoma range from
380 to 1,540 feet (Friedman, 1982). As of March 2000, there were 377 coalbed methane
wells in eastern Oklahoma, ranging in depth from 589 to 3,726 feet (Oklahoma
Geological Survey, 2001).
8.1.2 Cherokee Basin Coals
The primary coal seams targeted by operators in Kansas are the Riverton Coal of the
Krebs Formation and the Weir-Pittsburg and Mulky coals of the Cabaniss Formation
(Quarterly Review, 1993). The Riverton and Weir-Pittsburg seams are about 3 to 5 feet
thick and range from 800 to 1,200 feet deep (Quarterly Review, 1993). The Mulky Coal,
which ranges up to 2 feet thick, occurs at depths of 600 to 1,000 feet (Quarterly Review,
1993).
8.1.3 Forest City Basin Coals
Individual coal seams in the Cherokee Group in the Forest City Basin range from a few
inches to about 4 feet thick, with some seams up to 6 feet thick (Brady, 2002; Smith,
2002). Cumulative maximum coal thickness within the Cherokee Group is about 25 to
30 feet (Brady, 2002; Smith, 2002). Depths to the top of the Cherokee Group coals range
from surface exposures in the shallower portion of the basin in southeastern Iowa, to
about 1,220 feet in the deeper part of the basin, in northeastern Kansas (Bostic et al.,
1993). At one location in Nebraska, the depth to the Cherokee Group is about 1,396 feet,
and the base is at a depth of 2,096 feet (Condra and Reed, 1959). Maximum thickness of
the Cherokee and Marmaton Groups is about 800 feet in the southeastern tip of Nebraska
(Burchett, unpublished paper).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-2
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
8.2 Basin Hydrology and USDW Identification
8.2.1 Arkoma Basin Hydrology and USDW Identification
In Arkansas, the Arkoma Basin falls within the Interior Highlands physiographic
province (Figure A8-5). According to the National Water Summary (1984), there is no
principal aquifer in this area, only small alluvial aquifers bounding the Arkansas River
(Figure A8-5). In these alluvial aquifers, water wells typically penetrate to depths of 100
to 150 feet, and common well yields are in the order of 1,000 to 2,000 gallons of water
per minute (National Water Summary, 1984). In Oklahoma, the Arkoma Basin is
contained within the Ouachita and Central Lowland physiographic province (Figure A8-
6). Much like in Arkansas, there are no principal aquifers in this portion of the state, but
there are smaller alluvium and terrace deposits along the Arkansas, North Canadian, and
Canadian Rivers (National Water Summary, 1984) that serve as aquifers (Figure A8-6).
Marcher (1969) also identifies these smaller deposits as the most favorable for
groundwater supplies. Water well depths in the alluvium and terrace deposits of the
Arkansas River in Oklahoma typically range from 50 to 100 feet (National Water
Summary, 1984). Water well production rates in all three aquifers commonly range from
100 to 600 gallons of water per minute in alluvium, and 50 to 300 gallons of water per
minute in terrace deposits (National Water Summary, 1984).
Bill Prior, a geologist with the Arkansas Geological Commission, stated that within
Arkansas, the Arkoma Basin was in the Arkansas River Physiographic Province, which
lacks a true aquifer. Most of the rocks within this physiographic province are tight
sandstones and shales, and most communities within the province use surface water
supplies (Prior, 2001). Doug Hansen of the Arkansas Geological Commission said that
there were a few scattered bedrock wells within the Arkoma Basin (Hansen, 2001). Total
dissolved solids (TDS) levels in the McAlester Formation in Arkansas (which contains
the Hartshorne coals; Potts, 1987) range between 55 to 534 milligrams per liter (mg/L) at
depths ranging from 32.4 to 190 feet below land surface (Cordova, 1963). The base of
fresh water in the area is about 500 to 2,000 feet below ground surface (Cordova, 1963).
However, Cordova (1963) does not define "fresh water;" therefore, it is difficult to
determine if the depths reported by Cordova coincide with the base of an underground
source of drinking water (USDW).
Water quality test results from the targeted Hartshorne seam in Oklahoma have shown
the water to be highly saline (Quarterly Review, 1993). Ken Luza, a geologist with the
Oklahoma Geological Survey, stated that a hydrologic atlas prepared by the Oklahoma
Geological Survey delineated a 5,000 mg/L TDS water quality contour line in a portion
of the state, including the Arkoma Basin (Marcher, 1969; Marcher and Bingham, 1971).
Maps such as these atlas maps show that, based on water quality and rock type, very little
of the area falls within a zone "most favorable for groundwater supplies" or "moderately
favorable for groundwater supplies." Most of the area falls within a zone designated as
"least favorable for groundwater supplies" (Cardott, 2001). Pam Hudson, Manager of the
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-3
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
Geologic Section of the Oklahoma Corporation Commission, stated that the Commission
has a series of maps, one for each county in Oklahoma, showing the depth to the 10,000
mg/L IDS line (Hudson, 2001). The water quality criterion for a USDW is a IDS level
of less than 10,000 mg/L. The Oklahoma Corporation Commission maps are used to
assist drillers in complying with state regulations that require oil and gas wells to be
cased through USDWs.
The following table contains information concerning the relative location of potential
USDWs and methane-bearing coalbeds in the Arkoma Basin.
Table A8-1 Relative Locations of USDWs and Potential Methane-Bearing
Coalbeds, Arkoma Basin
Arkoma Coal
Basin, States and
Coal Group
Hartshorne Coals
Arkansas
Depth to top of
Coal1 (ft)
0 to < 4,500
Depth to base
of Fresh
Water 2'3 (ft)
500 to 2000
Oklahoma
Depth to top of
Coal l (ft)
>~1000
Depth to
base of
USDW4 (ft)
<~900
1 Andrews et al., 1998
2 Note: The base of "fresh water" is not the base of the USDW (depth to the base of the USDW is unknown
or not available). Fresh water is within the USDW and the base of fresh water is above the base of the
USDW. Cordova (1963) does not define "fresh water."
3 Cordova, 1963
4 Oklahoma Corporation Commission Depth to Base of Treatable Water Map Series (2001)
Based on Table A8-1, it can be determined that in Arkansas, there is a possibility for the
Hartshorne Coals to be located within a USDW, allowing the potential for impacts. The
potential for impacts from fracturing coalbeds below the USDW is not known. Cordova
(1963) does not specify the TDS level used to determine the depth of the base of fresh
water in the Arkansas Valley region; he merely states that it is the depth to salt water, and
he does not provide a definition of "salt water." The position of a coalbed methane well
within the basin would ultimately determine if coals and USDWs coincide, as the
Hartshorne Coals are typically shallower on basin margins (Andrews et al., 1998) and
progressively increase in depth toward the basin's center (where they are potentially too
deep to be located within a USDW).
8.2.2 Cherokee Basin Hydrology and USDW Identification
The Cherokee Basin underlies parts of the States of Kansas, Missouri, and Oklahoma. In
Kansas, the Cherokee Basin is part of the Central Lowlands and Ozark Plateaus
physiographic provinces (Figure A8-7). While the majority of this area does not contain
a principal aquifer, the Ozark and Douglas aquifers (Figure A8-7) are contained in the
basin (National Water Summary, 1984). The confined Ozark Aquifer, composed of
weathered and sandy dolomites, typically contains water wells that extend from 500 to
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-4
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
1,800 feet in depth, commonly yielding 30 to 150 gallons of water per minute (National
Water Summary, 1984). The usually unconfined Douglas Aquifer is channel sandstone
of Pennsylvanian Age (National Water Summary, 1984). Water wells are usually 5 to
400 feet deep in this aquifer and typically produce 10 to 40 gallons of water per minute
(National Water Summary, 1984).
In Missouri, only a very small portion of the basin falls within the Osage Plains area of
the Central Lowlands physiographic province (Figure A8-8). The principal aquifers in
this portion of Missouri are the Ozark and Pennsylvanian-Mississippian age aquifers
(National Water Summary, 1984) (Figure A8-8). Water well depths in the Ozark Aquifer
typically range from 200 to 1,700 feet, and those in the Pennsylvanian-Mississippian age
aquifers typically range from 100 to 400 feet in depth (National Water Summary, 1984).
Common well yields are 15 to 700 gallons of water per minute and 1 to 15 gallons of
water per minute in the Ozark and Pennsylvanian-Mississippian aquifers, respectively
(National Water Summary, 1984). Only a very small portion of the Cherokee Basin,
bounded from the Forest City Basin to its north by the Bourbon Arch, falls within the
State of Missouri (Figure A8-9). Jim Vandike, Chief of Missouri's Water Resources
Branch at the Missouri Geological Survey, stated that only two public water supplies
obtain water from Pennsylvanian strata, and those wells were outside of the Cherokee
Basin (Vandike, 2001).
In Oklahoma, the Cherokee Basin lies within the Central Lowland physiographic
province (Figure A8-6). In addition to the alluvium and terrace deposit aquifers
previously discussed in the Arkoma Basin aquifer descriptions, this area also contains the
Garber-Wellington and Vamoosa-Ada Aquifers (Figure A8-6), which are unconfined to
confined sandstone with shale and siltstone aquifers (National Water Summary, 1984).
The Vamoosa-Ada Aquifer contains some conglomerate aquifers as well. Water well
depths in these two aquifers usually range from 100 to 900 feet, and wells typically
produce from 100 to 300 gallons of water per minute (National Water Summary, 1984).
At least half of the area of this basin in Oklahoma does not contain a principal aquifer
(National Water Summary, 1984).
In Kansas, Al Macfarlane, of the Kansas Geological Survey, stated that the Ozark
Aquifer was located in the Cherokee Basin in Kansas (Macfarlane, 2001). An Ozark
Aquifer Extent map indicates that the "usable" part of the aquifer (defined as having less
than 10,000 mg/L of TDS per Macfarlane; no definition of "usable" is provided by the
map) covers the three southeastern-most counties (Bourbon, Crawford, and Cherokee) of
the state (Figure A8-7) and parts of the adjacent four counties (Linn, Allen, Neosho, and
Labette) (DASC Ozark Aquifer Extent Map, 200Ic). Because the land surface elevation
in that portion of the state is roughly 850 feet above sea level (DASC Kansas Elevation
Map, 200 Ib) and the elevation of the base of the Ozark Aquifer is roughly 900 feet below
sea level (Ozark Aquifer Base Map, 200 Ic), the base of the Ozark aquifer is roughly
1,750 feet below ground surface. Groundwater samples taken from lower Paleozoic
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-5
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
aquifers in Kansas show TDS levels ranging from <500 to 5,000 mg/L (Figure A8-10)
(Macfarlane and Hathaway, 1987), well within the range for a USDW.
Table A8-2 contains information concerning the relative location of potential USDWs
and methane-bearing coalbeds in the Cherokee Basin. The table shows that all or part of
the targeted coal seams could be coincident with a USDW, allowing the potential for
impacts. Most past coalbed methane production activity within the Cherokee Basin took
place in Kansas (Quarterly Review, 1993). However, coalbed methane production
activity within the Cherokee Basin in Oklahoma has increased markedly in recent years
(Hudson, 2001).
Table A8-2 Relative Locations of USDWs and Potential Methane-Bearing
Coalbeds, Cherokee Basin
Coal Group
Mulky
Weir-
Pittsburg
Riverton
Kansas
Depth to top
of Coal1
(ft)
600 to 1000
800 to 1200
800 to 1200
Depth to
base of
Fresh Water
(USDW) 2
(ft)
-1750
Missouri
Depth to
top of
Coal1
(ft)
600 to
1000
800 to
1200
800 to
1200
Depth to
base of
Fresh
Water 3
(ft)
N/A4
Oklahoma
Depth to top
of Coal l
(ft)
600 to 1000
800 to 1200
800 to 1200
Depth to
base of
Fresh
Water
(ft)
N/A4
1 Quarterly Review, 1993
2 Ozark Aquifer extent and base, and Kansas elevation maps from the Kansas Data Access and Support
Center (DASC) 200 Ib
above
3 Missouri's Geological Survey, Water Resources Branch, claims no water supplies in these
strata
4 Not Available
8.2.3 Forest City Basin USDW Identification
The Forest City Basin includes parts of the States of Iowa, Kansas, Missouri, and
Nebraska. In Iowa, the Forest City Basin lies within the Southern Iowa Drift Plain
physiographic province (Figure A8-11). The most productive aquifer in this area is the
dolomite and sandstone Jordan Aquifer (Figure A8-11). Wells in this aquifer commonly
range in depth from 300 to 2,000 feet (some are as deep as 3,000 feet) and usually
produce 100 to 1,000 gallons of water per minute (National Water Summary, 1984). This
aquifer usually contains in excess of 1,500 mg/L TDS in the southern portion of the state
(National Water Summary, 1984). Other aquifers used at various locations in the basin
are found in the Silurian-Devonian age and in the Mississippian-age strata (Figure A8-
11). Water wells in these aquifers range from 150 to 750 feet deep with variable
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-6
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
production (Howes, 2002). Also contained within this basin in Iowa is a portion of the
confined, poorly cemented sandstone Dakota aquifer (National Water Summary, 1984)
(Figure A8-11). Water wells in this aquifer are typically 100 to 600 feet in depth, and
commonly produce 100 to 250 gallons of water per minute (National Water Summary,
1984). An Iowa Division of Natural Resources Geological Survey Bureau geologist,
Mary Howes, said that few towns in Iowa use Pennsylvanian strata for water, as they
typically contain high concentrations of sulfate and TDSs (Howes, 2001). Most
community water supplies in the southern portion of Iowa use surface water and shallow
alluvial aquifers as drinking water sources, and there are a few wells in fractured bedrock.
Private water supplies typically are derived from seepage wells, shallow bedrock wells,
or purchased from a public supply (Howes, 2002).
In Kansas, the basin is located in the Lowlands physiographic province (Figure A8-7),
and only the northeastern corner of the state falls within the Forest City Basin boundary.
In addition to the Douglas Aquifers described above in the Cherokee Basin Aquifer
descriptions, this portion of the Forest City Basin in Kansas also contains a glacial drift
aquifer (is this and some alluvial aquifers adjacent to the Kansas River (National Water
Summary, 1984) (Figure A8-7). In the glacial drift, wells are typically 10 to 300 feet in
depth and usually produce 10 to 100 gallons of water per minute (National Water
Summary, 1984). Wells in the alluvium are usually 10 to 150 feet deep and typically
produce 10 to 500 gallons of water per minute (National Water Summary, 1984). The
glacial drift aquifer's base varies from about 850 to 1,300 feet above sea level (DASC,
Glacial Drift Base Map, 200 la). Since the elevation of the land surface in this portion of
Kansas is roughly between 1,000 and 1,400 feet above sea level (DASC, Kansas
Elevation Map, 200Ib), the aquifer appears to extend only to an approximate maximum
depth of 150 feet below the ground surface.
In Missouri, the basin lies within the Central Lowland physiographic province (Figure
A8-8). The principal aquifer in this area is a glacial-drift aquifer (Figure A8-8). In this
aquifer, water wells are typically 100 to 250 feet in depth and produce 5 to 200 gallons of
water per minute. In addition to this aquifer, alluvial deposits along the Missouri River
are also developed for water (National Water Summary, 1984)(Figure A8-8). Well
depths in the alluvium usually range from 80 to 100 feet in depth, and the wells typically
produce 100 to 1,000 gallons of water per minute (National Water Summary, 1984).
Two public supply wells in Cass County, Missouri, extract water from Pennsylvanian
strata for the town of East Lynn. A map of groundwater quality within Paleozoic aquifers
of Missouri (Figure A8-12) shows that within the Forest City Basin, water quality ranges
from about 500 mg/L TDS to 40,000 mg/L TDS in deeper portions of the basin (Missouri
Division of Geological Survey and Water Resources, 1967). A 10,000 mg/L TDS
boundary line delineated in the Mississippian aquifers of Missouri (located directly below
Pennsylvanian-age strata) includes portions of Cass, Jackson, Lafayette, Carroll, Saline,
Ray, Clay, Caldwell, Clinton, and Platte Counties (Netzler, 1982) (Figure A8-8).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-7
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
Only the southeastern tip of Nebraska (primarily Richardson County) falls within the
limits of the Forest City Basin. The principal aquifers in this area are undifferentiated
aquifers in Paleozoic-age rocks (National Water Summary, 1984) (Figure A8-13).
Locally overlain by saturated Quaternary-age sand and gravel deposits, wells within this
area are commonly 30 to 2,200 feet in depth, and produce about 10 to 200 gallons of
water per minute. TDS levels in the water can be as high as 6,000 mg/L, but are usually
less than 1,500 mg/L (National Water Summary, 1984). The Ground Water Atlas of
Nebraska (Flowerday et al., 1998) indicates that Richardson County is within the
Southeastern Nebraska Glacial Drift rock unit. The thickness of the aquifer in
Richardson County is less than 100 feet and the depth to water is 30 to 200 feet. The
information in the Ground Water Atlas of Nebraska (Flowerday et al., 1998) appears to
be in conflict with the data presented by the U.S. Geological Survey in the National
Water Summary (1984). Matt Jokel of the Nebraska Conservation and Survey Division
said it is very difficult to obtain water in this portion of the state, and most people use
valley fill materials and paleochannels as water supply sources. He also believes that the
coal resources, which could possibly be used for methane extraction, are probably too
deep to be located coincident with the shallow water supplies in the area (Jokel, 2001).
Table A8-3 contains information concerning the relative location of potential USDWs
and potential methane-bearing coalbeds in the Forest City Basin.
Table A8-3 Relative Locations of USDWs and Potential Methane-Bearing
Coalbeds, Forest City Basin
Coal Group
Cherokee
Group
Iowa
Depth
to top
of
Coal1
(ft)
Oto
>230
Depth
to base
of
fresh
water3
(ft)
N/A8
Kansas
Depth
to top
of
Coal1
(ft)
720 to
1220
Depth
to base
of
fresh
water 4
(ft)
-150
Missouri
Depth
to top
of
Coal1
(ft)
300 to
1100
Depth
to base
of
fresh
water 5
(ft)
N/A8
Nebraska
Depth
to top
of
Coal1'6
(ft)
1220
to
1396
Depth to
base of
fresh
water 2>?
(ft)
129 to
299
1 Bostic et al., 1993
2 Note: The base of "fresh water" is not the base of the USDW. Fresh water is within the USDW and the
base of fresh water is above the base of the USDW.
3 Howes, Iowa Geological Survey Bureau (2001) believes water quality data may be available to define this
depth
4 Glacial Drift base and Kansas elevation maps from the Kansas Data Access and Support Center (DASC),
200 Ib
5 Maps (Netzler, 1982) sent by Missouri show the extent of aquifers containing less than 10,000 mg/L of
TDS, but not depths
6CondraandReed, 1959
7 The Groundwater Atlas of Nebraska, (Flowerday et al., 1998)
8 Not Available
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Presently, there does not appear to be a USDW located at the same depth as the coals of
the Cherokee Group in the Forest City Basin. However, very little is known about the
coal resources of this basin (Quarterly Review, 1993). Further research is required to
delineate the possible link between coalbed methane resources and USDWs in the Forest
City Basin.
8.3 Coalbed Methane Production Activity
GTI places total coalbed gas production in the Western Interior Coal region at 6.5 Bcf for
the year 2000 (GTI, 2002).
8.3.1 Arkoma Basin Production Activity
In 1989, Bear Production Company became the first company to target coalbed methane
production from the Hartshorne Coals of the Arkoma Basin in Haskell County,
Oklahoma (Quarterly Review, 1993). As of 1993, Bear Production had 38 wells in
operation, Aztec Energy Corporation had 19 wells, and Redwine Resources, Inc. had 40
wells in the Arkoma Basin (Quarterly Review, 1993).
As of 1993, Bear Production was not fracturing its wells, but rather completing them as
open holes without perforated casings (Quarterly Review, 1993). However, other
production companies were fracturing their wells for methane production. Before 1992,
water, linear gel, acid, and nitrogen foam fracturing fluids were used, with most operators
using foam with small sand volumes (35,000 to 60,000 Ibs) (Quarterly Review, 1993). In
1993, slick water fracturing fluids containing no proppant were becoming more common
(Quarterly Review, 1993). Well fracturing data from 36 wells in the Spiro Southeast
Field of LeFlore County, Oklahoma show that either water or nitrogen foam was the base
fracturing fluid used to carry sand proppant into coal cleats (Andrews et al., 1998).
Fracturing continues in the Arkoma Basin today, at least in Oklahoma, where undisclosed
amounts of initial water production are "firac" waters introduced during fracture
stimulation (Cardott, 2001). Both Wendell (2001) and Marshall (2001) outline current
hydraulic fracturing practices within the Arkoma Basin, and Wendell (2001) includes
acids, benzene, xylene, toluene, gasoline, diesel, solvents, bleach, and surfactants as
detrimental hydraulic fracturing substances in his "lessons learned" category.
A search of the Oklahoma Coal Database, updated on January 17, 2001, indicated that
over 360 coalbed methane wells had been completed in Haskell, Le Flore, and Pittsburg
counties alone, targeting the Hartshorne, McAlester, and Savanna coals. Additional
operators in the Arkoma Basin today include Continental Resources, SJM Inc., Brower O
& G, Mannix Oil, and OGP Operating (Oklahoma Coal Database, 2001).
Apparently there is little to no coalbed methane activity in the Arkoma Basin in
Arkansas, based on the Arkansas Geological Commission's Web site, which states,
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-9
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
".. .there exists the potential for coalbed methane production in this area of the state"
(Arkansas Geological Commission, 2001). The low coalbed methane activity in this
Basin is further confirmed by Andrews et al. (1998), which outlines Arkansas' restrictive
field-spacing policy from the 1930s of only one well per 640-acre section for each
producing zone in the Hartshorne. This policy effectively made exploration
uneconomical. A change in field-spacing rules in 1995 has stimulated new interest
among independent producers in Arkansas to develop methane from the Hartshorne coals
(Andrews et al., 1998).
8.3.2 Cherokee Basin Production Activity
In the Cherokee Basin, unknown amounts of coalbed methane gas have been produced
with conventional natural gas for over 50 years (Quarterly Review, 1993). Targeted
coalbed methane production increased in the late 1980s, and at least 232 coalbed methane
wells had been completed as of January 1993 (Quarterly Review, 1993). During this
timeframe, development was centered on Montgomery County, Kansas, with the most
active operators being Great Eastern Energy and Development Corporation with 81 wells,
Kan Map Inc. with 47 wells, and Stroud Oil Properties Inc. with 35 wells, (Quarterly
Review, 1993). In addition to these operators, Bonanza Energy Corporation, Conquest
Oil Company, Foster Oil & Gas, Hunter, Quantum Energy, Uranus, and U.S. Exploration
had active development programs, and Derrick Industries was planning a program
(Quarterly Review, 1993).
The coalbed methane wells were typically fractured with water or nitrogen-based fluids
and sand, although the shallower Mulky coal received fracturing treatments of 40-pound
linear gel and sand (Quarterly Review, 1993). On average, 5,000 pounds of sand were
used per foot of coal (Quarterly Review, 1993). Another technique used in Kansas
consists of injecting 4 barrels of 15 percent hydrochloric acid mixed with 16 barrels of
potassium chloride and 15,000 standard cubic feet of nitrogen (Stoeckinger, 1990). In the
Sycamore Valley field in Kansas, Stroud Oil Properties used 426 barrels of cross-linked
fluid with 52 percent pad and 3 percent flush, and 30,000 pounds of 12/20 sand mixed at
one to nine pounds per gallon injected at 20 barrels per minute. Operators were avoiding
large-volume treatments due to a fear that fractures could be induced in thick water-
bearing sands above and below the coals, which would have created excess water
production (Quarterly Review, 1993). Stoeckinger (2000) reports that current hydraulic
fracturing practices in the Cherokee Basin in Kansas are water only, no gel, with nitrogen
being popular and "slick-water down tubing."
Pam Hudson, of the Oklahoma Corporation Commission, indicated that coalbed methane
extraction was beginning to grow in the Cherokee Basin in the northeastern section of
Oklahoma, and more development was now centered on that region as opposed to the
Arkoma Basin to the south. Ms. Hudson expected that much of the development would
be focused on Washington, Nowata, and Craig Counties (Hudson, 2001).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-10
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
In Missouri, there appears to be little to no coalbed methane extraction within the
Cherokee Basin. David Smith, a geologist with the Missouri Geological Survey, stated
that coalbed methane extraction in Missouri is essentially non-existent (Smith, 2001).
8.3.3 Forest City Basin Production Activity
The Forest City Basin was relatively unexplored in 1993, with about ten coalbed wells
concentrated in Kansas' Atchison, Jefferson, Miami, Leavenworth, and Franklin Counties
(Quarterly Review, 1993). The wells were hydraulically fractured with 500 to 30,000
pounds (with an average of 5000 pounds) of sand proppant. The types of fluids used
during the fracturing process were not mentioned (Quarterly Review, 1993).
David Smith, believes that at one time there were some coalbed methane wells just south
of Kansas City in Cass County (Smith, 2001). Sherri Stoner, of the Missouri Geological
Survey, confirmed this in February 2001, and remarked that they were no longer in
operation (Stoner, 2001). An Iowa Division of Natural Resources Geological Survey
Bureau geologist, Mary Howes, stated that presently there was no coalbed methane
production in Iowa (Howes, 2001).
Information concerning coalbed methane production activity in Nebraska could not be
found.
8.4 Summary
Based on depths to the Hartshorne Coal and the base of fresh water presented in Table
A8-1, it appears that coalbed methane extraction wells in the Arkoma Basin could be
coincident with potential USDWs in Arkansas, potentially allowing for impacts. Based
on maps provided by the Oklahoma Corporation Commission (2001), which depicts the
depths to the 10,000 mg/L of TDS groundwater quality boundary in Oklahoma, the
location of coalbed methane wells and USDWs would most likely not coincide in
Oklahoma. This is based on depths to coals typically greater than 1,000 feet (Andrews et
al., 1998) and depths to the base of the USDW typically shallower than 900 feet
(Oklahoma Corporation Commission, 2001).
Table A8-2 supports the possibility that coalbed methane wells in the Cherokee Basin
targeting the Cherokee Group coals in Kansas may coincide with USDWs, indicating the
potential for impacts to drinking water. In Missouri, more water quality data is required
prior to any determination of coalbed methane well/USDW conflict. In addition, since
only a very small portion of the Cherokee Basin falls within the state, this portion of the
basin needs to be delineated more precisely to see which USDWs lay within this small
part of the basin. However, current levels of coalbed methane activity in Missouri are
minimal.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-11
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Last, in the Forest City Basin, there appears to be little physical relationship between
coalbeds that may be used for coalbed methane extraction and water supplies. However,
aquifer and well information from the National Water Summary (1984) indicate that a co-
location of the two could exist in Nebraska. More information would be needed to fully
investigate the relationship between coalbeds and USDWs in the Forest City
Basin.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-12
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
FOREST CITY
MISSOURI
CHEROKEE
ARKOMA
0 100 200 300 400 500 MILES
«- Coal Basin
Western Interior Coal Basin - Index Map of the Arkoma, Cherokee and Forest City Basins
(Quarterly Review, 1993)
Figure A8-1
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-13
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
5
UJ
V)
&
Z
<
^_
z
«€
>•
1
=••••
5*-
CO
Z
z
UJ
CL
UJ
K
LU
W3
2
-
LU
Z
2E
C/J
UJ
O
z
5
o
<
n
*LjL.
O
oc
o
O
S
Of
O
u_
1
j
CO
EC
UJ
ce
^r
£
i-
CO
UJ
1
S
" u
oe
o
X
8Q
2
if) ^ n
I Z Q
tz fe uj
_J O CD
.-.--.--:- *.-.-•.!
-:t™rr:-*7-rq
ammmmmmmml
-ij.'-lL-'^j-zLJ
r f_t
,"f.J'f:".''*T'f
•••S^"
FORMALLY
AND
Keota SandsEone Member
Tamaha Sandstone Member «
Upper McA osier coal §
. "8
Cameron Sandstone Member 3
Lequrre ^anosiane Memoer m
_, ICw^teri(7) cos)
Warner Sandstone Member
McCurtain State Member
unnamed siltstorve
•••••••••T^. A Up|»r Harlshame coal
.'.* v'-v •'•"•' *-'l ^ S ypper Hartshoinc sanislofw
Is
£ ^^
s
2
P
<
;•--*- >-.•*'•' •
:•::.:•.:•-:;;•;
rr-^.t-T^^7
• '' ;' ''•'-'/•". ,- ,- '
fe jj Lower Hartohome coal
3 s lower Hartsnome sandstofie
Gllcrease sandttonis
Red Oak sandstone
Spiro sandstone
VI.. -Mil' !.i
V,'<Mlerr Intero? Cca' Ba*,-"< - Ge^MTftd alfahg'apr c CO'JTH bf'f'e -'entyks-iiar S/srir'v
It l*if- A'*.::ft(,'l "i3;,ir iFs'M'i'tVin, "SSJ
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-14
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
On
40'
60
ao-
100-
200-
300-
400-
tt
Iron PBS* sh
iron Posl coai
1*
£
a
"0
o
l_ _
E
1
a.
o»
1
0
(J
«
«s
B
B»moi-oe
Q
IA
1 1
c >
ffl m
C C
-^ c
*»**'"*' ^
II
Q. =
i*
i
sh - hiir.
fni «•
HP « Group
Cu-ii B.isin - Slr*li!^?sphiC Ctfunm of IMS Cherokee Group
•i Ihi: ChercAftft fiaain (Tadescu, 1*32)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-15
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
U L l_U
A,'«a Shown
'/I'eitem Inleror Coal Basin - Forest City Basin Sludy Area
ing Localran of Dnll HolaE Dscwssed in Bosfc fit al. 1993)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-16
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
-' ter, .,„. _
^
i'-l "/I "v-
• ^ xx.
-/j^j*'
HP
3,,.,,
,
. and
Waior Summaiy,
Fitrufe AB-5
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-17
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
.,
EXPLANATION
n Alluvium and terrace deposits along
major streams
J High Plains aquifer
J Antlers and Rush Springs aquifers
I Dog Creek Blsine aquifer
I Garber - Wellington and Vamoo&a -Ada aquifers
J Keokuk -Reedl Spring IBoonel aquifers
I Roubidoux aquifer
n Arbuckle • Simpson and Arbuckle • Timbered Hills aquifers
J Not a principal aquifer
Boundary of aquifer uncertain
^J
GREAT PLAINS
PROVINCE
OUACHITA
PROVINCE
\
COASTAL PLAIN
Counties, Aquifers, and Physiographic Provinces of Oklahoma
(National Water Summary, 1984)
u 1027-VYI-S.Mr
Figure A8-6
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-18
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
100 MILES
GREAT PLAINS PROVINCE
DISSECTED TILL PLAINS
SECTION OF CENTRAL
LOWLANDS PROVINCE
OS AGE PLAINS
SECTION OF CENTRAL
LOWLANDS PROVINCE
02ARK PLATEAUS
PROVINCE
EXPLANATION
J Alluvial aquifers
J Glacial drift aquifers
^| High Plaint aquifer
J Great Plains aquifer
~| Chase and Council Grove aquifer;
~| Doug I OS aquifer
Ozark aquifer
Not a principal aquifer
Counties, Aquifers, and Physiographic Provinces of Kansas
(National Water Summary, 1984)
612*02 Me 1027-WI-T ojr
Figure A8-7
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-19
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
CENTRAL LOWLAND
Dt«tct«d Till Plains
3 Os«g* Pltini
s
-I
, ft *
PI
1 .«., •* •
•
t _":-^SSfi_. ..^
EXPLANATION
PRINCIPAL AQUIFERS
| | Major river valleys
~] Alluvial
| Wllcox and Claibome
B McNairv
^] Kimmiwick-Potosi
OTHER AQUIFERS
"J Glacial-drllt. Pennsvlvanlan
Mississlppian age. Springlield
Plateau, and St. Francois
"] NOT A PRINCIPAL AQUIFER
A—A'Tr»c« of cross section
Dissolved-solids concentration
greater than 1000 milligrams
per liter (approximate location)
-4OOO'
Counties, Aquifers, and Physiographic Provinces of Missouri
(National Water Summary, 1984)
Figure A8-8
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-20
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
Sioux ridge
» ••'
':.::
Iowa
v\
Wisconsin
dome
3re-Pennsy(vanian production .»:
Nemaha uplift
Nebraska
McClain oil field
Wehking gas field
Pennsylvanian production
Ozark uplift
7
Cherokee Basin
Basin Boundary
Western Interior - Detail of Forest City Basin with Detail of
Cherokee Basin in Missouri (Tedesco, 1992)
S/29/02 kit 1027-WI-9.cdr
Figure A8-9
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-21
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
MONTGOMERY
***
ALLEN
NEOSHO
BOURBON l^auau
CRAWFORD
I BARTON
' '"*1 " ' >c 1
' *••!- X
' -1^ " r* '
CHEROKEE
OKLAHOMA
NQWATA
TCRAJG nr
s^
•^ «•
JASPER
NEWTON
Top of Mississippian Outcrop
<500
500-1000
1000-5000
>5000
25 mi
25 km
Water Quality (TDS) of Lower Paleozoic Aquifers in Kansas
(Macfarlane and Hathaway, 1987}
Figure A8-10
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-22
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
EXPLANATION
] Surliclal aquifers
] Dakota aquifer
J Mississippian aquifer
J Silurian — Devonian aquifer
J Jordan aquifer
^| Not a principal aquifer
A—A'Trace of cross section
NORTHWEST IOWA
PLAINS
SOUTHERN iOWfl
DRIFT PLAIN
Counties, Aquifers, and Physiographic Provinces of Iowa
(National Water Summary, 1984}
5129102 Ml 1027-Wl.tl.slr
Figure A8-11
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-23
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
EXPLANATION
Western Interior Coal Basin - Quality of Ground water in the Palezoic Aquifers of Missouri
(Missouri Division of Geological Survey & Water Resources, 1967)
Figure A8-12
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-24
image:
EPA816-R-04-003
Attachment 8
The Western Interior Basin
N
.- " '••"',
'
- .=; L
•" .-
i»il» "=J*- '"'-"•"" jjS
s
I ""**
I
H iBtMiLii
LntCOrOULiaATtl] ALLUVIAL Mhll
BfflJTMEM I
]
MCKMUWtM
H^ff Him xiNilii ik
I HKAIbii MHI|BP
AMD EJtflBDMATI
]
j MlC I MilXlUI «iK. I.I
EU W lilDM
.•• MfiFNI IVSTEM, ,n hn
.
'
I I
Couniies. Aquifers, and PhyEiograirfik: Provinces of Nebraska
^National Water Stirmwy, 19641
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A8-25
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
REFERENCES
Andrews, Richard D., Cardott, Brian J., and Storm, Taylor. 1998. The Hartshorne Play
in Southeastern Oklahoma: regional and detailed sandstone reservoir analysis and
coalbed-methane resources. Oklahoma Geological Survey, Special Publication
98-7.
Arkansas Geological Commission Web site. 2001. http://www.state.ar.us/agc.htm.
Bostic, Joy L., Brady, L. L., Howes, M. R., Burchett, R. R., and Pierce, B. S. 1993.
Investigation of the coal properties and the potential for coal-bed methane in the
Forest City Basin. U. S. Geological Survey, Open File Report 93-576.
Brady, L. L. 2002. Kansas Geological Survey. Personal communication.
Burchett, Raymond R. No date specified. Coalbed methane potential in the Nebraska
portion of the Forest City Basin. Institute of Agriculture and Natural Resources,
University of Nebraska-Lincoln.
Cardott, Brian J. 2001. Coalbed-Methane Activity in Oklahoma, 2001. Oklahoma
Coalbed-Methane Workshop 2001: Oklahoma Geological Survey, Open File
Report 2-2001, p. 93-118.
Charpentier, Ronald R. 1995. Cherokee Platform Province. U. S. Geological Survey,
National Assessment of United States Oil and Gas Resources.
Condra, G. E. and Reed, E. C. 1959. The geological section of Nebraska. Nebraska
Geological Survey Bulletin 14A, 1959.
Cordova, Robert M. 1963. Reconnaissance of the ground-water resources of the
Arkansas Valley Region, Arkansas. Contributions to the Hydrology of the United
States, Geological Survey Water-Supply Paper 1669-BB.
DASC website. 200la. Glacial drift base map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html.
DASC website. 2001b. Kansas elevation map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html.
DASC website. 2001c. Ozark Aquifer base map.
http://gisdasc.kgs.ukans.edu/dasc/kanview.html.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-26
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Flowerday, C. F., Kuzelka, R. D., and Pederson, D. T., compilers. 1998. The Ground
Water Atlas of Nebraska.
Friedman, Samuel A. 1982. Determination of reserves of methane from coalbeds for use
in rural communities in eastern Oklahoma. Oklahoma Geological Survey, Special
Publication 82-3, 1982.
Gas Technology Institute (GTI) website. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org
Hansen, D. 2001. Arkansas Geological Commission. Personal communication.
Howes, M. R. 2001. Iowa Geological Survey Bureau. Personal communication.
Howes, M. R. 2002. Iowa Geological Survey Bureau. Personal communication.
Hudson, P. 2001. Oklahoma Corporation Commission. Personal communication.
lannacchione, A.T., and D.G. Puglio. 1979. Methane content and geology of the
Hartshorne coalbed in Haskell and Le Flore Counties, Oklahoma: U.S. Bureau of
Mines Report of Investigations 8407, 14 p.
Jokel, M. 2001. Nebraska Conservation and Survey Division. Personal communication.
Luza, K. 2001. Oklahoma Geological Survey. Personal communication.
Macfarlane, A. 2001. Kansas Geological Survey. Personal communication.
Macfarlane, P. A. and Hathaway, L. R. 1987. The Hydrologic and Chemical Quality of
Ground Waters from the Lower Paleozoic Aquifers in the Tri-State Region of
Kansas, Missouri, and Oklahoma: Kansas Geological Survey Groundwater Series
9.
Marcher, M. V. 1969. Reconnaissance of the Water Resources of the Fort Smith
Quadrangle, East-Central Oklahoma: Oklahoma Geological Survey Hydrologic
Atlas 1.
Marcher, M. V. and Bingham. 1971. Reconnaissance of the Water Resources of the
Tulsa Quadrangle, Northeastern Oklahoma: Oklahoma Geological Survey
Hydrologic Atlas 2.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-27
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Marshall, R. 2001. Midcontinent Evolving Coalbed-Methane Completion Techniques
and Practices. Oklahoma Coalbed-Methane Workshop 2001: Oklahoma
Geological Survey, Open File Report 2-2001, p. 140-150.
Missouri Division of Geological Survey and Water Resources. 1967. Mineral & Water
Resources of Missouri, 43(2).
Nelson, Charles R. 1999. Changing perceptions regarding the size and production
potential of coalbed methane resources. Gas Research Institute, June 1999.
Netzler, Bruce W. 1982. Maps of total dissolved solids concentrations in ground water
from the Mississippian aquifers, the Jefferson City, Cotter and Powell Dolomites,
and the Roubidoux Formation in Missouri.
National Water Summary. 1984. Hydrologic events, selected water-quality trends, and
ground-water resources. United States Geological Survey Water-Supply Paper
No. 2275.
Oklahoma Coal Database, January 17, 2001.
Oklahoma Corporation Commission, Depth to Base of Treatable Water Map Series,
2001.
Oklahoma Geological Survey website. 2001. http://www.ou.edu/special/ogs-pttc.
Petroleum Technology Transfer Council website. 1999. http://www.pttc.org.
Potts, Ronald. 1987. Water Quality and Quantity in Abandoned Underground Coal
Mines of West-Central Arkansas and Use of Surface Electrical Resistivity in
Attempting Quality Determinations. Arkansas Geological Commission -
Information Circular 20-N.
Prior, W. 2001. Arkansas Geological Commission. Personal communication.
Quarterly Review. 1993. Coalbed Methane - State of the Industry. Methane From Coal
Seams Technology, August 1993.
Stoeckinger, William T. 1990. Kansas coalbed methane comes on stream. Oil & Gas
Journal, June 4, 1990.
Stoeckinger, William T. 2000. Coalbed Methane Completion Practices on the Cherokee
Platform. Oklahoma Coalbed-Methane Workshop: Oklahoma Geological Survey,
Open-File Report OF 2-2000, pp. 36-51.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-28
image:
EPA 816-R-04-003 Attachment 8
The Western Interior Basin
Smith, D. 2001. Missouri Geological Survey. Personal communication.
Smith, D. 2002. Missouri Geological Survey. Personal communication.
Stoner, S. 2001. Missouri Geological Survey. Personal communication.
Tedesco, Steven A. 1992. Coalbed methane potential assessed in Forest City Basin. Oil
& Gas Journal, Exploration^ February 10, 1992.
Vandike, J. 2001. Missouri Geological Survey. Personal communication.
Wendell, John H. JR. 2001. Arkoma Basin Coalbed-Methane Potential and Practices.
Oklahoma Coalbed-Methane Workshop 2001: Oklahoma Geological Survey,
Open File Report 2-2001, pp. 119-139.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A8-29
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
Attachment 9
The Raton Basin
The Raton Basin covers an area of about 2,200 square miles in southeastern Colorado and
northeastern New Mexico (Figure A9-1). It is the southernmost of several major coal-
bearing basins along the eastern margin of the Rocky Mountains. The basin extends 80
miles north to south and as much as 50 miles east and west (Stevens et al., 1992). It is an
elongate asymmetric syncline, with 20,000 to 25,000 feet of sedimentary rock in the
deepest part. Coalbed methane resources in the basin, which have been estimated at
approximately 10.2 trillion cubic feet (Tcf), are contained in the upper Cretaceous
Vermejo Formation and upper Cretaceous and Paleocene Raton Formation (Stevens et
al., 1992). In 2000, the average gas production rate per well in the Raton Basin was close
to 300,000 cubic feet per day, and annual production was 30.8 billion cubic feet (Bcf)
(GTI, 2002).
9.1 Basin Geology
The Raton structural basin is an asymmetric synclinal sedimentary basin containing
sedimentary rocks as old as Devonian overlying basement Precambrian rocks, with
Holocene sediments at the surface. The coal occurs in the Vermejo and the Raton
Formations, which overlie the Trinidad Sandstone, a basin-wide regressive marine
sandstone (Figure A9-2). The Vermejo and Raton Formations consist of deltaic lower
coastal plain and fluvial deposits (Flores and Pillmore, 1987). Numerous discontinuous
and thin coalbeds are located in the Vermejo Formation and the Raton Formation, which
overlie the Trinidad Sandstone (Figure A9-3). The top of the Trinidad Sandstone forms
the lower boundary of the Raton coal basin as shown in Figure A9-1. Development of
coalbed methane wells has focused on development of the Vermejo coals rather than the
Raton coals because the former are thicker and more abundant. The coalbeds are of
limited extent and cannot be correlated over more than a few miles.
Individual coalbeds in the Vermejo Formation range from a few inches to about 14 feet
thick, and total coal thickness typically ranges from 5 to 35 feet. An isopach map of total
coal thickness in the Vermejo Formation, based on 92 well logs and measured sections,
was published by Stevens et al. (1992) (Figure A9-4). Total coal thickness in the Raton
Formation ranges from 10 feet to greater than 140 feet, with individual seams ranging
from several inches to greater than 10 feet thick. Although the Raton Formation is much
thicker and contains more total coal than the Vermejo Formation, individual coal seams
in the Raton are less continuous and generally thinner. Additionally, because of
extensive erosion of the Raton Formation, particularly in the eastern part of the basin,
much of the original coal is no longer present (Stevens et al., 1992). Between 5 and 15
individual coalbeds produce coalbed methane for wells in the basin (Hemborg, 1996).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-1
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
Middle Tertiary igneous intrusions are present in the central part of the basin (Steven,
1975). Sills and dikes have invaded sediments of the basin including both the Vermejo
and Raton Formations. Sills have intruded along the coal seams destroying tremendous
quantities of coal (Carter, 1956).
Coal seam depth is an important variable used to estimate gas production potential.
Figure A9-5 is a thickness of overburden map from Stevens et al. (1992). The map
shows the depth below land surface to the midpoint depth of the coal-bearing interval,
using coal thickness as a weighting factor. Overburden thickness ranges from less than
500 feet near the basin perimeter to greater than 4,100 feet in the deep northwestern part
of the basin. Many of the differences in thickness of overburden can be attributed to
variations in topography and are thus a consequence of erosion and not necessarily
subsurface geologic structure.
Stratigraphic cross-sections constructed to illustrate the regional subsurface geologic
structure and the distribution of coal seams and igneous intrusions, as well as the areal
locations of these cross-sections, are shown in Figures A9-6 through A9-8. The cross-
sections use the top of the Trinidad Sandstone as the horizontal datum. The Vermejo
Formation has a relatively uniform thickness of about 350 feet throughout the basin. The
Raton Formation varies from about 0 to 2,100 feet thick. It grades westward into and is
overlain by the conglomeratic Poison Canyon Formation (Flores, 1987; Flores and
Pillmore, 1987).
A study of the relationship between coal cleat orientation and the compression stresses
due to tectonic forces can indicate areas likely to have increased coal seam permeability
and provide increased coalbed methane yield (Stevens et al., 1992). Cleats, or small-
scale fractures in the coal, are commonly oriented perpendicularly to the maximum stress.
These fractures tend to expand, thereby providing greater permeability and coalbed
methane yields on the axes of the anticlines, such as the Vermejo Park anticline. Wells
drilled near the axis of the La Veta syncline, in contrast, did not encounter adequate
permeability (Stevens et al., 1992). Initially it was thought that sills that intrude along the
bedding plane of the coal seams would reduce methane production, but several operators
have noted that elevated methane contents have sometimes been measured in coal seams
that have been intruded by igneous rocks (Stevens et al., 1992).
9.2 Basin Hydrology and USDW Identification
Regional groundwater flow in the Raton Basin is dependent on geologic structure and
topography. Regional flow is generally down-slope from west to east or southeast
(Figure A9-9). In the northern part of the basin, however, flow is radial away from
Spanish Peaks (Howard, 1982; Geldon, 1990). Additionally, along the eastern margin of
the basin, sediments dip to the west and groundwater flow is locally down-dip to the
west. While recharge occurs primarily at elevations greater than 7,500 feet, discharge is
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-2
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
mainly through streams and by evapotranspiration in the central and eastern parts of the
basin.
Principle bedrock aquifers in the basin are the Cuchara-Poison Canyon, the Raton-
Verm ejo-Trinidad, the Fort Hayes-Codell, the Dakota-Purgatoire, and the Entrada
(Geldon, 1990) (Figure A9-3). The pressure regime in the basin is poorly understood.
Under-pressured conditions, or hydraulic heads in deep bedrock aquifers that are lower
than those in shallow formations, appear to exist throughout much of the basin (Howard,
1982; Geldon, 1990; Tyler et al., 1995). This hydraulic head difference suggests that the
deep bedrock aquifers are not in communication with shallow formations. Meteoric
circulation, however, is indicated by the regional freshness of the produced waters
(Stevens et al., 1992; Tyler et al., 1995).
All of the water produced along with coalbed methane in the Raton Basin has a total
dissolved solids (TDS) content of less than 10,000 milligrams per liter (mg/L) (the water
quality criterion for an underground source of drinking water (USDW)), and the aquifers
from which the gas is produced meet the water quality criterion for a USDW (National
Water Summary, 1984). A scatter diagram of potentiometric head versus TDS from
coalbed methane wells in the Raton Basin (Figure A9-10) shows little correlation
between potentiometric head and water quality. More importantly, this figure shows that
all of the water had less than 10,000 mg/L of TDS, nearly all had a TDS of less than
2,500 mg/L, and more than half had a TDS of less than 1,000 mg/L. Two producers used
injection wells for disposal, but operating permits issued to one gas producer (Evergreen
Resources, Inc.) by the Colorado Department of Public Health and Environment allowed
discharge of produced water into streambeds and stock ponds, indicating that the water
was not too saline for surface discharge. Hemborg (1998) suggests that the wells
yielding larger quantities of groundwater might be connected to the underlying water-
bearing Trinidad Sandstone.
9.3 Coalbed Methane Production Activity
Hydraulic fracturing employed for enhancement of coalbed methane production is
designed to enable gas within the rock to flow more readily to an extraction well.
Coalbed methane well stimulation using hydraulic fracturing techniques is a common
practice in the Raton Basin. Records show that fluids used are typically gels and water
with sand proppants.
Hemborg (1996) reported that the average water production from coalbed methane wells
in the Raton Basin was 700 barrels per million cubic feet (Mcf), and average daily
production for 42 wells in the Spanish Peak Field was 0.309 Mcf (Hemborg, 1998).
Conversion of these rates from coalbed methane industry units to those commonly used
for water supplies gives an average water production rate for those wells of only 6.3
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-3
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
gallons per minute. These rates are generally not considered sufficient for public water
supply or irrigation; however, they meet the water supply volume criterion for a USDW.
Hemborg (1998) showed that in most cases water yield decreased dramatically as coalbed
methane production continued over time (Figure A9-11). However, some wells exhibited
increased water production as coalbed methane production continued or increased over
time (Figure A9-12). Two causal factors were suggested (Hemborg, 1998) for the rise in
water production:
1. Well stimulation had increased the well's zone of capture to include adjacent
water-bearing sills or sandstones that were hydraulically connected to
recharge areas; or
2. Well stimulation had created a connection between the coal seams and the
underlying water-bearing Trinidad Sandstone.
The Trinidad Sandstone is a bedrock aquifer confined by the Pierre Shale below and the
shales and siltstones of the Vermejo Formation above (Figure A9-2). The Trinidad
Sandstone exhibits low vertical and horizontal permeabilities of 0.186 and 0.109 meters
per day, respectively, as reported by Howard (1982) in Stevens et al. (1992). One gas
company reported that lower water production and improved gas production were
achieved by avoiding known water-bearing horizons and by selectively completing the
coal zones (Quarterly Review, 1993).
In-place coalbed methane resources in the Vermejo and Raton Formations were estimated
by Stevens (1992) to be between 8.4 and 12.1 Tcf with a mean estimate of 10.2 Tcf. As
of 1992, 114 coalbed methane exploration wells had been drilled in the basin (Quarterly
Review, 1993). Soon after the Picketwire Lateral was constructed to convey gas from the
fields to Trinidad and then to markets, gas well development in the basin increased
significantly. The Purgatoire River Valley (Figure A9-1), which had been identified as
having the highest coalbed methane potential in the basin, up to 8 Bcf per square mile
(Stevens et al., 1992), became the focus of development. The Purgatoire Valley area was
considered favorable for development because total coal thickness ranges from 5 to over
15 feet, drilling depths are shallow and coalbed methane content is high. The New
Mexico portion of the basin was estimated to have methane resources ranging from 4 Bcf
per square mile in the southern and eastern margins of the basin to more than 8 Bcf per
square mile in the area south of the Vermejo Park anticline. Coal seams in the Vermejo
Park area (Figure A9-1) are relatively thick, but shallow and of low rank, making
estimates of coalbed methane content relatively low (Stevens et al., 1992).
The Spanish Peak Field, in the Purgatoire River development area in Las Animas County,
Colorado (Figure A9-1), had 53 active wells in December 1996. Plans had been
announced by Evergreen Resources, Inc. to drill and complete an additional 40 wells in
1997 (Hemborg, 1998). In 1996, the Purgatoire development area was projected to be
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-4
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
capable of producing 122-137 Mcf per day in 3 to 4 years (Figure A9-1) (Hemborg,
1996). Total coalbed methane production within the Raton Basin was 30.8 Bcf per year
in 2000 (GTI, 2002).
Methane production wells have generally been completed with 5.5-inch (outer diameter)
casing with two to eight perforations per foot through the casing at the depths of the coal
seams. The coal seams are stimulated with hydraulic fracturing treatments of sand and
gelled-water, but detailed information on the nature, volumes, and use of hydraulic
fracturing fluids in gas well development in this basin are not readily available. Water
and gels with 10/40-mesh sand proppant seem to be the fluids of choice for fracturing
practices in the Raton Basin. Stevens et al. (1992) report that multiple zones in one well
are typically developed with 200,000 pounds of 10/20 or 20/40-mesh sand with 100,000
gallons of cross-linked gel per well. In one series of tests, wells were hydraulically
fractured with 283,000 to 532,000 pounds of 12/20 and 20/40-mesh sand as proppant and
110,000 to 769,000 barrels of water or gel. The wells were fractured in two stages, one
for a 25-foot thick upper zone and another for a 75-foot thick lower zone (Quarterly
Review, 1993). Relatively high rates of water flow in these wells may be the result of
fractures penetrating sandstones as well as coal seams. Another set of tests led a different
methane producer to conclude that high water production was the consequence of
induced fractures that intercept water-bearing sandstone and intrusive rocks. While
operators initially assumed that large hydraulic fracture stimulations were necessary to
link the thin and widely-spaced coal seams, it was found that such fracturing increased
unwanted water production from associated sandstones, sills and water-bearing faults
(Quarterly Review, 1993).
9.4 Summary
There are two major coal formations in the Raton Basin, the Vermejo Formation and the
Raton Formation. The Vermejo coals range in thickness from 5 to 35 feet while the
Raton coal layers range from 10 to over 140 feet thick. The coal seams of the Vermejo
and Raton Formations, developed for methane production, also contain water that meets
the water quality criteria for a USDW; therefore, it can be assumed that the Raton Basin
coals are located within a USDW. The Cuchara-Poison Canyon, Fort Hayes-Codell,
Dakota-Purgatoire, Entrada and Trinidad Sandstone and other sandstone beds within the
Vermejo and Raton Formations, as well as intrusive dikes and sills, also contain water of
sufficient quality to meet the USDW water quality criteria. Hydraulic fracturing may
create connections to the Trinidad Sandstone, as shown by increases in water withdrawal
from production wells over time. On the other hand, hydraulic connections to other
adjacent water-bearing formations may also account for the increase in water production.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-5
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
Thrust ^a«^
Anitri'sni
I**- SynsHst* —|— Anitri'sni
J - -$('u*lur>i GIISSI l*cllan
\ A»:^*n*,«B RMI^SI*^*** islpiB^pgd^**, **£ Mffla1
Structure Contour Map On Top Trinidad Sandstone (Stevens et aL 1992)
I Fgi*»A9-i
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A9-6
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
& 35 *>
111
o p. cl
sia
ti- o
° ^
l|
3. en
OQ g
O O
^+J ft
C/l
1
K>
O
O
Structure! Cross-Section A - A1
I 1
!7t%ifci
Structural Crass Secion (Stevens el a-.,
oo
K^
ON
s
O
Cd
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
<
ee
UJ
CENOZOiC
( MESOZOtC !
?$%&
Recent
Miocene
ONgoccnc
Eocene
Palsocene
llppe*
Cretaceous
Lowisr \ — .
Crwt«*0U5
Junssie
Trii&ale
PALEOZOIC
FORMATION
P0¥^S MO"^ ^O-'^Of'-On
F»n^if.ia ^i^mtth^rs
<My€>r1a.<"to FormK'Or
Cucnara Forma1iQf>
POilOF CAPyOn
Fe'if-aiion
y,> -^SHjg» t oriF^ilir "i
"r *iiiiia Eai'JUi. lit
Ptsr^e Shilo
-.P
A f ' ^(".•m'.j H<» MJ-I
>cl!ic'tiin¥C'5_ neslcne
i Cl-flCl) Si'ifls.!f.t*
~ "• i,in !" ytMi*
^ ^J Qlfe,* f ' y !• U !• «9 Ik i .s
Gune'cs Snail-
Qafcc-a 3ar ds!y«-e
^•^'ly&lis 'r ^K^rnal-r^"
Mp-*i ^r^n 1 ^r^~ifi*icn
M^'^lnr S.-^C'k »orma*A^
Ec-'aSl Si iflilir *
DsChyftl &»i5;^|i
UNDIVIDED
THICKNESS
CFT|
6— 30
2B - 1 , 330
D-1 20C
o- z aoo
0-i.»0
o-?,s«D
U i- C.'S
0 3»
0 2S&
1 , 3DO-
2 BOO
«.*f. fl-'.i
u"-hs
D 3D
IBS 23S
3D OD
IBS 4CC.
ioa-?t.«j
I3t 1SU~'
'SC <DO
20 ict:
JKI -'-;t:i
0-1. ZOO
5. om~
to.ooo
LITHOLOG¥
AHwtfiyr^, tiaii^^ Ho**H
Ligtl q'ay c jugljmiji^ic h,t jrc CD»J >-ufiiit
Bun >:f>3 ^moraii! jrj «sanfl4lunff
Mr fH^Y^hflr«
Red. pipfc, and wH»»a.f«liiona, BPO rwi, Dr*¥-
and ion cinyslone
Hud arttow conokmigraiD trj wndi-icn?,
ytlv* Bilrtpno, jr>d t^igln
S jy'sl-r^stB M b^ SdniH .^V^ .!*£ p -.i^j %Ui!«f;iis
H's**^ M *A i~i *tj ij,f*Jn>mu&m & !:s*&4"
Ctii'i qray LitVAiracua'i snn-i E^-H la 3"tt|.
Mdnraujous rtsHw ,-JTC sss-ancre aoz :aif
t«i|?i*. ^r^v !o ^ufl ^tit"iUSiCJ""it
DDrt-gmv HBHIO malo and iillstDira
I'ffi^irr* f-'iet^ munriet gj^ay ir^™0 mnl fin hHn*JS
*ftumlur^ Hru jijt* -y'^ii ^nSLr-p fe" li«*»si»
3"o»<". si" 5a'">3si5no ca'h g'av i>~'if 3"Jy
nfTiailci"'C anc jiav shjlo
3v.H w»tayjo» &.*< co%>c»~ciar: &an3i;:-«
ixl dJ'k, g-ay snair
VMT'l'^MluC "IIUICU'1 Ia*ili'*!, l^fa1, »'^KSU^'J8
«in! 'iill^lsii'e1 yy{!5i4u^n ui u j'tiy s«ii^£ia'^ »
HHr! %.nrAli,!y!iM c«*!i.i»tv*j.jp s* *lHb« **«'c
Shn -tMrnitifma
M'ai «-aaiCd I-'S*! h'kOW ccrciCTBrt'i-i.
ii'ia ihr Tiar^e !.•*•<«»-*
Generalized Stratigraphy of Cenozoicand Mesozoic Units (Hen-borg. 1998 1
'•.. ... • •-,- :*,• i -. • Fgirc- ftp-".
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
._..ii/_,*,i~,_v ^--:.l.-:
; —'/? \' "'^-*.V.r:. -'/
-Total Coat isopach (Stevens da!., 1992)
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A9-9
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
vti'iK.roon Its Caa riieivn1 (St^vu^s t.'t al., 1992}
Figirc- ft.9-5
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A9-10
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
'. undf an o* St*a*;prapi"ir ClTosH BeciiDns [Htev-Hns fit ^: 1^
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A9-11
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ~-
en ^ en
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. oo
C3 O
OQ g
o ri
«5 rt
> I
i O
i—- O
K< -1^
*
5F __ j^r. x
i2 _ *»'«-«a-ts~ii ^ ^»- ^^ ^
.'."r^ -rd^"""
1
1
' 3
• - u . "t"
? v M - jj
-" « I ^ '
*• * ^ - I-J
* r^f 1 ' O
J , n -a«Kpi I s F|
1
,,,=!.,
'*, "" , H
f.,1'* 11 ",
:..- " — I
"^
g ? 3 '
i; -
~ * "
4> \: ^
•t_ r _--• a — — •<-
H*™ "~~"h~""
^ H ^
L |^. |
Cross Section A-A (Stevens
.-• rv - • - - -J
". , ijyf I ,
,"* •• '
'i II |
H
I
- ---_ji r
j ;
'
setaL 1992}
i «t *.**
Ts. -"''* *** !-
1" ' - '"'
n V.
H ^ .
1 " \
y , * i ***** *i
j ^ ,*JJ!T!-^\ ^ 5:M?
r n — i^Linjj
H
™i
Ls "
^ -i / !'
f
RJlfSX §A3Ik
n,'.^"C*./-**.tr- i,-.^.
oo
K^
ON
s
O
Cd
I
5'
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
li°
g p? y
ft ?t ng
«^ ^
2. B- C
a i a
-
0 OQ
^ 3
3 §
o id
3. oo
C3 O
OQ g
o ri
«5 rt
> I
i O
i—- O
U) -1^
2 s|
sr i
II,
J
M
H
u
a»r»« p{
1 ,
»,
*
-
t
4t *—
a -
"
'
I— 1
h
I !
|
3 U
" L
__~ —
ikiiati i«ii**»^9,fi ib*i
*-tBi r* p
if*i j*i «i^
la** C««r«p-' «$+ *is*
PIATdM
Section C-C' (Stevens el al.( 1992)
uj'c A9-U
cc
H-'
0\
s
O
Cd
I
5'
image:
EPA816-R-04-003
Attachment 9
The Raton Basin
--,-• - \h
*'•<•'•' v£ 3>x"
Potefitometnc Surface Map tor Raton Basin (Qldaker et al,, 1993J
fi.9-6
Evaluation of Impacts to Underground Sources
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs
June 2004
A9-14
image:
Q2 C^ ^
a- 3. £T
rt B to
o- R- ~-
en ^ en
2. B- C
a i a
ti- ft
0 OQ
^ 3
3 §
o id
£ &
3. oo
C3 O
OQ g
o ri
^+1 rt
> I
i O
i—- O
(Jl -1^
9000;- -
— BOOO -
w
ro 1
w
— -
"O
ca j
_P |
CL 6000 J"^
•
10
*
•
,»,||lt A ji i $?•, • *
rf> V^flV'2' 4^%* * /„', A A
if* *
*V .. JJ-^ „ .^ .
« Poison C; ny
& Raton
« Verrnejo
4 Trinidad
« -'.- t ^t
+ " . ^
i, -tjL
Ml J
J
ft
A
- — ....
A * "' ft
x *» *„ 4"
***« *x
>*. *^, + •
100
Total DfsoIvccJ Solids, (rng/l}
RaiatiOpts^hip Batwean Gaa Contact gno Depth EieSom1 Pontenhometnc Surface far Two Gmupg of Coal RarK
lOidaceret al.. 1S93;«
oo
ON
O
U)
s
O
Cd
8
image:
EL a EL
C? 3. G
ft B ?5
P. R- £••
5 S 3
ft f?-H
.^
2. & C
a i a
fj- fO
0 OQ
^ 3
3 §
o id
5 o
OQ g
O O
^+J ft
C/l
> I
i O
i—- O
ON -l^
aoa.
BOO
700-
/ i
f '
i
1 l
i 1
> l
5 1 / '
W 1 ' |
0- BOO I ,' I
fT*" ' * 'f
M- !
t^ 1
x !
^T 5QC • \
Q> I
* 1
;| 4-j.; - ^
t> i
•o
c
s
3
O
t,
^ ZOO —
o
1OO-
1
I
4r ^
~ ^.^-^P'
j ^:'*''
*-i^-,--* * \ y *"*•„
\^/ ~ "^
1 3- 3 4 & 6 7 8 0 10 tl
i
iti5
__ _ _ ^^ =_ _ «-i___a«._A>^.™«
*__^ W*»f P»tf Mdl*n
/^^"^N^,
J/ ^S.
*--•
_^-
/
/
*-• "* — *_ »_ ,
12 i z 3 4 es 6 r
l
1«H
HistO' en- G;is dix) Walt's' P't>.it of ion *0' T/p'ca We i BH
How Watfrr 'fV.thcJra*3l IfiC'fjases a^t.
Merhai^ Pror^jruoo Mn-easeF
ASOO
4000
350D
1
o
3
o
i C
s LJSQO -^
.—
h_
5 a
^ • £
V ^
^^-s^«^^
* -*•- *^ * -*
I ' T I
« 9 10 11 12
•
^
«
4j
- ^
w
IE
- 5<3O
"}'ft i'q
. iHe:-norq 1998J
• •. •>. .-••«.."... hgore A9-T1
oo
K^
ON
I
ft 5
s
P4. 3
O vo
Cd
I
5'
image:
EL a EL
C? 3.
ft B
P. R-
ft ft
2. B- C
a i a
-
0 OQ
^ 3
3 §
o id
-
OQ
o
1
K»
O
O
s»*
at
Q
a
o
s
81
it,
o
'j3
C
nj
A
3
O
C3
700 •
8QG-
500
40fl
3DQ
200
100
700
500
w
u.
J*
3
Q'
C
a
e
a.
3CM fc
200 "i
ka
Q
D
'1OD 5
^
1 2
121 33 4Si78«JlOM"2
Historical Gas and Water Production Icr Oz/e o*li"-1 Vtfe Shading
Water Withdrawal Increasing with Gas Procuction iHeinbary, 1S98)
oo
K^
ON
I
ft 5
s
P4. 3
O vo
Cd
I
5'
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
REFERENCES
Carter, D. A. 1956. Coal deposits of the Raton Basin; in McGinnis, C. J., ed., Geology
of the Raton Basin, Colorado: Rocky Mountain Association of Geologists
Guidebook, pp. 89-92
Gas Technology Institute (GTI) Web site. 2002. Drilling and Production Statistics for
Major US Coalbed Methane and Gas Shale Reservoirs.
http://www.gastechnology.org.
Geldon, A. L. 1990. Ground-water hydrology of the Central Raton Basin, Colorado and
New Mexico: U.S. Geological Survey Water-Supply Paper 2288, 81 p.
Flores, R. M. 1987. Sedimentology of Upper Cretaceous and Tertiary sillicielastics and
coals in the Raton Basin, New Mexico and Colorado, in Lucas, S. G., and Hunt,
A. PI, eds., Northeastern New Mexico: New Mexico Geological Society Annual
38th Field Conference, pp. 255-264.
Flores, R. M., and Pillmore, C. L. 1987. Tectonic control on alluvial paleoarchitecture
of the Cretaceous and Tertiary Raton Basin, Colorado and New Mexico, in
Ethridge, F. G., Flores, R. M., and Harvey, M. D., eds., Recent developments in
Fluvial Sedimentology: Society of Economic Paleontologists and Mineralogists
Special Publication 39, pp. 311-321.
Hemborg, H. T. 1998. Spanish Peak Field, Las Animas County, Colorado: Geologic
setting and early development of a coalbed methane reservoir in the Central Raton
Basin. Colorado Geological Survey, Dept. of Natural Resources, Denver, CO,
Resource Series 33, 34 p.
Hemborg, H. T. 1996. Raton Basin coalbed methane production picking up in Colorado.
Oil & Gas Journal, pp. 101-102 (Nov 11. 1996).
Howard, W.B. 1982. The Hydrogeology of the Raton Basin, South-Central Colorado.
M.A. Thesis, Department of Geology, Indiana University.
National Water Summary. 1984. Hydrologic events, selected water-quality trends, and
ground-water resources. United States Geological Survey Water-Supply Paper
No. 2275.
Oldaker, P., Stevens, S.H., Lombardi, T.E., Kelso, B.S., and McBane, R.A. 1993.
Geologic and hydrologic controls on coalbed methane resources in the Raton
Basin. Proceedings of the 1993 International Coalbed Methane Symposium,
Tuscaloosa, AL, pp. 69-78 (May 17-21, 1993).
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-18
image:
EPA 816-R-04-003 Attachment 9
The Raton Basin
Quarterly Review of Methane from Coal Seams Technology. 1993. Raton Basin
Colorado and New Mexico. Methane from Coal Seams Technology, pp. 33-36
(August).
Steven, T. A., 1975. Middle Tertiary Volcanic field in the southern Rocky Mountains: in
Curtis, B. F., ed. Cenozoic History of the Southern Rocky Mountains: Geological
Society of America Memoir 144, pp. 75-91.
Stevens, S., Lombard!, T. E., Kelso B. S., and Coates, J. M. 1992. A geologic
assessment of natural gas from coal seams in the Raton and Vermejo Formations,
Raton Basin. GRI Topical Report 92/0345, 84 pp.
Tyler, R., Kaiser, W. R., Scott, A. R., Hamilton, D. S., and Ambrose, W. A. 1995.
Geologic and hydrologic assessment of natural gas from coal: Greater Green
River, Piceance, Powder River, and Raton Basins, Western United States: Austin,
Tex., Bureau of Economic Geology, Report of Investigations 228, 219 p.
Evaluation of Impacts to Underground Sources June 2004
of Drinking Water by Hydraulic Fracturing of
Coalbed Methane Reservoirs A9-19
image:
