Well Intergity

Final Technical Report
Integrated Wellbore Integrity Analysis Program for CO2 Storage Applications
U.S. Department of Energy National Energy Technology Laboratory

DOE Award No. DE-FE0026585
Principal Investigator: J.R. Sminchak
[email protected] 614-424-7392
September 28, 2018
Recipient: Battelle
505 King Avenue
Columbus, OH 43201-2696
DUNS Number: 00 790 1598
Project Grant Period: October 1, 2015, and ending September 30, 2018
U.S. Department of Energy Disclaimer
This report was prepared as an account of work sponsored by an agency of

the United States Government. Neither the United States Government nor
any agency thereof, nor any of their employees, makes any warranty,
express or implied, or assumes any legal liability or responsibility for the
accuracy, completeness, or usefulness of any information, apparatus,
product, or process disclosed, or represents that its use would not infringe
privately owned rights. Reference herein to any specific commercial
product, process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein do not
necessarily state or reflect those of the United States Government or any
agency thereof.
Battelle Disclaimer
Battelle does not engage in research for advertising, sales promotion, or

endorsement of our clients’ interests including raising investment capital
or recommending investments decisions, or other publicity purposes, or for
any use in litigation. Battelle endeavors at all times to produce work of the
highest quality, consistent with our contract commitments. However,
because of the research and/or experimental nature of this work, the client
undertakes the sole responsibility for the consequence of any use or misuse
of, or inability to use, any information, apparatus, process or result
obtained from Battelle, and Battelle, its employees, officers, or Trustees
have no legal liability for the accuracy, adequacy, or efficacy thereof.
Acknowledgements
Support for this project was provided by the United States Department of
Energy National Energy Technology Laboratory (NETL) under award
#DE-FE0026585. The project team very much thanks NETL for supporting
this effort, which is an important issue for CO2 storage projects. Project
guidance was provided by Bill O’Dowd (NETL). The project team thanks
Core Energy and the Petroleum Technology Research Council for
providing access to CO2 wells for testing and other support information.
Support for the Appalachian Basin site characterization was provided by
the West Virginia Geologic and Economic Survey, with special thanks to
Phil Dinterman, Gary Daft, Jessica Moore, and Ron McDowell.
Project results reflect contributions from many people on the project team.
The project lead was Battelle, and J.R. Sminchak was the project manager.
Mark Moody and Dr. Neeraj Gupta were technical advisors. Task leads
were Dr. Andrew Duguid, Matt Place, and Jared Hawkins. Other
contributions were made by Autumn Haagsma, Desiree Padgett, Kelly
Quigley, Leo Chiques, Jen O’Brian, Jacob Markiewicz, and others. Carol
Brantley, Colin Kegel, Theresa Schaff, and Jackie Gerst assisted with
project administration and management.
Abstract/Plain Language Summary
Exposure to CO2 in the subsurface is a concern for wellbore integrity at

CO2 storage sites. This project completed a review of 1,500 wells, field
survey of 83 wells, and detailed wellhead testing on 23 wells at three CO2
field sites. The field testing results did not show significant well defects.
Well construction and/or cement carbonation sealing may have contributed
to well integrity. Geochemical analysis suggests subsurface conditions at
the field sites were suitable for cement sealing of gas migration pathways
via calcium carbonate precipitation. Results support effective management
of CO2 storage applications in areas with many legacy oil and gas wells.
Table of Contents
Page
Chapter 1.0 Introduction ........................................................................................................... 1
1.1 Project Background ...................................................................................................................... 1
1.2 Objectives .................................................................................................................................... 1
1.3 Research on CO2 Storage and Wellbore Integrity........................................................................ 2
1.4 Summary of Selected Field Sites ................................................................................................. 7
Chapter 2.0 Well Integrity Registry ........................................................................................ 10
2.1 Well Construction Methods ....................................................................................................... 10
2.2 Well Casing Integrity ................................................................................................................. 11
2.3 Well Cement Integrity................................................................................................................ 12
2.4 Geologic Processes .................................................................................................................... 15
2.5 Key Findings of Well Integrity Registry.................................................................................... 15
Chapter 3.0 Well Record Data Collection and Review .......................................................... 18
3.1 Field Site Descriptions ............................................................................................................... 18
3.2 Michigan Basin Site ................................................................................................................... 19
3.3 Appalachian Basin Site .............................................................................................................. 22
3.4 Williston Basin Site ................................................................................................................... 25
3.5 Key Findings of Well Record Data Collection and Review ...................................................... 28
Chapter 4.0 Log- and Testing-Based Well Integrity Assessment ........................................ 29
4.1 Log Analysis .............................................................................................................................. 29
4.2 Well Record Analysis ................................................................................................................ 30
4.3 Well Integrity Analysis .............................................................................................................. 31
4.4 Key Findings of Log- and Testing-Based Well Integrity Assessment ....................................... 35
Chapter 5.0 SCP Analysis....................................................................................................... 37
5.1 Michigan Basin SCP Field Testing ............................................................................................ 37
5.2 Williston Basin SCP Field Testing ............................................................................................ 43
5.3 Key Findings of SCP Analysis................................................................................................... 51
Chapter 6.0 Field Analysis of CO2 Cement Sealing and Well Integrity ................................ 53
6.1 Analysis of Subsurface Setting for Cement Sealing .................................................................. 53
6.2 Geochemical Analysis to Predict Cement Sealing Conditions .................................................. 56
6.3 Review of Well Defects and Pore Network Dimensions for Field Sites ................................... 62
6.4 Key Findings of Field Analysis of CO2 Cement Sealing and Well Integrity ............................. 66
Chapter 7.0 Wellbore Integrity Sealing Factor Uncertainty Analysis .................................. 68
7.1 CO2 Batch Model Indicator Geochemical Analysis ................................................................... 68
7.2 CO2 Reaction Batch Model Indicator Geochemical Analysis ................................................... 72
Battelle | September 28, 2018 i

Table of Contents (cont)
Page
7.3 Geochemical Analysis Meta-Modeling ..................................................................................... 73
7.4 Key Findings of Wellbore Integrity Sealing Factor Uncertainty Analysis ................................ 80
Chapter 8.0 Reporting and Technology Transfer ................................................................. 82
8.1 Reporting.................................................................................................................................... 82
8.2 Technology Transfer .................................................................................................................. 82
Chapter 9.0 Conclusions ........................................................................................................ 84
Chapter 10.0 References ........................................................................................................ 86
Battelle | September 28, 2018 ii

List of Tables
Page
Table 1-1. Project task objectives and deliverables. ..................................................................................... 3
Table 1-2. Summary of research on wellbore integrity for CO2 storage....................................................... 3
Table 1-3. Summary of field sites. ................................................................................................................ 7
Table 2-1. Wellbore integrity registry of identified integrity issues. .......................................................... 16
Table 3-1. Summary of field sites. .............................................................................................................. 19
Table 4-1. Number of logs used for Otsego, Michigan, and Indian Creek, West Virginia, fields. ............. 29
Table 4-2. Number of wells in each field with known data for each category ranked in the well
integrity evaluation. ............................................................................................................... 30
Table 5-1. Construction specifications for Michigan Basin wells surveyed. .............................................. 38
Table 5-2. Field test results for Michigan Basin wells................................................................................ 39
Table 5-3. Gas sampling results of six tested wells. ................................................................................... 41
Table 5-4. Williston Basin well specifications. .......................................................................................... 44
Table 5-5. Gas testing records for the Williston Basin wells. ..................................................................... 46
Table 5-6. SCP analysis results for Williston Basin field site. ................................................................... 51
Table 6-1. Summary of mineralogy for the field test sites. ......................................................................... 53
Table 6-2. Summary of brine geochemistry for reservoir samples from the field test sites. ...................... 54
Table 6-3. Composition of typical Portland cement clinkers. ..................................................................... 55
Table 6-4. General hydrologic conditions for three field sites.................................................................... 56
Table 6-5. Questions to answer with model. ............................................................................................... 57
Table 6-6. Appalachian Basin chemistry summary data. Midpoint values were used for
geochemical modeling. .......................................................................................................... 58
Table 6-7. Michigan Basin chemistry summary data. Midpoint values were used for
geochemical modeling. .......................................................................................................... 58
Table 6-8. Batch reaction parameters. ........................................................................................................ 58
Table 6-9. Appalachian Basin solid equilibrium phases. ............................................................................ 59
Table 6-10. Michigan Basin solid equilibrium phases. ............................................................................... 60
Table 6-11. Amount of each mineral phase that precipitates from Appalachian Basin summary
solutions, with and without 1.1 mol of CO2. Precipitates are reported in mol/kg
water. Columns labeled “Δ with CO2” show the change in precipitation when the
solution is reacted with CO2. .................................................................................................. 62
Table 6-12. Amount of each mineral phase that precipitates from Michigan Basin summary
solutions, with and without 1.1 mol of CO2. Precipitates are reported in mol/kg
water. Columns labeled “Δ with CO2” show the change in precipitation when the
solution is reacted with CO2. .................................................................................................. 62
Table 6-13. Descriptive statistics of MICP-derived pore throat radius distribution for lithologies
at the Dover 33 reef field. ...................................................................................................... 63
Table 6-14. Summary of aperture data derived from fractures located within the caprock interval
of the Dover 33 FMI log. ....................................................................................................... 64
Table 7-1. Summary of four test study areas. ............................................................................................. 74
Table 7-2. Summary of test study area well construction specifications. ................................................... 75
Battelle | September 28, 2018 iii

List of Tables (cont)
Page
Table 7-3. Summary of general mineralogy for test study areas. ............................................................... 77
Table 7-4. Summary of brine geochemistry for test study areas................................................................. 77
Table 8-1. Summary of deliverables. .......................................................................................................... 82
Battelle | September 28, 2018 iv

List of Figures
Page
Figure 1-1. Project task organization chart. .................................................................................................. 2
Figure 1-2. Location of the Indian Creek field in Kanawha County, West Virginia. ................................... 8
Figure 1-3. Structure map of the Brown Niagaran showing the pinnacle reef trend (gray),
Antrim shale trend (red), and the Michigan Basin site (star). .................................................. 9
Figure 1-4. Location of the Williston Basin field site. .................................................................................. 9
Figure 2-1. Diagram illustrating possible leakage pathways within a well. ............................................... 11
Figure 2-2. Photos showing the results of well cement exposed to flowing carbonated brine at
50°C and pH 3. Left photo shows the outer orange and brown silicate-hydrate zone
outside of a CH-depleted cement zone, with an unreacted cement in the center.
Right photo shows the outer orange and brown silicate-hydrate zones, the white
CaCO3-rich zone, and a CH-depleted cement zone in the center. .......................................... 13
Figure 2-3. Wellbore schematic showing types of wellbore integrity issues and where they occur
(Sminchak et al., 2016). ......................................................................................................... 17
Figure 3-1. Summary of well record collection parameters. ....................................................................... 18
Figure 3-2. Typical Antrim (left) and Niagaran (right) well construction. ................................................. 20
Figure 3-3. Monthly and cumulative oil production from Niagaran wells showing greatest
monthly production in the early 1980s and a steady monthly decline through 2016. ............ 21
Figure 3-4. Wellbore diagram of an Indian Creek well showing common well construction. ................... 23
Figure 3-5. Histogram of sacks of cement used for each casing string in the Indian Creek wells.............. 24
Figure 3-6. Annual and cumulative production in the Indian Creek field since 1985. ............................... 25
Figure 3-7. Status of wells at the Williston Basin site. ............................................................................... 26
Figure 3-8. Cement volumes for wells at the Williston Basin site.............................................................. 27
Figure 3-9. Casing pressure category for wells at the Williston Basin site. ............................................... 27
Figure 4-1. Histograms of Total Likelihood and Total Severity Scores (top) and Total Risk
Scores (bottom), Northern reef trend, Otsego County, Michigan. ......................................... 32
Figure 4-2. Total Risk Scores, Northern reef trend, Otsego County, Michigan. ........................................ 33
Scores (bottom), Indian Creek Field, Kanawha County, West Virginia. ............................... 34
Figure 4-4. Total Risk Scores, Indian Creek Field, Kanawha County, West Virginia. .............................. 34
Scores (bottom), Weyburn Field, Saskatchewan, Canada. .................................................... 35
Figure 5-1. Timeline of production and EOR of Michigan Basin reefs, outlining the primary
production, pressurization period when CO2 was being injected without
production, and EOR period. ................................................................................................. 37
Figure 5-2. Map of tested wells in the Michigan Basin, among many other wells in the area. .................. 38
Figure 5-3. Well diagrams for Michigan Basin wells tested for SCP buildup. ........................................... 40
Figure 5-4. Pressure and temperature testing data from Michigan Basin wells. ......................................... 42
Figure 5-5. Map of studied wells in the Williston Basin field test site. ...................................................... 43
Figure 5-6. Williston Basin wellbore diagrams. ......................................................................................... 44
Figure 5-7. Well testing apparatus at the Williston Basin site. ................................................................... 45
Figure 5-8. Pressure-temperature testing data from Williston Basin wells WB-1 though WB-8. .............. 47
Battelle | September 28, 2018 v

List of Figures (cont)
Page
Figure 5-9. Pressure-temperature testing data from Williston Basin wells WB-9 though WB-16. ............ 48
Figure 5-10. General SCP buildup curve. ................................................................................................... 49
Figure 5-11. SCP plots for Williston Basin wells. ...................................................................................... 50
Figure 6-1. Log SI values for cement mineral phases in initial Appalachian Basin summary
solutions. ................................................................................................................................ 61
Figure 6-2. Log SI values for cement mineral phases in initial Michigan Basin summary
solutions. ................................................................................................................................ 61
Figure 6-3. Cement channel/fracture identified within the Charlton 4-30 well from 3,052 to
3,080 ft, highlighted by the yellow circle. Note the low attenuation (blue color) and
its gas saturation (red) illustrated on the Solid Liquid Gas Map (right-most track)............... 65
Figure 6-4. Distribution of measured calculated defects (mm) within the intermediate casing
interval of the Charlton 4-30 well. ......................................................................................... 66
Figure 7-1. pH and pe of the Appalachian Basin summary solutions, CO2 batch reactions.
Reference pH and pe are simulated for CO2 added to pure water for comparison. ............... 68
Figure 7-2. Log SI values for cement mineral phases in Appalachian Basin summary solutions,
CO2 batch reactions. ............................................................................................................... 69
Figure 7-3. CO2 batch delta molar mass (moles of mineral phases that precipitated or dissolved
from previous CO2 batch step) and moles of mineral phases with midpoint
summary solution for Tuscarora sandstone, Rose Hill formation, and cement
phases. .................................................................................................................................... 70
from previous CO2 batch step) (top) and moles of mineral phases (bottom), State
Chester limestone, Dover 33 limestone, Dover 33 A2 carbonate, and cement
mineral phases with midpoint summary solution. The reference pH and pe (Ref.
pH and Ref. pe) were found by reacting the CO2 batch with pure water alone. .................... 71
Figure 7-5. Log SI values for cement minerals batch reaction for pH, temperature, and pressure,
Appalachian Basin. ................................................................................................................ 72
Figure 7-6. Map showing four test study areas. .......................................................................................... 73
Figure 7-7. Map of production casing cement thickness across test study areas. ....................................... 76
Figure 7-8. Baseline pressure conditions at test study areas. ...................................................................... 78
Figure 7-9. Flow chart illustrating meta-modeling process for evaluating geochemical cement
sealing conditions across the four test study areas. ................................................................ 79
Figure 7-10. Meta-modeling output for test study area 4 at pressure +0, temperature +10 °C, and
pH -2. ..................................................................................................................................... 80
Figure 8-1. EDX data upload submission summary. .................................................................................. 83
Battelle | September 28, 2018 vi

List of Acronyms and Abbreviations
Bcf billion standard cubic feet
CaCl2 calcium chloride
CaCO3 calcium carbonate
Ca(OH)2 Portlandite
CBL cement bond log
CCS carbon capture and storage
CCUS carbon capture, utilization, and storage
CO2 carbon dioxide
cm3 cubic centimeter
CH calcium hydroxide
C-S-H calcium-silicate-hydrate
DEP Department of Environmental Protection (State of West Virginia)
DEQ Department of Environmental Quality (State of Michigan)
DOE U.S. Department of Energy
EDX Energy Data Exchange
EOR enhanced oil recovery
FEP Features, Events, and Processes
ft foot
ft3 cubic foot
kg kilogram
kg/L kilograms per liter
km kilometer
LCM lost circulation material
LLNL Lawrence Livermore National Laboratory
LTSI long term shut in
μm micrometer
m meter
MD measured depth
mg/kg milligrams per kilogram
mg/L milligrams per liter
Battelle | September 28, 2018 vii

List of Acronyms and Abbreviations (cont)
MICP mercury injection capillary pressure
mm millimeter
MPa megapascal
MRCSP Midwest Regional Carbon Sequestration Partnership
MSCFD thousand standard cubic feet per day
NETL National Energy Technology Laboratory
nm nanometer
NRAP National Risk Assessment Partnership
psi pounds per square inch
PTRC Petroleum Technology Research Centre
SCP sustained casing pressure
SI saturation index
SiOxOHx silicate-hydrate
TD total depth
TOC top of cement
TVD true vertical depth
WAG water alternating gas
WVGES West Virginia Geologic and Economic Survey
Battelle | September 28, 2018 viii

Executive Summary
Carbon dioxide (CO2) injection for geologic carbon storage and enhanced oil recovery (EOR) may result
in mixtures of CO2 and water contacting new and legacy wells in the deep subsurface. Exposure to CO2 is
a concern for wellbore integrity, because CO2 can corrode well materials and migrate along defects
around the borehole potentially reaching near surface groundwater resources or the atmosphere. This
project evaluated well integrity in CO2 wells with a combination of direct field testing and well records
analysis. Key accomplishments and results of this project are summarized as follows:
• Approximately 1,500 wells at three field sites were reviewed in terms of well construction,
history of exposure to CO2, geochemistry, mineralogy, and well materials:
▪ Appalachian Basin. The Appalachian Basin Indian Creek site is a methane and natural
CO2 field located in Kanawha County, West Virginia. The field contains 58 wells at total
depths between 6,200 feet (ft) and 6,700 ft in the Tuscarora sandstone. The field has
natural pockets of CO2 at levels up to 60%.
▪ Michigan Basin. The Michigan Basin field site is located in Otsego County, Michigan.
Carbonate reef fields were developed since the 1960s in the region, and selected reefs
have been subject to CO2 EOR since the 1990s. CO2 is present in the Antrim gas wells
between 5% and 30% at depths between 1,000 ft and 1,500 ft and in the Niagaran Reef
EOR wells at depths ranging from 5,000 ft to 7,000 ft.
▪ Williston Basin. The Williston Basin field is in Saskatchewan Canada with operations
dating back to 1954. More than 3,000 wells are present at the Williston Basin testing site,
completed at depth of approximately 6,000 ft to 7,000 ft. CO2 EOR was started in 2000 at
the site, expanding to additional areas over time.
• A total of 83 CO2 wells were surveyed at the Michigan Basin site (23 wells) and the Williston
Basin site (60 wells) for wellhead casing pressures that may indicate well defects. The
Appalachian Basin site was not available for testing, because the asset was sold to a new operator.
• Detailed sustained casing pressure (SCP) testing was completed on 23 wells that had indications
of significant SCP.
• The testing results did not show significant well defects, with casing pressures less than
1 megapascals (MPa) and minor pressure buildup patterns. There was no evidence of significant
defects or CO2 migration in the wells that were tested.
• Additional geochemical modeling and meta-modeling for the three field sites and four test study
areas indicated that mineralogy, hydrologic conditions, cement blends, and brine geochemistry
were not critical factors to the cement carbonation process.
• Well construction and/or cement carbonation sealing appears to have contributed to well
integrity. Results support effective management of CO2 storage applications in areas with many
legacy oil and gas wells.
The three field sites have wells which have been exposed to CO2, either naturally or through EOR
operations, for 5 to 50+ years under different geologic settings and subsurface conditions. These datasets
provide unique opportunities to study the influence of CO2 on wellbore integrity. The field testing was
completed on a subsample of wells and does not mean all CO2 wells would be free of defects. In addition,
the SCP testing methodology requires defects that would lead to gas migration to the wellhead. Therefore,
there may be existing downhole defects not revealed by the testing. Project results demonstrate that well
construction procedures, well design, and well logging/testing for defects are important considerations for
wellbore integrity in CO2 environments in the subsurface. Additional work on the life-cycle effects of
CO2 would help highlight changes over time due to subsurface exposure to CO2 in wells.
Battelle | September 28, 2018 ix

Chapter 1.0 Introduction
This Final Technical Report presents the findings for the project Integrated Wellbore Integrity Analysis
Program for CO2 Storage Applications (FE0026585). The project is part of a U.S. Department of Energy
(DOE) National Energy Technology Laboratory (NETL) program to develop and advance technologies
that will significantly improve the effectiveness and reduce the cost of implementing carbon storage. The
project is designed to assess well integrity for wells exposed to carbon dioxide (CO2) environments in the
subsurface through a combination of field testing and record analysis.
1.1 Project Background

Legacy oil and gas wells are considered a key risk factor for carbon capture, utilization, and storage
(CCUS) projects because they represent the most likely migration pathway out of a reservoir into
overlying formations. Each well at a CCUS project, whether a legacy well or a well used for injection,
production, or monitoring of the CO2 plume, must effectively ensure that 99% of the CO2 is accounted for
within the subsurface. The goal of ensuring that CO2 is effectively captured and stored on longer, more
geologic time scales must overcome leakage risk through proper engineering and construction of
wellbores.
The overall objective of this research project was to develop and validate a program for identifying and
characterizing wellbore integrity issues for potential CO2 storage applications based on analysis of well
records validated with SCP field testing. The project involved analyzing existing well data for several
fields where wells have been exposed to CO2 and analyzing new data collected from SCP testing.
Together, these data and analyses were used with geochemical analysis to identify trends that lead to
better understanding and prediction of well integrity issues. The project was designed to result in
predictive methods to survey, identify, characterize, and remediate wellbore integrity issues for CO2
storage applications. The project was divided into the eight tasks shown in Figure 1-1. Full reports were
prepared for each major task and the field testing. As such, this report focuses on presenting results of the
project tasks. The task reports are available in case more details are required.
1.2 Objectives
The project goal was to develop an integrated program to identify, survey, measure, and analyze CO2
migration in wellbores. After a well has been constructed and/or plugged, the only indication of migration
through the outer well materials may be pressure buildup on the well, referred to as sustained casing
pressure (SCP). The impact of CO2 on wellbore integrity was determined by integrating field casing
pressure test results with analysis of cement sealing potential, well construction details, well logs, cement
bond logs (CBLs), and well history. In addition, the types of well defects (micro-annulus, cracks, porous
cement, and incomplete cement coverage) were explored by analyzing casing pressure buildup curves
measured in the field on CO2 wells. Meta-modeling methods were used on CO2 storage test fields to
investigate the impact of pressure, gas saturation, and chemistry on well integrity.
Battelle | September 28, 2018 1

Figure 1-1. Project task organization chart.
The project was organized into eight main tasks with more detailed subtasks. Table 1-1 lists the task
objectives, technical approach, and corresponding deliverables. A key objective of the project was to
complete field measurements of casing pressure at several locations with existing boreholes that have
wells exposed to CO2 in the Michigan Basin, Appalachian Basin, and Williston Basin. The casing
pressure testing results were integrated with analysis of cement sealing to predict well integrity problems
in wells exposed to CO2 in terms of leak location, nature, and severity. Based on SCP test results at the
field sites, Tasks 6 and 7 were modified to evaluate cement sealing conditions.
1.3 Research on CO2 Storage and Wellbore Integrity

Many research studies have evaluated the effects of CO2 on wellbores in the subsurface (Table 1-2). In
general, the studies have focused on laboratory testing of cement samples, surveys of existing well data,
modeling of CO2 exposure and/or migration, and field studies on wells (Zhang & Bachu, 2011). Given all
this research, there are relatively few direct field studies with field testing of multiple CO2 wells, because
it is difficult to access wells in the subsurface and to collect samples from these wells. In addition,
operators are hesitant to participate in research on wellbore integrity.
Therefore, many researchers have completed laboratory tests on prepared cement samples. Laboratory
tests have generally confirmed the potential for geochemical reactions with Portland cements resulting in
dissolution of cement and precipitation of calcium carbonate (CaCO3). Laboratory tests have included
diffusion-based tests in pressure vessels and flow-through tests with CO2 and brine mixtures. Field studies
have taken sidewall cores through casing and cement that have shown a dissolution front and
mineralization front. Modeling studies have had more varied results simulating CO2 migration within well
boreholes, since it is difficult to determine cement permeability and pathways in boreholes. Studies have
also examined cement sealing potential for different fracture aperture widths, concluding that CaCO3
mineralization may reduce permeability in fractures less than approximately 50 to 200 nanometers (nm)
wide.

Table 1-1. Project task objectives and deliverables.
Milestone
Task Objective Deliverable
Description
Project Project Management Plan (Oct
Coordinate project schedule, budget, progress
1 Management & 2015), Quarterly Research
reporting, and planning.
Planning Performance Progress Reports
Develop a registry of potential well defects for
CO2 storage applications based on well
Wellbore Integrity Well Integrity Registry Summary
2 construction methods, well casing integrity
Registry Report (June 2016)
issues, well cement issues, geologic
processes, and CO2 subsurface environments.
Well Record Data Describe field sites based on cementing,
Well Record Data Summary Report
3 Collection & drilling, operational, and well workover
(June 2017)
Review records.
Log & Testing
Complete analysis of field sites based on well Log & Testing Based Well Integrity
Based Well
4 logs, well historical records, and quantitative Assessment Summary Report
Integrity
well integrity indicator analysis. (November 2017)
Assessment
Complete field testing of SCP in 20+ existing
Michigan Basin Field Testing
wells at field sites in Michigan Basin,
Summary Report (March 2017),
5 SCP Analysis Appalachian Basin, and Williston Basin.
Williston Basin Field Testing
Analyze results for quantitative indicators of
Summary report (February 2018)
wellbore integrity defects.
Field-Based Analyze field data on mineralogy, fluids,
Field-Based Analysis of CO2
Analysis of CO2 cement, hydrologic conditions & CO2 exposure
6 Cement Sealing Summary Report
Cement Sealing for the three field sites to determine cement
(June 2018)
Conditions sealing & well integrity relationship
Wellbore Integrity Field-Based Analysis of CO2
Examine sealing conditions uncertainty for
Sealing Conditions Cement Sealing Summary Report
7 CCS projects in areas where there are a large
Uncertainty (June 2018), Final Technical
number of existing wells with meta modeling.
Analysis Report
Reporting & Document project results and distribute project Technical Reports, Presentations,
8 Technology data for other CO2 storage research and Final Technical Report (September
Transfer applications 2018)
Table 1-2. Summary of research on wellbore integrity for CO2 storage.

Author Date Topic Category
Bruckdorfer 1984 Carbon dioxide corrosion in oilfield cements Lab testing
Burke 1984 Synopsis: Recent Progress in the Understanding of CO2 Data survey
corrosion
Onan 1984 Effects of supercritical carbon dioxide on well cements Data survey,
field study
Shen and Pye 1989 Effects of CO2 attack on cement in high-temperature Data survey
applications
Bonett & Pafitis 1996 Getting to the root of gas migration Data survey
Chen et al. 2002 CO2 corrosion for oil tube steel Lab testing

Rochelle et al. 2002 Geochemical interactions between supercritical CO2 and the Lab testing
Midale Formation. II: Initial results
Midale Formation. I: Intro to fluid-rock experiments
Kermani & Morshed 2003 Carbon dioxide corrosion in oil and gas production—a Data survey
compendium
Midale Formation. III: Midale Fmt
Bateman et al. 2004 Geochemical interactions between supercritical CO2 and the Lab testing
Midale Formation. VI: Midale Marly
Boukhelifa 2004 Evaluation of Cement Systems for Oil and Gas Well Zonal Data survey
Isolation in a Full-Scale Annular Geometry
Duguid et al. 2004 The effect of CO2 sequestration on oil well cements. Lab testing
Gasda et al. 2004 Spatial characterization of the location of potentially leaky Data survey
wells penetrating a deep saline aquifer in a mature sed.
basin
Rocelle et al. 2004 Interactions between supercritical CO2 and borehole Lab testing
cements used at the Weyburn oilfield
Ladva et al. 2005 The Cement-to-Formation Interface in Zonal Isolation Modeling,
lab tests
Cui et al. 2006 Study on corrosion properties of pipelines in simulated Lab testing
produced water saturated with supercritical CO2
Czernichowski- 2006 Geochemical Interactions between CO2, Pore-Waters and Lab testing
Lauriol Reservoir Rocks
Duguid et al. 2006 The effect of carbonated brine on the interface between well Lab testing
cement and geo. formations under diffusion-controlled
conditions.
U.S. DOE 2006 Degradation of wellbore cement due to CO2 injection Modeling,
lab tests
Vignes et al. 2006 PSA Well Integrity Survey, Phase 1 summary Report Data survey
Carey et al. 2007 Analysis and performance of oil well cement with 30 years Field study
of CO2 exposure from the SACROC unit, West Texas, USA
Kutchko et al. 2007 Degradation of well cement by CO2 under geological Lab testing
sequestration conditions
Randhol et al. 2007 Ensuring well integrity in connection with CO2 injection Data survey
Bachu & Watson 2008 Review of failures for wells used for CO2 and acid gas Data survey
injection in Alberta, Canada
Carey et al. 2008 Wellbore integrity and CO2-brine flow along the casing- Field study
cement microannulus
Duguid 2008 An estimate of the time to degrade the cement sheath in a Modeling,
well exposed to carbonated brine lab tests
Kutchko et al. 2008 Rate of CO2 attack on hydrated Class H well cement under Lab testing
geological sequestration conditions
Lecolier et al. 2008 Behavior of permeable steel/cement interface in contact with Lab testing
CO2-saturated brine
Liteanu et al. 2008 Failure behaviour of wellbore cement in the presence of Lab testing
water and supercritical CO2

Rimmele et al. 2008 Heterogeneous porosity distribution in Portland cement Lab testing
exposed to CO2-rich fluids
Strazisar et al. 2008 Chemical reactions of wellbore cement under CO2 storage Lab testing
conditions: effects of cement additives
Barlet-Gouédard et 2009 A solution against well cement degradation under CO2 Lab testing
al. geological storage environment
Watson & Bachu 2009 Evaluation of the Potential for Gas and CO2 Leakage Along Modeling
Wellbores
Wigand et al. 2009 Geochemical effects of CO2 sequestration on fractured Lab testing
wellbore cement at the cement/caprock interface
Huerta et al. 2009 Utilizing SCP Analog to Provide Parameters to Study CO2 Modeling
Leakage Rates Along a Wellbore study
Crow et al. 2010 Wellbore integrity analysis of a natural CO2 producer Field study
Krupka et al. 2010 Thermodynamic Data for Geochemical Modeling of Data survey
Carbonate Reactions Associated with CO2 Sequestration -
Lit Rev
Carey and Lichtner 2011 Computational studies of two-phase cement–CO2–brine Modeling
interaction in wellbore environment study
Han et al. 2011 A coupled electrochemical–geochemical model of corrosion Modeling
for mild steel in high-pressure CO2–saline environments study
Han et al. 2011 Effect of debonded interfaces on corrosion of mild steel Lab testing
composites in supercritical CO2-saturated brines
Liteanu & Spears 2011 Fracture healing and transport properties of wellbore Lab testing
cement in the presence of supercritical CO2
Schaef et al. 2011 Brucite [Mg(OH)2] carbonation in wet supercritical CO2: An Lab testing
in situ high pressure X-ray diffraction study
Scherer et al. 2011 Characterization of cement from a well at Teapot Dome Oil Field study
Field: implications for geological sequestration
Yalcinkaya et al. 2011 Experimental study on a single cement-fracture using CO2- Lab testing
rich brine
Zhang & Bachu 2011 Review of integrity of existing wells in relation to CO2 Data survey
geological storage: What do we know?
Pan et al. 2011 Transient CO2 leakage and injection in wellbore-reservoir Modeling
systems for geologic carbon sequestration study
Agbasimalo & 2012 Experimental Study of Portland Cement/Rock Interface in Lab testing
Radonjic Relation to Wellbore Stability for Carbon Capture and
Storage
Han et al. 2012 Degradation of cement–steel composite at bonded steel- Lab testing
cement interfaces in supercritical CO2 saturated brines
Jacquemet et al. 2012 Armouring of well cement in H2S–CO2 saturated brine by Modeling,
calcite coating – experiments and numerical modelling lab tests
Cao et al. 2013 Dynamic alterations in wellbore cement integrity due to Lab testing
geochemical reaction in CO2-rich environments
Carey 2013 Geochemistry of wellbore integrity in CO2 sequestration: Lab testing
Portland cement–steel–brine–CO2 interactions
Choi et al. 2013 Wellbore integrity and corrosion of carbon steel in CO2 Data survey
geologic storage environments: A literature review
Hawkes & Gardner 2013 Pressure transient testing for assessment of wellbore Field study
integrity in the IEAGHG Weyburn-Midale CO2 Project

Huerta et al. 2013 Exp. evidence for self-limiting reactive flow thru a fractured Lab testing
cement core: Implications for time-dep wellbore leakage
Jung et al. 2013 Imaging wellbore cement degradation by CO2 under geo. Lab testing
sequestration conditions using X-ray computed
microtomography
Jung et al. 2013 Experimental study of potential wellbore cement carbonation Lab testing
by various phases of carbon dioxide during geologic seq
Luguot et al. 2013 Hydro-dynamically controlled alteration of fractured Portland Modeling
cements flowed by CO2-rich brine study
Mason et al. 2013 Chemical and mechanical properties of wellbore cement Modeling
altered by CO2-rich brine using a multi-analytical approach study
Newell & Carey 2013 Experimental evaluation of wellbore integrity along the Lab testing
cement-rock boundary
Walsh et al. 2013 Permeability of wellbore-cement fractures following Lab testing
degradation by carbonated brine
Wenning et al. 2013 Reactive flow channelization in fractured cement- Modeling
implications for wellbore integrity study
Duguid et al. 2014 Well integrity assessment of a 68 year old well at a CO2 Field study
injection project.
Nygaard et al. 2014 Effect of dynamic loading on wellbore leakage for the Modeling
Wabamun Area CO2 Sequestration Project study
Sminchak et al. 2014 Investigation of wellbore integrity factors in historical oil and Data survey
gas wells for CO2 geosequestration in the Midwestern U.S.
Glazewski et al. 2015 Wellbore Evaluation of the basal Cambrian System Data survey
Haagsma et al. 2015 Utilizing Cement Bond Logs to Evaluate Wellbore Integrity Data survey
for CO2 Storage
Jordan et al. 2015 A response surface model to predict CO2 and brine leakage Modeling
along cemented wellbores study
Moody & Dotson 2015 SCP Diagnosis Using the Wellhead Model Field study
Zhang et al. 2015 Wellbore cement integrity under geologic carbon storage Lab testing
conditions
Brunet et al. 2016 Fracture opening or self-sealing: Critical residence time as a Modeling
unifying parameter for cement-CO2-brine interactions study
Brunet et al. 2016 Cement fracture opening or self-sealing: critical residence Modeling
time unifies observations under different conditions study
Huerta et al. 2016 Reactive transport of CO2-saturated water in a cement Modeling,
fracture: Application to wellbore leakage during geo. CO2 lab tests
storage
Wolterbeek et al. 2016 Reactive transport of CO2-rich fluids in simulated wellbore Lab testing
interfaces: Flowthrough experiments on the 1− to 6-meter
(m) length scale
Wolterbeek et al. 2016 Effect of CO2- induced reactions on the mechanical Lab testing
behavior of fractured wellbore cement
Carroll et al. 2017 Influence of Chemical, Mechanical, and Transport Data survey
Processes on Wellbore Leakage from Geologic CO2
Reservoirs
Iyer et al. 2017 Incorporating reaction-rate dependence in reaction-front Modeling
models of wellbore-cement/carbonated-brine systems study

The research for this project integrates direct SCP testing of wells along with examination of defects and
geochemical reactions along the length of the borehole, providing a more holistic examination of the
effects of CO2 on wellbore integrity. Many studies have verified that CO2 brine mixtures may dissolve
oilfield cements and precipitate CaCO3. However, it is difficult to examine this process in situ along the
borehole length, since the wells are cemented in place 3,000 to 15,000 feet (ft) deep in the subsurface. For
example, CO2 may mix with brine in the reservoir zone and dissolve cement, but then precipitate CaCO3
in overlying caprock zones. Or, certain minerals in carbonate rock layers may buffer CO2 brine mixtures,
reducing potential for CO2 to corrode well materials. Consequently, it is useful to evaluate the interaction
of CO2 with fluids, pressure conditions, geological layers, cement blends, and well construction materials.
1.4 Summary of Appalachian Basin, Michigan Basin, and Williston Basin Field
Sites
A key part of the project was testing and analysis at three field test sites. The three field test sites
identified for wellhead SCP testing were located in the Appalachian, Michigan, and Williston
sedimentary basins. These sites have wells exposed to CO2 at depths of 1,000 to 7,000 ft and 5 to 50+
years of age (Table 1-3). Therefore, they provided an excellent opportunity to examine CO2 storage
wellbore integrity at field sites. The wells were surveyed for indications of SCP, and a subsample of wells
were tested for wellhead casing pressure buildup. The field test sites were characterized for geologic
setting, field history, well construction specifications, and hydrologic conditions. Field site characteristics
are summarized in Table 1-3 and are discussed in the following subsections.
Table 1-3. Summary of field sites.

Parameter Appalachian Basin Michigan Basin Williston Basin
Field area (acres) 30,000 3,000 45,000
Reservoir depth (ft) 6,200-7,000 1,000 & 6,000 5,000
Reservoir type Sandstone Carbonate reefs Carbonate
Caprock Shale/carbonate Evaporite Evaporite
CO2 type Natural gas & CO2 CO2 EOR CO2 WAG EOR
Temperature (°F) 140 105 145
Discovery pressure (psi) 2,900 3,000 2,000
Discovery year 1973 1960 1954
# Wells 58 ~45 ~3,000
Note: WAG = water alternating gas; EOR = enhanced oil recovery; psi = pounds per square inch.
Appalachian Basin Site
The Appalachian Basin Indian Creek site is a natural CO2 and methane field located in Kanawha County,
West Virginia (Figure 1-2). The field contains approximately 58 wells at total depths between 6,200 ft
and 6,700 ft. The Indian Creek field produces in the Tuscarora sandstone, where the percent of CO2 in
some wells ranged from 44% to 83% and nitrogen ranged from 13.9% to 35% (Avary, 1996). A
completion report for the discovery well, API number 4703901684, listed CO2 at 65%. Two scout cards,
for wells 4703902718 and 4703902719, both listed CO2 at 60%. Hamak & Sigler (1991) and Hamak &
Gage (1992) reported produced gas with an average CO2 at 65.8%, 305 British thermal units per cubic
foot, and 1.214 grams per centimeter average gas gravity. One gas sample in the field taken by Jenden et
al. (1993) reported a CO2 content of over 61%. At depth, pressure conditions in the Indian Creek field
were likely >2,800 pounds per square inch (psi) and temperatures were >110 °F. Thus, the CO2 would be
in supercritical state.

Indian Creek Field
Figure 1-2. Location of the Indian Creek field in Kanawha County, West Virginia.
Michigan Basin Site
The Michigan Basin study site is located in the northern portion of the Niagaran reef trend in Otsego
County, Michigan (Figure 1-3). The fields have been developed since the 1960s in the region, and
selected reefs have been subject to CO2 enhanced oil recovery (EOR) since the 1990s. There are several
hundred Antrim gas wells at depths between 1,000 ft and 1,500 ft, and there are dozens of Niagaran reef
wells at depths ranging from 5,000 ft to 7,000 ft in Otsego County. The Antrim shale wells produce
methane and CO2. The CO2 volume in the produced gas is between 5% and 30%. Some of the Niagaran
reefs in the area have been subject to CO2 EOR over the past 10 to 20 years. Overall, the Antrim shale
wells and Niagaran reef CO2 EOR wells are attractive candidates for well integrity analysis.
Williston Basin Site
The Williston Basin is a large sedimentary basin which spans the southern portion of Saskatchewan,
Canada, and the north-central United States. The Weyburn field is located on the northwestern edge of the
Williston Basin geologic feature and was a major oil play which is now used for CO2 EOR (Figure 1-4).
The main reservoir for the Weyburn oil field consists of the Marly Midale and Vuggy Midale beds, which
are part of the Mississippian-aged Madison group (Wilson & Monea, 2004). Approximately 3,000 wells
are located in the Williston Basin testing site. The testing site is a mature oil field that began production in
1954. Prior to the start of CO2 EOR, the Williston Basin testing site produced roughly 340 million barrels
of oil, or an estimated 25% of the field’s total reserve. As such, conventional methods for oil production
are no longer viable for profit. To keep the oil field active, CO2 EOR was started at the Weyburn field in
2000 using CO2 transported via pipeline from a gasification synfuels plant in North Dakota.

Figure 1-3. Structure map of the Brown Niagaran showing the pinnacle reef trend (gray),
Antrim shale trend (red), and the Michigan Basin site (star).
Figure 1-4. Location of the Williston Basin field site.

Chapter 2.0 Well Integrity Registry
Well integrity issues may arise from the materials, construction, operations, and subsurface conditions
specific to a well. The objective of this task was to develop a wellbore integrity registry that describes
actual or potential well integrity issues that may affect carbon storage projects. The registry was based on
existing research and on experience related to well construction methods, well casing integrity issues,
well cement issues, and geologic processes for CO2 environments.
2.1 Well Construction Methods

Historical and current well construction methods for deep wells are important components to
understanding well integrity. Important factors can include hole conditions, tubular design, cement slurry
properties, hydrated cement properties, geologic conditions, and operational conditions. The cemented
annulus, the cement space between the casing and formation, represents most of the leakage risk
associated with the well. Leakage pathways through the cement matrix and many leakage pathways
around the cement are described in this section.
Production wells, injection wells, and other deep wells generally consist of a conductor casing, a surface
casing, a production casing, and intermediate casing strings in unstable hole conditions. A liner, a string
of casing that does not extend to the surface, may also be used to case the bottom section of the well.
Most casing used in wells is metal, but fiberglass, coatings, and alloys are sometimes used. The size and
strength of casing in a well are determined by the production rate and loads on the casing in the well.
Most wells are constructed using mild carbon steel, but corrosion-resistant alloys are used in corrosive
environments. Primary cement is placed in the annulus between the casing and formation or the annulus
between the inner and outer casing while the well is being constructed.
Well repair failures and defects within the well architecture may lead to well integrity issues that can
cause gas and/or fluid migration along the borehole. Remedial cement may be emplaced with a squeeze
job to address well defects. Cement plugs are used to close off portions of a well or to abandon the well.
Plugs are generally set across the perforated zone, across resource zones, at casing seats, and at the top of
the well. Multiple methods may be used to set cement plugs in a well, including a balanced plug method,
a dump bailer method, and a two-plug method.
Multiple researchers, including Gasda et al. (2004), Duguid et al. (2012), and Carroll et al. (2016), have
characterized migration pathways. In general, this research concludes that the main migration pathway
occurs through or around the cement matrix that makes up the primary and plug cement in a well
(Figure 2-1). Leakage through the cement matrix may include flow through the cement matrix, degraded
cement, cement-casing interface, formation-cement interface, mud channels, and fractures/cracks in the
cement sheath, and as well as through casing defects in the open wellbore. The results of the Duguid et al.
(2012) study indicate that higher-permeability flow pathways around the cement matrix are the major
pathway for gas migration.

Figure 2-1. Diagram illustrating possible leakage pathways within a well.
2.2 Well Casing Integrity

Casing is employed to keep the wellbore from closing after it is drilled and allow a pathway for access to
the reservoir. Casing represents a barrier between the cemented annulus and the open inside of the
wellbore. At carbon storage projects, the casing may be exposed to CO2 and formation fluids in injection
wells and legacy oil and gas wells that penetrate the storage zone. Exposure to operational and geologic
factors, including temperature and pressure cycles, carbonated brines, chlorides, and hydrogen sulfide,
can lead to corrosion, wear, leaky collars, and other casing issues. Holes and cracks in the casing can
allow fluids to enter the well and use the wellbore as a leakage pathway. Debonding of the casing from
the primary cement can lead to microannuli, allowing leakage between the cement and casing. Well
casing failure mechanisms may include the following processes:
Thermomechanical cycling. Thermomechanical cycling due to production or injection can cause

well casing to debond from the cement surrounding it. Differences in thermal expansion and
engineering properties cause the cement and steel to expand and contract to different degrees when
exposed to the same conditions, causing the bond between the cement and steel to break and leading
to a microannulus.
Wear. Physical wear to the casing can occur when the casing is run into the well after drilling, when
tools are run in the casing during workover operations, or when the production tubing and the
production casing rub together. Wear can weaken the casing, making the casing more likely to burst
or collapse, and leading to either a casing breach or separation of the casing from the cement.
Corrosion. Corrosion needs to be considered when designing new wells or when repurposing
existing wells for CCUS. Corrosion of the casing can create holes, allowing CO2 being injected into
the formation to enter the wellbore and migrate from the reservoir to the wellhead or into overlying
formations. This process could then lead to SCP. CO2-saturated fluids are corrosive to mild steel and
can cause a failure in the wellbore integrity. Casing steel in contact with supercritical CO2 was found
to corrode at a rate of 20 millimeters (mm) per year (Carey, 2013). A Plains CO2 Reduction
Partnership study (PCOR, 2014) also found corrosion of steel casing, but lower corrosion in J55 and
N80 grade steels.

Corrosion, cracks, and leaking connections can be inferred from the existence of SCP or leaks near or at
the surface. Dead vegetation around a well can indicate a leak. Gas bubbles leaking from an abandoned
well head are a strong indication of a casing integrity problem (Carey, 2013). Corrosion can be detected
using corrosion logging tools to monitor the condition of the inside and outside of the casing. Several
different types of tools exist, including multi-finger caliper tools that measure pitting and defects on the
inside of the casing, ultrasonic tools that measure the location of the inside and outside of the casing,
electromagnetic tools that measure the amount of metal in the casing, and magnetic flux leakage tools that
measure inner and outer corrosion.
2.3 Well Cement Integrity

Well cements are typically Portland cements similar to those used in the construction industry. Well
cements are made by combining calcareous materials (such as limestone) and argillaceous materials (such
as shale) and heating them to create clinker, then grinding the cooled clinker with calcium sulfate. Many
research projects have focused on using historical well data, changes in economy, and regulatory changes
to predict or assess the risk of leakage in CO2 storage fields; however, these studies have not integrated
field testing with data analysis.
The addition of water causes a hydration reaction that creates calcium-silicate-hydrate (C-S-H), calcium
hydroxide (CH), and other phases that make up set cement. Neat Portland is stable at high pH values
owing to the hydroxide phases that are byproducts of hydration. Portland cement is susceptible to
carbonation and acid attack at lower pHs, including those common to CO2 storage conditions (pH 2 to 5).
Exposure to CO2 and carbonated brine will lead to carbonation and, under the right conditions, further
reaction and degradation of the carbonate and cement minerals. In brine or water, the first reaction is CO2
dissociation (Equation 1):
CO2 + H2O ↔ H2CO3* ↔ H+ + HCO3- ↔ 2H+ + CO32- [Equation 1]
The carbonate species from Equation 1 interact with the cement, C-S-H, and CH to create CaCO3
(Equations 2 through 5):
Ca(OH)2(s) + 2H+ + CO32- → CaCO3(s) + 2H2O [Equation 2]
Ca3Si2O7H•4H2O(s) + 2H+ + CO32- → CaCO3(s) + SiOxOHx(s) [Equation 3]
Ca(OH)2(s) + H+ + HCO3- → CaCO3(s) + 2H2O [Equation 4]
Ca3Si2O7H•4H2O(s) + H + HCO3 → CaCO3(s) + SiOxOHx(s)
+ -
[Equation 5]
For cement exposed to wet supercritical CO2 and carbonated brine that is not refreshed often (diffusion
conditions with a small volume of brine as compared to the volume of cement), the reaction with CO2
stops with CaCO3 being created within the cement pores. However, if cement is exposed to flowing
carbonated brine or the carbonated brine is refreshed often, the carbonate that formed in the pores can be
reacted away (Equations 6 and 7), leaving a soft silicate-hydrate gel (SiOxOHx in Equations 3 and 5).
CO2 + H2O + CaCO3(s) ↔ Ca2+ + 2HCO3- [Equation 6]
2H+ + CaCO3(s) ↔ CO2 + Ca2+ + H2O [Equation 7]

Cement carbonation has been identified in both the laboratory and field setting. Kutchko et al. (2007),
Duguid and Scherer (2010), and others have conducted experiments on cements under batch and flowing
conditions. Carbonic acid can weaken wellbore integrity when it comes in contact with Portland cement.
Portland cement is an alkaline substance (pH greater than 12.5) which is incompatible with CO2 fluids
(pH less than 6) (Carey, 2013). An experimental study conducted by Kutchko et al. (2007) showed three
distinct zones around the cement following a contact with CO2 saturated fluids. Zone 1 showed an
increase in porosity and a depletion of the Portlandite (Ca(OH)2) in the cement. Zone 2 exhibited
decreased porosity and was dominated by CaCO3 polymorphs. Zone 3 was completely leached of CaCO3,
leaving an amorphous silica layer. Barring mechanical failure of the cement or casing, the decreased
porosity of zone 2 acts as a protective barrier slowing further degradation of the cement (Duguid et al.,
2011). The primary geochemical threat to cement exists along the interfaces (Carey, 2013).
The reactions represented in Equations 2 through 7 were visible in multiple fronts moving from the
exposed edge of the cement toward the center of the samples. In samples exposed to flowing carbonated
brine, moving from the outside to the inside of the samples, Duguid et al. saw two zones totally depleted
of calcium, indicating that the reaction had progressed to the point of leaving the soft silicate-hydrate gel
mentioned above (Figure 2-2). Inside the CaCO3-enriched zone is a zone depleted of CH and then an
unreacted zone in the center of the sample. Cement degradation can lead to large increases in
permeability.
Figure 2-2. Photos showing the results of well cement exposed to flowing carbonated brine
at 50°C and pH 3. Left photo shows the outer orange and brown silicate-hydrate zone
outside of a CH-depleted cement zone, with an unreacted cement in the center.
Right photo shows the outer orange and brown silicate-hydrate zones, the
white CaCO3-rich zone, and a CH-depleted cement zone in the center.
Multiple field investigations have been conducted on wells exposed to CO2 or generally on well
construction as it relates to carbon capture and storage (CCS) leakage risk. Carey et al. (2007), Crow et al.
(2010), and Duguid et al. (2014) have all studied wells to identify well integrity defects and identify
changes brought on by exposure to CO2. Carey et al., Crow et al., and Duguid et al. all found evidence of
carbonation in cements in and above the CO2 reservoir.
• Carey et al. (2007) looked at a 50-year-old well that was exposed to CO2 for 30 years in the
SACROC oil field in Scurry County, Texas. The cement was collected at the surface during a
sidetracking operation. The cement in the well was a neat Portland Type 1. The authors saw
carbonation and discoloration of the samples and measured a permeability of an air-dried sample
of 0.1 millidarcy.

• Crow et al. (2010) studied a 30-year-old natural CO2 producer in southern Colorado. The authors
analyzed seven sidewall core samples collected from the well. The well cement was a Portland
cement with fly ash and bentonite additives. The permeability ranged between 0.3 and
32 microdarcys, with the highest values of 5, 27, and 32 microdarcys measured in samples
collected in or adjacent to the CO2 zone.
• Duguid et al. (2014) studied a 68-year-old well that was a producer, was plugged and abandoned,
and then was converted to a CO2 storage test monitoring well and exposed to CO2 for 5 years. Two
whole sidewall core samples were collected about 400 ft above the CO2 zone. Both samples
showed incomplete carbonation with both CaCO3 and CH being present in the samples. The upper
sample was likely squeeze cement and still contained unhydrated cement phases. The sample
collected 4.5 ft lower was likely primary cement with heavy alteration.
The integrity of the cement in wells with CCUS applications can be damaged by exposure to CO2 and
carbonated brines or as the result of physical processes during construction, operation, and abandonment.
Defects in the wells may be brought about by physical processes such as microannuli, gas contamination,
mud contamination, and cracking. Many of these defects can result in SCP, defined as the persistent
buildup of pressure over time by an intake of fluid into a well’s annulus. Early-stage SCP is related to
well completion methods; for example, poor casing centralization may cause mud displacement, leaving it
on the borehole wall. Long-term SCP is typically attributed to temperature cycling causing the casing to
expand and contract; this process subsequently leads to a detachment between the elastic steel and brittle
cement interface (Huerta et al., 2009). A variety of processes may affect cement integrity during initial
well construction and further well life cycle events. These processes may lead to the following leakage
pathways:
Microannulus. A microannulus is a small gap between the casing and primary cement. It occurs
when the cement and casing debond or the bond is never established. A microannulus can extend part
or all of the way around the casing (circumference) and can act as a leakage path. A microannulus can
be created at the time of well construction or after construction, during operations or workover.
Cracking. Cracking of the cement sheath can be caused by thermomechanical cycling of the casing,
shrinking of the primary cement, or geomechanical forces on the well. The removal of overburden in
or near river valleys can change geomechanical forces on the well by reducing the lithostatic pressure
that controls flow to the surface.
Eccentering. Eccentering of casing occurs when the casing is not centered in the borehole. In severe
cases, the casing may be in contact with the borehole wall. Centralizers can be employed in the casing
as it is run into the borehole to reduce eccentering. Eccentered casing can lead to poor mud cleanout
and/or poor cement placement on the narrow side of the hole.
Mud contamination and mud channels. During the drilling process, thick muds such as oil-based
muds can build up on the borehole wall, creating a mudcake. If the well is not thoroughly cleaned out,
mud can keep the cement from contacting the borehole wall, keeping a cement-to-formation bond
from forming. The interface between the cement and mud and between the mud and formation can act
as a leakage path. Mud channels can also form when cement slurry fingers through the mud annulus
or eccentered casing.
Fluid/gas invasion. Invasion of fluids or gas during hydration can damage cement’s isolation
capacity. When cement is pumped, it acts as a liquid with a hydrostatic head. As cement hydrates, its
ability to provide overbalanced pressure against the surrounding formation can be lost, allowing fluid
to enter the cement and creating connected pathways, sometimes referred to as “gas-cut” cement.
Well integrity problems in cement can be detected through testing, monitoring, logging, or a combination.
Successful detection and mitigation procedures are important to test and prepare a wellbore for CO2
storage. Common detection methods include casing inspection logs, temperature logs, mechanical

integrity tests, radioactive tracer surveys, and SCP measurements. Recorded successful mitigation
processes include using pressure-activated sealant, cement squeezing techniques, swelling technologies,
and self-healing cements; controlling pH levels; and using plug replacements and chemical
enhancements.
2.4 Geologic Processes
Geologic processes can greatly influence the condition of a wellbore and create potential fluid migration
pathways. These processes include formation lithology, influence of CO2 on wellbore cement and
surrounding formations, ambient conditions, lost circulation zones, geomechanical stresses, and
geochemical environments. Monitoring and mitigation processes can detect and fix problems caused by
geologic processes.
Lithology can influence the integrity of a wellbore based on the condition of the wellbore and the bond
between the wellbore and cement. Borehole breakout and drilling-induced fractures are common during
drilling. Borehole breakout is the enlargement of a borehole due to the removal of more material from an
interval than from the overlying and underlying intervals. This can cause an increase in the stress in
adjacent rocks (Bell and Gough, 1979). Drilling-induced fractures are fractures created around a borehole
which are parallel to the direction of drilling (Aadnoy, 1990). These fractures are created when the
induced stress of the drilling exceeds the maximum stress of the formation. Borehole breakout and
induced fractures could both lead to potential fluid migration pathways along the borehole and caprocks.
These occur most frequently in brittle and soft rocks.
The introduction of CO2 into a reservoir changes the local stress field. This geomechanical shift in
pressure affects both the caprock and the wells penetrating the target formation. The changing pressures
within the reservoir could lead to cement or casing failure, resulting in fractures or microannuli (Zhang
and Bachu, 2011). Radial, axial, and shear deformations can cause the cement and casing to debond or
crack. Radial deformation is caused by an increase in temperature and pressure, as well as external
pressures from viscous movement of the surrounding rock. Axial deformation is caused by compaction
during production and subsequent expansion during injection. Due to the brittle nature of cement, this
deformation may cause the cement-casing bond to fail. Shear deformation is caused by the presence of
local faults and fractures; these can cause the casing to shear and the cement to fracture (Orlic, 2009).
This stress is particularly damaging in deformable lithologies such as salt and shale because stress is more
readily transmitted to the well.
2.5 Key Findings of Well Integrity Registry
Many research projects have focused on using historical well data, changes in economy, and regulatory
changes to predict or assess the risk of leakage in CO2 storage fields; however, those studies have not
always identified the well integrity issues that can lead to a leak. The wellbore integrity registry presented
in Table 2-1 identifies the well component, integrity issues, causes, timing, and leakage pathways that
may occur in wells. Most wellbore integrity problems are located in the casing, cement, or interface
between the two components (Figure 2-3). Other problems arise due to geological processes such as
formation lithology and geomechanical stresses.

Table 2-1. Wellbore integrity registry of identified integrity issues.
Well Component Integrity Issue Description Causes When Leakage Pathway

Thermo- Contraction and Differences Construction, Debonding along
mechanical expansion of between properties operation, cement interface
cycling well casing of materials workover, (microannulus)
abandonment
Wear Wear to the Casing interactions After drilling, Burst, collapse,
Casing
casing with wellbore and during holes in casing

tools workovers
Corrosion Corrosion of Contact with Construction, Holes in casing,
casing corrosive fluids operation, cracking
saturated with CO2 workover,
abandonment
Degradation Dissolution or Contact with Construction, Pores in cement or
alteration of corrosive fluids operation, along degraded
cement saturated with CO2 workover, cement at
abandonment interfaces
Microannulus A small gap Casing and cement Construction, Along casing-
and cracking between casing debond, or bond operation, cement interface
and cement and was never workover,
cracks in the established or was abandonment
cement broken
Mud Poor mud Poor cement job During Along interfaces or
Cement
contamination removal before design, poor hole construction through bulk

cementing cleanout cement
Eccentering Casing is not Poor centralization During Along casing,
centered in the construction cement, or mud
borehole interfaces
Mud channels Cement slurry Poor cement job During Along mud channel
fingers through design construction interface or
the mud in the through flowing
annulus mud
Fluid invasion Invasion of fluids Poor cement slurry During Poor zonal
into cement design and loss of construction isolation
hydrostatic pressure
Formation Borehole Induced stress During drilling Poor cement bond
Borehole wall (Geologic
lithology breakout and greater than to borehole wall

drilling induced maximum of the
Processes)
fractures formation stress

Geomechanical Changes in Pressure gradient Construction, Cement and
stresses stress field changes and creep operation, casing damage or
workover, failure
abandonment

Figure 2-3. Wellbore schematic showing types of wellbore integrity issues
and where they occur (Sminchak et al., 2016).

Chapter 3.0 Well Record Data Collection and Review
Task 3 was divided into three subtasks to compile available information on existing wells at the field test
sites in Otsego County, Michigan (Michigan Basin site), Kanawha County, West Virginia (Appalachian
Basin site), and Saskatchewan, Canada (Williston Basin site). The well data collection task was
completed in June 2017. Information was compiled and evaluated for well cementing/drilling, operation,
and well workover records. This dataset included over 1,000 items related to wellbore construction in the
three study sites. The information from the well collection task will be used to evaluate the overall
condition of boreholes in the study areas and as input for future tasks (Figure 3-1).
Oil and gas records were acquired from publicly available resources such as the Michigan Department of
Environmental Quality (DEQ), the West Virginia Geologic Survey (WVGES), and the Government of
Saskatchewan. The primary focus was collecting data on active and producing wells. Operational and
workover records were also acquired from the same sources. Additional, proprietary information was
shared by select operators.
Well Cementing and Well Workover/Leakage

Operational Records
Drilling Records
•Location •Producing Formation •Date
•Age •Production History •Reason
•Depth •Pressure •Type of workover
•Type
•Status
•Construction
•Treatments
•Perforations
•Cement quantity
•Cement bond logs
Figure 3-1. Summary of well record collection parameters.
3.1 Field Site Descriptions

The field sites were selected, because they have wells exposed to CO2 at depths of 1,000 to 7,000 ft and
5 to 50+ years of age (Table 3-1). Therefore, they provided an excellent opportunity to examine CO2
storage wellbore integrity at field sites. The wells were surveyed for indications of SCP, and a subsample
of wells were tested for wellhead casing pressure buildup. The field test sites were characterized for
geologic setting, field history, well construction specifications, and hydrologic conditions.

Table 3-1. Summary of field sites.
Parameter Appalachian Basin Michigan Basin Williston Basin
Field area (acres) 30,000 3,000 45,000
(across multiple reefs)
Reservoir depth (ft) 6,000-7,000 1,000 & 6,000 5,000
Reservoir type Sandstone Carbonate reefs Carbonate
Caprock Shale/carbonate Evaporite Evaporite
CO2 type Natural gas & CO2 CO2 EOR CO2 WAG EOR
Temperature (°F) 140 105 145
Discovery pressure (psi) 2,900 3,000 2,000
Discovery year 1973 1960 1954
# Wells 58 ~45 ~3,000
Note: WAG = water alternating gas; EOR = enhanced oil recovery; psi = pounds per square inch.
3.2 Michigan Basin Site

A significant number of oil and gas wells were drilled at the Michigan Basin site in Otsego County,
Michigan. Records for 1,204 Antrim formation wells and for 418 Niagaran wells were reviewed for well
construction and cementing details. The earliest Antrim well in the study area was drilled and completed
in the 1950s; however, very few Antrim wells were drilled until the 1980s, when the Antrim play became
more economic. Since the 1980s, the number of Antrim wells drilled per year has been reduced by
approximately two-thirds. Within the study area, the Antrim play is relatively shallow, with the deepest
well reaching a depth of 2,250 ft. The Niagaran play experienced a boom in the 1970s; nearly half of the
Niagaran wells were installed in that decade. The number of Niagaran wells installed has dropped
significantly since the 1980s; currently, only a small fraction of Niagaran wells are installed annually
compared with the numbers completed in the 1970s and 1980s. The Niagaran wells vary more in depth
due to the influence of the basin, with a range of 5,050 ft to 6,700 ft.
The majority of Antrim wells (94%) were drilled as gas wells, while 5% were drilled for brine disposal.
The remaining 1% were dry holes. Michigan well records through 2016 show that 90% of the wells are
still producing methane and 4% have been plugged and abandoned. Records also show that 324 wells
listed with a terminated permit were never drilled.
Niagaran wells were drilled for a wide range of purposes. Nearly 60% of Niagaran wells were dry holes
because of the isolated, compartmentalized reef structures and the variable lithology within the reefs.
However, the percentage of dry holes drilled has decreased over time as a result of improved
characterization techniques and a better understanding of the reef structure and composition. Wells that
produced were mostly oil, with some recorded as gas wells. A small percentage of Niagaran wells were
used for brine disposal and injection. As of 2016, 79% of the Niagaran wells were plugged and
abandoned; only 18% are currently active and/or producing. The remaining 3% were temporarily
abandoned or shut in.
Due to the large number of Antrim wells in the general area, a subset of wells was randomly selected for
more detailed analysis. Antrim wells were consistently constructed with three casing strings: conductor,
surface, and production. The conductor casing was 13.375 inches in diameter in nearly half of the wells,
while 30% of wells used larger-diameter casing and 24% used smaller-diameter casing. The surface
casing was set at the base of the glacial drift (550 to 1,150 ft) and generally used a casing diameter of
8.625 or 7 inches. Approximately 3% had an intermediate casing diameter of 11.75 inches. The
production casing was often 5.5 inches or 4.5 inches in diameter.

The Niagaran wells were consistently constructed with four casing strings in the general area. The
conductor casing was typically 16 inches in diameter, with 15% of wells having smaller or larger
diameters. The total depth of the conductor casing was set between 38 and 147 ft deep within the glacial
drift. The surface casing string was predominantly 13.375 inches in diameter and was set at the base of
the glacial drift (474 to 1,142 ft). Next, the intermediate casing was mostly 8.625 inches in diameter, with
a few wells having larger or smaller diameters. The casing was typically run to the Bass Islands or within
the Salina group (~3,400 ft deep). Finally, the deep casing string was typically 5.5 inches in diameter and
was set in the Niagaran with some wells set higher in an open hole or barefoot completion. Figure 3-2
illustrates common well construction used for the Antrim and Niagaran wells.
Figure 3-2. Typical Antrim (left) and Niagaran (right) well construction.
CBLs were available for 87 wells in the Michigan Basin field site area. The types of cement were not
typically entered in well records. The surface casing in Antrim wells was mostly driven (95%) with only
5% having recorded using cement. The amount of cement used ranged from 350 to 460 sacks. The cement
used in the intermediate casing ranged from 160 to 700 sacks of cement, with an average of 360 sacks.
Some wells did not have cement data available but were recorded to have had cement circulate to the
surface. The deep casing had recorded cement ranging from 55 to 660 sacks with an average of 243 sacks.
Occasionally, wells were recorded to have had cement circulate to the surface.
The conductor casing in Niagaran wells was mostly driven (70%) with 18% having recorded using
cement. The amount of cement ranged from 100 to 200 sacks. The remaining 12% of wells did not have
recorded cement. The cement used in the second surface casing string ranged from 200 to 1,130 sacks,
with an average of around 600 sacks. A subset of wells (9%) did not have a second casing string. The

cement used in the intermediate casing ranged from 200 to 950 sacks with an average of 406 sacks. The
cement used in the deep casing string ranged from 75 to 950 sacks with an average of 484 sacks. A deep
casing string was not recorded for 9% of wells.
Production data were available for 2,541 Antrim wells and 414 Niagaran wells at the Michigan Basin site,
which included monthly production of oil, gas, and water since 1982. Cumulative gas production by well
ranged from 2,355 thousand cubic feet (MCF) to 31 million MCF for Antrim wells. On average, a well
had a cumulative gas production of 5.1 million MCF. Wells experienced different ranges of production
time and many starts and stops. Gas production records showed that wells produced from 1 month up to
28 years, with an average lifespan of 15 years. Production was greatest in the southeastern corner of the
general study area and lowest toward the west. Water cut and CO2 content increased throughout the
lifespan of the average well; the more gas produced, the more water and CO2 were produced. Water
production ranged from 0 to 2 million barrels with an average of 370,000 barrels. CO2 content ranged
from 2%-30%.
Over 110 million barrels of oil have been produced from Niagaran wells at the Michigan Basin site since
1982. The cumulative oil production by well ranged from 0 barrels to 5.2 million barrels with an average
of 270,000 barrels. There are several individual production fields which are composed of single to
multiple reefs. The producing interval and length varied greatly by well and by field. The wells had
recorded production from 0 months to 26 years with an average of 10 years. Many wells also experienced
periods of stopped production after primary production, and then periods of production during secondary
and tertiary recovery periods. Production of oil was greatest along the northern trend of reefs and during
the early 1980s. Monthly production steadily declined into 2016 (Figure 3-3).
Figure 3-3. Monthly and cumulative oil production from Niagaran wells showing
greatest monthly production in the early 1980s and a steady monthly decline through 2016.

Over 380 million MCF of gas have been produced from Niagaran wells at the Michigan Basin site since
1982. The cumulative gas production by well ranged from 0 MCF to 10 million MCF with an average of
930,000 MCF. Niagaran wells had incremental periods of gas production since 1982 with varying
production lengths. The recorded production ranged from 0 to 26 years with an average of 10 years,
following the oil production trend. Gas production remained high through the 1980s and began tapering
off in the 1990s, with lowest production from 2005 to 2016. The greatest gas production occurred in the
largest fields without a strong trend of direction. Water production was also recorded for the Niagaran
wells. The produced water ranged from 0 to 2.5 million barrels with an average of 150,000 barrels per
well. The greatest water production was recorded in the late 1990s through early 2000s, which tapered off
toward 2016. No production was recorded prior to 1992. The greatest water production occurred in the
center of the reef trend, where some wells penetrated the oil-water contact.
As part of this study, records for 57 Antrim wells and 54 Niagaran wells were reviewed for post-
completion workovers (casing replacement, tie-back strings, etc.).
Antrim Wells. In general, very few “true” workovers were performed on the Antrim wells following
initial completion; however, perforations were frequently added to access more productive zones in the
Antrim and improve well productivity. Of the 57 wells investigated, perforations were commonly added
to the wells. Seven wells were deepened, typically by underboring the well beneath the existing casing;
the bottom of the well was left as open hole. Cast-iron bridge plugs were added to two of the wells to seal
off deeper (likely unproductive) portions of the well. Any time the wells were reperforated, the wells were
also refractured with a frac fluid/foam and a proppant, and acidized after the perforations were complete.
Niagaran Wells. Similar to the Antrim wells, relatively few workovers have been performed on the wells
completed in the Niagaran reefs. Of the 54 wells investigated, only three were part of a workover, which
involved running additional casing/line (Well 21137299580000), running a tie-back string (Well
21137578160000), or performing a cement squeeze job for a casing leak (Well 21137290910000). Eight
wells were extended through a kickoff or had a portion of the well temporarily abandoned with a bridge
plug. However, slightly more than 60% of the wells had additional perforations to access more productive
zones. Typically, acid was pumped into the wells when new perforations were added.
3.3 Appalachian Basin Site

The earliest Indian Creek well was drilled in the 1940s. Drilling activity remained low until the 1970s and
1980s, when drilling increased substantially. Activity tapered off after the 1980s, when only an additional
eight wells were drilled. Indian Creek wells primary targeted the Tuscarora sandstone, with a few wells
producing from nearby sandstones (Clinton, Oriskany). The resulting depth of the wells was mostly
between 6,300 and 7,300 ft; however, one well was drilled to 8,075 feet. Indian Creek wells produced
natural gas; 85% were recorded as gas wells while the remaining 15% were dry holes. As of 2016, 20% of
the wells were recorded as plugged and abandoned with 80% still operational.
The wells in the Indian Creek field were commonly constructed with four to five casing strings. The
conductor casing diameter was mostly 13.375 or 20 inches in diameter, with 30% of wells having
recorded smaller or larger sizes. The depth of the conductor casing ranged from 18 to 930 feet. The
conductor casing depths averaged 183 feet, placing the casing in the undifferentiated
Pennsylvanian/Mississippian strata. The second surface casing string was typically 9.625 or 13.375 inches
in diameter. The depth of the surface casing ranged from 77 to 2,670 feet, placing it across several shale
formations that could produce gas. The intermediate casing was either 7 or 9.625 inches in diameter and
was run through the Devonian shale or Helderberg formations. When a fourth, or second intermediate,
casing string was used, it was 7 inches in diameter and set in the Helderberg through the Tuscarora. The
deepest casing string, or production string, was 4.5 inches in diameter and set in the Tuscarora. A typical
wellbore diagram is illustrated in Figure 3-4.

Figure 3-4. Wellbore diagram of an Indian Creek well showing common well construction.
In addition to construction data, information on perforations, treatments, and treatment volumes were
recorded. These data provide information on well stimulations and additional stresses applied to the
wellbores. Only 22% of the wells were recorded to have perforations, while many wells were completed
open hole. Nearly 60% of the wells were treated with acid and fractured. The volumes and types of
materials used for treatments varied greatly by well.
CBLs were available for 15 wells in the Indian Creek field. The CBL analyses were conducted, and are
summarized, under Task 4. Results showed moderate thickness of good quality cement with frequent
intervals of poor cement (Figure 3-5). Additionally, caliper logs were collected to be used in Task 4 with
CBLs and cement sacks to determine the volume and thickness of a cement column. The number of sacks
of cement was recorded for each casing string; cement type was not readily available. Eight wells did not
have cement data available. The conductor casing was set with 10 to 826 sacks of cement with an average
of 183 sacks of cement. Wells were frequently recorded to have had cement circulate to the surface. The
surface casing string used 150 to 1,121 sacks of cement with an average of 548 sacks. It was also
commonly recorded to have been cemented to the surface. The intermediate casing used 148 to 1,226
sacks of cement with an average of 544 sacks. The fourth casing and deep casing strings used similar
amounts of cement, ranging from 75 to 280 sacks.

Figure 3-5. Histogram of sacks of cement used for each casing string in the Indian Creek wells.
Production data were available for 52 wells in the Indian Creek field on a monthly basis since 1985. Over
40 billion standard cubic feet (Bcf) of gas were produced since 1985 between the 52 production wells.
The cumulative production by well ranged from 0 MCF to over 4.3 million MCF. On average, a well had
a cumulative gas production of 780,000 MCF. The Indian Creek field experienced two major periods
without production: from January 1987 to October 1987 and again from January 2012 to October 2013.
Production ended in January 2015, and the field has recently been sold to new ownership.
Figure 3-6 shows the annual and cumulative production of gas for each well in the Indian Creek field; in
the figure, the periods of no production are visible. The production of gas was not consistent across the
field. The greatest amount of gas was produced in the southernmost wells, while the least amount was
produced in the northern wells.
No well workover or leakages were recorded for wells in the Indian Creek field. A search on the West
Virginia DEP oil and gas database showed no violations listed for Indian Creek wells. Detailed operator
records were unavailable for the field. Informal discussions with the field technician suggest that the wells
had few problems with no significant cause for frequent repairs or other corrective actions.

Figure 3-6. Annual and cumulative production in the Indian Creek field since 1985.
3.4 Williston Basin Site

Due to different reporting regulations and publicly available data, the dataset collected for the Williston
Basin site contained different information than those collected for the Michigan Basin and Appalachian
Basin sites. Information pertaining to well type, status, construction, and cementing was available. There
were over 3,000 identified wells in the Williston Basin site, and data were compiled for 1,432 wells.
Well drilling in the Williston Basin surged in the 1950s. Drilling tapered off through the 1990s, followed
by another increase in drilling in the 2000s. Approximately 60% of the wells were vertical, with 39%
recorded as horizontal. The remaining 1% of wells were not indicated or were listed as deviated. The
primary target formation was the Midale, which resulted in 97% of the wells reaching a true vertical depth
between 4,000 and 5,000 ft. The wells were mostly oil wells (68%); however, some were used for CO2,
H2O, or WAG injection. Nearly 22% of the wells have been abandoned, 42% are producing, and 22% are
operating (Figure 3-7). The remaining wells have been listed as observation, suspended, or long term shut
in (LTSI). Most wells were recorded as being exposed to oil (68%), and some wells were exposed to CO2,
gas, water, or a combination thereof.

Figure 3-7. Status of wells at the Williston Basin site.
Casing details were available for 1,328 wells at the Williston Basin site. The wellbores were consistently
constructed using two casing strings: a surface casing and a production casing. The diameter of the
surface casing ranged from 7.0 to 13.375 inches. The most common surface casing diameter was 8.625 or
9.625 inches, with 96% of the wells drilled no deeper than 1,000 feet. The production casing ranged in
diameter from 2.875 to 10.75 inches. Approximately 50% of the wells had 7-inch-diameter production
casing, with 30% of wells having production casing of 5.5 inches in diameter and 18% with 4.5-inch
casing. The production casing was set at depths up to 5,600 ft, with most wells set between 4,000 and
5,000 ft deep to target the Midale.
Cementing data were recorded as cement volumes rather than the amount of materials used. Cement
volumes were only available for 135 wells. The volumes ranged from 170 cubic feet (ft3) to 1,560 ft3,
with an average of 545 ft3. Most wells had less than 400 ft3 of cement. Figure 3-8 shows a histogram of
cement volumes for available wells at the Williston Basin site.
Full operational data and workover histories were not publicly available for Williston Basin wells.
However, numerous wells (80%) were tested for SCP as part of a monitoring program. The level of SCP
was categorized by leakage severity. Most of the wells tested did not show any SCP or leakage through
the wellbore; a small percentage had minor SCP associated with non-serious leakage. Some wells (34%)
showed significant SCP and were categorized as having serious leakage. Figure 3-9 illustrates the number
of wells that fell into each leakage category.

Figure 3-8. Cement volumes for wells at the Williston Basin site.
Figure 3-9. Casing pressure category for wells at the Williston Basin site.

3.5 Key Findings of Well Record Data Collection and Review
To assess wellbore integrity, publicly available data were collected for the three field test sites for wells
with different levels of exposure to CO2. The data included information on well construction, cementing,
workovers/leakage, and operations. Data availability varied by site due to differences in reporting
regulations on the state and country level.
Data Collection and Summary for the Michigan Basin
• Data were collected and reviewed for 1,622 wells (418 Niagaran and 1204 Antrim wells).
• Antrim wells occurred at shallow depths and were exposed to naturally occurring CO2. The
Antrim derived CO2 is removed in gas processing and partially utilized for CO2 EOR operations.
• Niagaran wells produced oil and gas from reef structures. Additionally, some were used for
CO2 EOR as either injection or producing wells.
• Drilling, construction, and cementing data were available for all wells.
• Well workovers were often recorded for wells but were typically maintenance or adding
perforations. No major workovers or leakages were reported.
• Operational data were available for all production wells on a monthly basis since 1982.
Data Collection and Summary for the Appalachian Basin
• Data were collected and reviewed for 55 wells.
• The primary target was the Tuscarora sandstone in the Indian Creek field.
• CO2 was naturally occurring in the Indian Creek field, exposing wells to varying amounts of CO2
over the well life cycle.
• Drilling, construction, and cementing data were available for all wells.
• No well workovers or leakages were reported in public files.
• Operational data were available for all wells on a monthly basis since 1985. The data showed
periods of production and non-production.
Data Collection and Summary for the Williston Basin
• Data were collected and reviewed for 1,432 wells.
• The primary target was the Midale carbonate in the Weyburn field.
• Wells were primarily used for oil and gas production, but many were later utilized for CO2 EOR
operations.
• Drilling and construction data were available for all wells. Cementing data were sparse and
recorded as cement volumes.
• No well workovers were recorded; however, many wells were tested for SCP as part of monitoring
regulations. The level of leakage associated with SCP was indicated for the wells.
Overall, the data collected for the three field test sites represent wells which have been exposed to CO2
either naturally or through EOR operations. These datasets provide unique opportunities to study the
influence of CO2 on wellbore integrity. The available data varied greatly by site, but the datasets were
valuable for developing wellbore integrity predictors for Tasks 4 through 7.

Chapter 4.0 Log- and Testing-Based Well Integrity
Assessment
The objective of Task 4 was to analyze well integrity for the three field sites based on logs and well
records. Task 4 was divided into three subtasks: Log Analysis, Well Record Analysis, and Well Integrity
Evaluation. The well risk analysis was focused on field test sites in Otsego County, Michigan (the
Michigan Basin site), Kanawha County, West Virginia (the Appalachian Basin site), and Saskatchewan,
Canada (the Williston Basin site). Task 4 was completed in October 2017. Well log data for the Michigan
Basin and Appalachian Basin field sites were collected in Task 3. Log data for Weyburn Field were
collected as part of the SaskCO2USER project (Duguid et al., 2011) and were provided to Battelle already
interpreted for risk assessment. The available log data are summarized in Table 4-1.
Table 4-1. Number of logs used for Otsego, Michigan, and Indian Creek, West Virginia, fields.
Log Type Otsego, MI Indian Creek, WV
Cement Bond Log (CBL) 1 12
Gamma Ray (GR) 54 11
Formation Bond (FBL) -- 7
Neutron Log (NEU) 43 1
Density (DEN) 38 3
Acoustic (XMAC) 26 --
Resistivity (RT) 49 --
Pulsed Neutron Capture (PNC) 3 --
Caliper Log (CAL) 53 36
Bit Size Log (BIT) 6 --
4.1 Log Analysis

Log analysis included caliper log analysis for borehole irregularities and CBL analysis for cement
emplacement. Analysis was performed on subset of available logs for the Michigan Basin and
Appalachian Basin sites.
Caliper Analysis
Caliper logs from the deep section of 53 wells within the Michigan Basin field test site and 36 wells
within the West Virginia field test site were reviewed to determine the variations over the depth. Where
possible, the reasons for variation were identified, including washouts due to formation properties. The
deep section of the wells was divided into different zones (over depth) based on variations in the diameter
of the wellbore shown in the caliper logs. For example, if sections of the well displayed different borehole
diameters, these sections would be split into different zones. For each zone, the average borehole diameter
was then estimated based on visual inspection of the caliper logs, and the annular volume was calculated
using the borehole diameters. Ultimately, the volumes for each zone were added together to determine the
annular volume over the entire well.
In the Michigan Basin site, the boreholes were close to gauge (bit diameter) over most of the deep section
of the well, often showing a borehole diameter of between 8 (gauge) and 10 inches. However, in the
F-Salt zone, which is dominated by halite deposits, the borehole shows significant washouts that increase
the borehole diameter to greater than 16 inches (or tool limit) in some cases. Like the Michigan Basin
wells, the wells in the Appalachian Basin site typically run close to gauge of the bit (often 6 to 7 inches in
these wells). However, washouts occur in the Rose Hill formation, where the borehole diameter may
increase to greater than 13 inches.

CBL Analysis
Twelve CBLs were available for the Appalachian Basin site. CBL analysis was completed with a
standardized CBL interpretation tool that was developed by Battelle to evaluate well cement quality
(Haagsma et al., 2015). The CBL interpretation tool uses a bond index method to provide a more
objective rating of cement quality in a well. The results of the tool may be used to determine cement-to-
casing bond in vertical zones within a well. Considering multiple wells in a field, spatial trends in cement-
to-casing bond may be apparent. CBLs for the Appalachian Basin field test site were imported into the
CBL interpretation tool, and the cement bond was rated for discrete intervals.
A total of 87 CBLs were available for the Michigan Basin field test site. The CBLs were also analyzed
with the Battelle CBL interpretation tool. Many of the Antrim well CBLs were not suitable for analysis
due to illegible log quality in the available raster images. The majority of the wells were in the 80% to
100% rating, but 20% of the wells had an index rating of less than 60%. The percentage of total casing
that was cemented for the Michigan Basin wells showed a similar distribution. Overall, the CBL analysis
suggests that there is a bit of a bimodal distribution of cement quality in the Michigan Basin field test site.
The deeper Niagaran reef wells had high-quality cement ratings, but the shallow Antrim shale wells had
lower cement ratings.
4.2 Well Record Analysis
Well records were collected from public databases and project partners. Data were collected from the
Michigan DEQ, the WVGES, and the Alberta Energy Utilities Board. Core Energy provided additional
details on their wells used in the study, and the Petroleum Technology Research Centre (PTRC) provided
the database used for the risk assessment performed in the SaskCO2USER project (Zaluski et al., 2016;
Duguid et al., 2017). The data included drilling records, cementing records, workover records, plugging
records, logs, and permits. The records provided a consistent set of data that could be used to develop a
proxy for likelihood of leakage and severity of impact for each well in each field. Records were collected
for 54 wells in Michigan, 47 wells in West Virginia, and 1,391 wells in Alberta. From these records, the
dataset was developed to include five categories common across each field. Table 4-2 lists the categories
and the number of wells in each category for each field.
Table 4-2. Number of wells in each field with known data for
each category ranked in the well integrity evaluation.
Risk Michigan Appalachian Williston
Field
Type Basin Basin Basin
Total number of wells (records collected) 54 47 1,391
Cemented through caprock 51 41 176
Likelihood of
leakage
Well deviation 54 47 1,391

Well age 54 47 1,390
Well status 54 47 1,391
Well type 54 47 1,391
Distance from developed populated areas 54 47 1,391
Severity of
impact
Distance to domestic groundwater well 54 47 1,391

Distance to environmentally sensitive areas (including
54 47 1,391
surface water sources)

4.3 Well Integrity Analysis
Wellbores with integrity issues in CO2 storage fields can lead to leakage out of the storage zone. Using
the data collected in Subtasks 4-1 and 4.2, a well integrity evaluation was conducted in order to determine
the overall risks for each well in the three field test sites. The well log data analyses, well record analyses,
and risk factors were integrated to calculate a Risk Score for each well. These Risk Scores were then used
to determine the wells that posed the greatest risk to wellbore integrity issues. A method to calculate the
risk from the individual wells in the field was adopted generally following that of Duguid et al. (2017).
The method used here calculates a Total Likelihood Score as the summation of the individual scores
judged to be proxies for the likelihood of leakage category (Equation 8) and calculates a Total Severity
Score as the summation of the severity of impact category each ranked between one and five
(Equation 9). Total Risk was calculated as the product of Total Likelihood and Total Severity
(Equation 10).
𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑘𝑒𝑙𝑖ℎ𝑜𝑜𝑑 = ∑𝑛𝑖=1 𝐿𝑖𝑘𝑒𝑙𝑖ℎ𝑜𝑜𝑑 (Equation 8)

𝑇𝑜𝑡𝑎𝑙 𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 = ∑𝑚
𝑗=1 𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 (Equation 9)
𝑇𝑜𝑡𝑎𝑙 𝑅𝑖𝑠𝑘 = 𝑇𝑜𝑡𝑎𝑙 𝐿𝑖𝑘𝑒𝑙𝑖ℎ𝑜𝑜𝑑 ∗ 𝑇𝑜𝑡𝑎𝑙 𝑆𝑒𝑣𝑒𝑟𝑖𝑡𝑦 (Equation 10)

The wells were divided into categories common to wells at each field test site. Equation 10 was used to
calculate a semiquantitative risk based on ranking each category between 1 and 5 for each well. Five
categories affecting the likelihood of leakage along a wellbore were identified: primary cement through
the caprock, deviation of the well, age of the well, status of the well (i.e., producing/injecting or
abandoned), and type of the well. These likelihood criteria were ranked from 1, the lowest likelihood of
contributing to a leak, to 5, the highest likelihood of contributing to a leak. To provide the most
conservative estimate of wellbore risk, any likelihood criteria that was unknown was automatically
assigned the highest value (5).
The depth of the top of cement (TOC) in a borehole was determined one of two ways: (1) through
volumetric calculation, using borehole and casing diameter from wells and the number of sacks of cement
based on construction records, or (2) picking the TOC based on CBLs. If a primary cement job placed at
least 40 ft of cement into the caprock, it was assumed that an effective seal was established. The
orientation of the well or the degree of deviation (vertical, horizontal, etc.) can affect the construction of
the well and the ability to isolate the well from different zones. Well orientation was classified as vertical,
horizontal, deviated, slant, and unknown, and each category was qualitatively ranked with respect to
leakage likelihood.
Well materials are subject to degradation over time through corrosion or other chemical processes. Thus,
older wells are considered to have a higher likelihood of leakage than more recently drilled wells. Older
well casing or cement is more likely to have degraded through exposure to ambient and produced fluids,
and older wells are more likely to be damaged by well operations such as workovers. In addition, changes
in regulations for oil and gas well construction have led to more effective environmental controls over
time, which also reduces the likelihood of leakage. Well age was binned in 13- to 15-year intervals,
beginning in 2017.
The status of the well (active, temporarily abandoned, plugged and abandoned, etc.) could also affect the
likelihood of leakage. The status of the well, whether producing or not producing, affects the pressure
gradient around the wellbore and also the amount of monitoring the well receives. Wells were categorized
in four groups for this assessment: active, shut-in, temporarily abandoned, and plugged and abandoned.
Well type likelihood scores were determined based the types of fluids near the well and the operations
being conducted. Monitoring wells do not have active injection or production but are exposed to
formation fluids. The monitoring wells included in this project were completed in a reservoir with CO2
brine and hydrocarbons.

Three potential receptor categories were assessed for severity of impact due to leakage: population centers
(i.e., people); groundwater; and surface water and environmentally sensitive areas (treated as a single
category). Although an entire aquifer formation would be considered an underground source of drinking
water, the actual groundwater receptors were wells used for potable water and, where applicable,
wellhead protection areas were used as receptor locations. Surface water receptor locations consisted of
intermittent and perennial streams, lakes and ponds, and surface water infrastructure (i.e., canals), where
applicable. Two designations in the dataset referred to inferred surface water pathways: artificial paths
and connectors. These were also treated as surface water features since they represent the most likely path
of surface water flow in areas with missing data.
Once the individual parameters were tallied, Likelihood Scores, Severity Scores, and Total Risk Scores
were calculated using Equations 8, 9, and 10, respectively. Likelihood and Severity Scores are the sum
the scores of all individual likelihood criteria and severity criteria, respectively. Total Risk Scores are
calculated by multiplying the sum of the Likelihood Scores by the sum of the Severity Scores.
This section presents the score results by field test site. Total Risk Scores for the Michigan Basin wells
ranged from 77 to 255. The histogram of Risk Scores for the wells included in this study is a bell curve
that centers around bins 151-175 and 176-200, each of which contains 16 wells (Figure 4-1). Differences
in Risk Score do not follow an obvious geographical pattern (Figure 4-2).
Figure 4-1. Histograms of Total Likelihood and Total Severity Scores (top) and
Total Risk Scores (bottom), Northern reef trend, Otsego County, Michigan.

Figure 4-2. Total Risk Scores, Northern reef trend, Otsego County, Michigan.
Total Risk Scores for the Appalachian Basin wells ranged from 55 to 165 (the minimum possible Total
Risk Score was 15, while the maximum possible Total Risk Score was 375). The average Total Risk
Score was 98.2 with a standard deviation of 45.3. The histogram for risk scores was a bell-shaped curve
centered between the bins 76-100 and 101-125 that skewed toward lower values (Figure 4-3). More than
half of the wells had a Total Risk Score of 100 or less (25 locations). The wells with lower Total Risk
Scores are located in the eastern and central portions of the field, away from population centers
(Figure 4-4), suggesting that distance to population was a determining factor for total risk.
Histograms for the Total Likelihood Scores, Total Severity Scores, and Total Risk Scores for Williston
Basin wells included in this study are presented in Figure 4-5. Maps are not presented for the Weyburn
Field at the request of the operator that provided the data. Total Risk Scores for the Williston Basin wells
included in this study range from 27 to 300 (the minimum possible Risk Score is 15, while the maximum
possible Risk Score is 375). The average Total Risk Score was 101.7 with a standard deviation of 25.6.
The histogram of Total Risk Scores for the wells included in this study is skewed toward lower values.
The largest number of wells has a Total Risk Score between 101 and 125. More than half of the wells
included in this study (756 locations) had Total Risk Scores of 100 or less.

Scores (bottom), Indian Creek Field, Kanawha County, West Virginia.
Figure 4-4. Total Risk Scores, Indian Creek Field, Kanawha County, West Virginia.

Scores (bottom), Weyburn Field, Saskatchewan, Canada.
The three field test sites were assessed using a Features, Events, and Processes (FEP)-based risk ranking
methodology. The fields varied in size, geography, and geology. The assessment showed that Total
Severity was probably more important in determining Total Risk. Correlation analyses of the Total
Likelihood data showed that cement through caprock, well age, and well status had the largest effect on
the Total Severity rank, with cement through caprock being the most important for each field. The CBL
analysis conducted for the Indian Creek and Otsego wells shows that some wells may have primary
cement that may pose a risk with values of less than 80% bond index.
The effect of unknown data was also most pronounced in the Alberta dataset, with only 176 of 1,391
wells with data for the cement through caprock category. However, for each of the other fields, the most
missing data were in the cement through caprock category as well. The strong correlation may be due to
giving the worst-case score to categories with missing data, but it also highlights the need to collect
additional data to ensure that these categories can be fairly and fully evaluated. Correlation analyses of
the data show that Total Risk follows more closely the Severity categories than the Likelihood categories.
This finding has implications for risk mitigation. Mitigation of severity is most likely to come from early
warnings from monitoring tools, implying that the receptors of concern (the local population,
groundwater, and surface water and environmentally sensitive areas) may need to be monitored.
4.4 Key Findings of Log- and Testing-Based Well Integrity Assessment
The three field test sites in Otsego, Michigan, Indian Creek, West Virginia, and Alberta, Saskatchewan,
were assessed using a risk ranking methodology. The fields varied in size, geography, and geology. The
assessment showed that Total Severity was more important in determining the Total Risk.
Correlation analyses of the Total Likelihood data showed that cement through caprock, well age, and well
status had the largest effect on the Total Severity rank, with cement through caprock being the most
important for each field. This finding may indicate that well isolation for each CCS project may need to
be better understood in order to properly assess leakage risk. The CBL analysis conducted for the Indian

Creek and Otsego wells shows that some wells may have, on average, primary cement that may pose a
risk with values of less than 80% bond index. However, in many cases there were portions of each well
that would be expected to isolate the CO2 zone from the surface. If a cement job is poor and no
information can be found, risk could be mitigated by working over wells to establish that there is isolation
or by monitoring wells to catch leaks early.
The effect of unknown data was most pronounced in the Alberta dataset, with only 176 of 1,391 wells
with data for the cement through caprock category. However, for each of the other fields, the most
missing data were in the cement through caprock category as well. The strong correlation may be due to
giving the worst-case score to categories with missing data, but it also highlights the need to collect
additional data to ensure that these categories can be fairly and fully evaluated.

Chapter 5.0 Sustained Casing Pressure Analysis
A key part of the project was testing SCP in CO2 wells at the field test sites. The tests provide a means to
measure the nature and severity of well defects in CO2 wells. The procedures, results, and analysis of SCP
field testing are provided in Sections 5.1 (Michigan Basin) and 5.2 (Williston Basin).
5.1 Michigan Basin SCP Field Testing
Niagaran reefs at the Michigan Basin field test site have been subject to CO2 EOR since 1996, with
operations expanding to a total of 10 reefs. The Niagaran fields were developed since the 1960s in the
region. The production history of the tested reefs is outlined in Figure 5-1. Overall, the Antrim
shale/Niagaran reef is an attractive area for examining wells exposed to different types of CO2
environments in the subsurface. The site also has other hydrocarbon wells and injection wells for
comparison testing (Figure 5-2).
Multiple wells in the Michigan Basin were measured for casing pressure. If well conditions indicated that
pressure might continue in the well, the well was selected to be part of enhanced testing. Six selected
wells were tested using a pressure bleed-down/buildup test, and gas samples were collected from the
wells. In addition, the gas chamber volume was measured in select wells. Table 5-1 summarizes the well
construction specifications for the 23 wells surveyed for indications of SCP. The majority of the tested
wells were drilled as primary production wells in the 1970s; however, a few of the wells were drilled as
part of recent CO2 EOR operations.
Reef Primary Pressurization EOR

Reef A 5/73 - 4/97 2/97 - 07/98 08/98 - Present
Reef B 5/74 - 4/96 5/96 - 12/96 01/97 - 11/12
Reef C 1/73 - 11/96 08/05 - 06/06 07/06 - Present
Reef D 08/71 - 10/09 11/09 - 09/11 10/11 - Present
Reef E 1/73 - 4/94 5/04 - 5/04 5/4 - Present
Reef F 1/73 - 8/87 5/11 - 5/12 06/12 - Present
Reef G 4/73 - 11/15 12/15 - Present N/A
Figure 5-1. Timeline of production and EOR of Michigan Basin reefs,
outlining the primary production, pressurization period when CO2 was being
injected without production, and EOR period.

Figure 5-2. Map of tested wells in the Michigan Basin, among many other wells in the area.
Table 5-1. Construction specifications for Michigan Basin wells surveyed.

Completion Total Casing
Well ID Type Comments
Year Depth (ft) Stringsa
MB-1 1973 Vertical 5,794 4 Intermediate annulus SCP test
MB-2 1974 Vertical 5,675 4 Deep annulus SCP test, cemented to surface
MB-3 1997 Deviated 6,450 4 Deep annulus SCP test
MB-4 2008 Vertical 5,850 4 Deep annulus SCP test
MB-5 1976 Vertical 5,655 4 Deep annulus SCP test, cemented to surface
MB-6 2012 Deviated 6,970 4 Deep annulus SCP test
MB-7 1996 Kickoff 7,134 4 Vertical with kickoff TVD 5,510 ft
MB-8 1976 Vertical 5,650 4
MB-9 1977 Vertical 6,255 4
MB-10 2006 Vertical 5,800 4
MB-12 1998 Vertical 5,700 4
MB-14 1973 Vertical 5,981 4
MB-15 2008 Vertical 6,202 4
MB-16 1973 Vertical 5,770 4 Deep casing cement to surface
MB-17 1984 Vertical 6,250 4
MB-18 1984 Vertical 6,324 4
MB-19 1984 Vertical 6,045 4
MB-20 1986 Vertical 6,130 4
MB-21 1986 Vertical 6,200 4
MB-22 1986 Vertical 6,000 4
MB-23 1975 Vertical 6,013 4
a. Including conductor. Note: TVD = true vertical depth.

Table 5-2 provides the sampling dates and field test results for the six wells (well IDs MB-1 through
MB-6) that underwent SCP testing and also describes the status of the remaining wells initially surveyed
for indications of significant casing pressures (well IDs MB-7 through MB-23). All the wells have a
similar casing design, with a conductor pipe to aid in the drilling operations, and surface casing cemented
back to the surface to protect potable aquifers. The intermediate casing is positioned with the shoe just
below the Bois Blanc formation. This standardization is in place because the formation has been known to
take fluid while drilling and circulating in certain parts of Michigan. The production casing is positioned
near or at total depth (TD). The completion is performed by perforating the zones of interest, and
chemically treating the perforations with acid and other chemicals when needed. Wellbore diagrams with
geologic stratigraphic columns of the six tested wells are shown in Figure 5-3.
Table 5-2. Field test results for Michigan Basin wells.

Name Reef Sampled Description
MB-1 Reef A 3/23/2016 32 psia initial pressure
MB-2 Reef B 3/23/2016 30 psia initial pressure
MB-3 Reef B 3/23/2016 162 psia initial pressure
MB-4 Reef C 5/1/2016 16 psia initial pressure
MB-5 Reef C 5/2/2016 35 psia initial pressure
MB-6 Reef D 5/3/2016 25 psia initial pressure
MB-7 Reef B 3/23/2016 inaccessible valve
MB-8 Reef C 3/23/2016 inaccessible valve
MB-9 Reef C 3/23/2016 no pressure on annulus
MB-10 Reef C 3/23/2016 inaccessible valve
MB-11 Reef E 3/23/2016 no pressure on annulus
MB-12 Reef A 3/23/2016 inaccessible valve
MB-13 Reef D 3/23/2016 inaccessible valve
MB-14 Reef D 3/23/2016 subgrade valve
MB-15 Reef F 3/23/2016 no pressure on annulus
MB-16 Reef F 3/23/2016 inaccessible valve
MB-17 Reef G 3/23/2016 subgrade valve
MB-20 Reef G 3/23/2016 no pressure on annulus
MB-23 Reef G 3/23/2016 27 psia initial pressure, drops to 17 psia after 2 mos
Note: Highlighted rows indicate the six Michigan Basin wells selected for SCP testing.

Figure 5-3. Well diagrams for Michigan Basin wells tested for SCP buildup.
Monitoring was performed by recording the temperature and pressure to have a record of pre-test
conditions. The process was continued through depleting the pressure in the annulus to atmospheric
conditions. In cases where gas chamber volume was needed, a flow meter was attached to the correct
valve, and the volume of flow out of the annulus space was recorded. Later, these data were converted
from atmospheric conditions (pressure and volume coming out of the wellhead into the atmosphere) using
the annulus pressure and a form of the Ideal Gas Law formula to calculate the gas volume released.
After opening the valve, venting the gas, and properly closing the valve, a memory gauge was attached.
This was done to monitor the pressure buildup in the annulus. The gauge was programmed to record a
pressure/temperature measurement at a preselected interval of time, such as a minute-by-minute
measurement. The well was then left alone for a minimum of 2.5 weeks to see if the pressure would build
up in the annulus. After this time period had fully passed, the gauge was removed and the pressure
buildup data were collected. The data revealed how long it took for the pressure to return to previous
levels, as well as the influence of thermal changes on the gas in the annulus.
If there was SCP in the annulus of the well, a gas sample was taken. This sample would have been
procured after the initial drawdown of the well to ensure that the sample was representative of the gas that
was making its way into the annulus. It was important to ensure that the sampling cylinder was visually

inspected and cleaned ahead of time, and it is often purged on site with annulus gas. Each sample was
tracked with a tag that included the well name, API, date/time, and all other pertinent identifiers. These
samples were then sent to a laboratory for testing to determine the individual gas compositions, such as
CO2, nitrogen, hydrogen, helium, oxygen, methane, ethane, butane, pentane, and other gases. Results of
these analyses are discussed in the following section. Table 5-3 lists the wells that were tested, along with
a short summary of the well. Some wells were inaccessible due to adverse conditions. Subgrade valves
did not allow for annulus testing.
Test results in Table 5-2 and alternatively in Figure 5-3 illustrate that six wells had no pressure, four wells
had subgrade valves, six wells were inaccessible, one well produced results inconclusive with SCP after
pressure monitoring, and six wells had gas samples collected. These six wells were also monitored for
pressure. One gas sample was inconclusive, which was taken from MB-6. Out of those five wells with
conclusive gas samples, only one well, MB-3, had a significant percentage of CO2 in the annulus.
However, further analysis will be needed to determine if this is due to CO2 injection, or if the CO2 buildup
is due to gas from the Antrim formation.
Table 5-3. Gas sampling results of six tested wells.
MB-1 MB-2 MB-3 MB-4 MB-5 MB-6a
Component
Mol %
Helium 0.23 NIL NIL NIL NIL N/A
Hydrogen 0.148 56.212 17.521 30.952 0.795 N/A
Carbon Monoxide NIL NIL NIL NIL NIL N/A
Oxygen 0.035 1.406 0.225 0.624 16.383 N/A
Nitrogen 1.198 6.506 1.153 51.035 74.058 N/A
Methane 95.456 35.378 1.976 16.39 2.81 N/A
Carbon Dioxide 1.175 0.031 75.263 0.191 0.607 N/A
Ethane 1.308 0.035 0.602 0.159 1.182 N/A
Propane 0.075 0.086 0.88 0.2 1.387 N/A
Iso-Butane 0.098 0.035 0.325 0.047 0.638 N/A
n-Butane 0.017 0.05 0.586 0.133 0.975 N/A
Iso-Butylene NIL NIL NIL 0.005 0.011 N/A
Iso-Pentane 0.06 0.03 0.462 0.062 0.573 N/A
n-Pentane 0.007 0.016 0.343 0.073 0.332 N/A
Hexanes Plus 0.193 0.215 0.664 0.129 0.249 N/A
a. Samples were taken of MB-6, but the sampling process was unsuccessful.
A pressure bleed-down/build-up test was performed on each well involved in this study to evaluate the
pressure-response curve related to the SCP. The casing valve associated with the annulus being tested was
then opened to allow the pressure to bleed down to near-atmospheric conditions. Next, a data-recording
pressure/temperature gauge was connected to the annular space to monitor the pressure recovery curve.
Figure 5-4 shows the comparison of pressure and temperature versus time for the Michigan Basin wells.
For MB-1, MB-4, and MB-6, it is apparent that temperature is driving the changes in pressure. MB-3
seems to be building pressure over time. MB-2 and MB-5 are inconclusive; these tests, for example, may
have been affected by the formation of ice, hydrate, or another substance. The beginning and end points
show where the gauge is reading standard conditions before and after being connected to the well.

Figure 5-4. Pressure and temperature testing data from Michigan Basin wells.
In two of the wells, the annular space was bled down through a flow meter to determine the volume of the
gas released; a valve at the outlet of the wellhead maintained a constant flow rate. Using the time required
to blow the well down to near-atmospheric levels and the initial pressure, the gas volume released was
calculated. The total volume and the initial pressure were used with the Ideal Gas Law to determine the
gas volume released. The volume of gas released from the annulus of well MB-1 was calculated to be
about 18 ft3, which was calculated using a well pressure of 17 psia, and bleeding down to a pressure of
15 psia. The total open annulus volume was calculated to be about 420 ft3, using standard casing and hole
sizes. The volume of gas released from the annulus of well MB-3 was calculated to be about 66 ft3, which
was calculated using a well pressure of 32 psia, and bleeding down to a pressure of 18 psia. The total
open annulus volume was calculated to be about 865 ft3, using standard casing sizes. The open annulus

volumes were calculated from the surface to the top of cement. These released volumes are much smaller
than the size of the open annulus.
5.2 Williston Basin SCP Field Testing

Approximately 3,000 wells are located in the Williston Basin test site. Field operational records were
reviewed to identify wells with reoccurring wellhead pressures on the deep annulus that would be
candidates for SCP testing. The screening effort identified 60 wells for this testing study. In order to
better pre-screen which wells to test, the operator performed bagging tests, where a bag was attached to
the annulus valve of the well to physically measure the amount of gas building up in the annulus. This
qualitative test helped to determine if the pressure drawdown test procedure listed above would have any
pressure buildup to measure. This survey effort resulted in 17 tested wells with significant casing
pressure. The wells were primarily oil producers with select wells used for H2O or CO2 injection.
Figure 5-5 shows the locations of the tested wells.
Figure 5-5. Map of studied wells in the Williston Basin field test site.
Table 5-4 summarizes the well construction specifications for the 17 Williston Basin wells tested for
SCP. Figure 5-6 shows the wellbore diagrams for WB-1, WB-4, WB-5, WB-10, WB-13, and WB-15
(other well diagrams are provided in the full field testing report). Testing showed that these six wells had
casing pressure buildup. The tested wells were between 1,381 and 1,582 meters deep (measured depth
[MD]). The tested wells in the field are being used for production, WAG, or water injection. Eight of the
wells were drilled in the 1950s, with the rest being drilled in the 1990s or 2000s. CO2 EOR operations
began in 2000, and the wells have been in CO2 environments in the subsurface since the early 2000s,
some of which have likely been exposed to CO2 since being drilled. Detailed operational data, well
histories, and certain details pertaining to well construction, such as cementing records, were not available
from the operator.

Table 5-4. Williston Basin well specifications.
Surface Prod Prod Csg Prod
Surface
Well Comp. TD, TVD Csg Csg OD Csg OD Csg
Well ID Csg
Profile Date (m) Strings Nominal Depth Nominal Cement
Depth (m)
(in) (m) (in) (m)
WB-1 Horizontal 2005 1,456 2 137* 9.625* 1,300* 5.5* 885*
WB-2 Vertical 1957 1,471 2 111 8.625 1,470 4.50 N/A
WB-3 Vertical 1955 1,582 2 152 10.75 1,411 7.00 N/A
WB-4 Vertical 1957 1,402 2 184 10.75 1,402 5.50 1,051*
WB-5 Vertical 1957 1,410 2 95 8.625 1,410 5.50 1,059*
WB-6 Vertical 1958 1,381 2 94 8.625 1,381 5.50 N/A
WB-7 Horizontal 2001 1,399 2 152 9.625 1,315 7.00 N/A
WB-8 Vertical 1959 1,480 2 91 8.625 1,480 5.50 N/A
WB-9 Horizontal 2005 1,424 2 137* 9.625* 1,300* 5.5* N/A
WB-10 Horizontal 1995 1,471 2 163 9.625 1,361 7.00 N/A
WB-11 Horizontal 2000 1,443 2 150 9.625 1,358 7.00 N/A
WB-12 Vertical 1997 1,420 2 137* 9.625* 1,300* 5.5* N/A
WB-13 Horizontal 1994 1,458 2 153 9.625 1,355 7.00 N/A
WB-14 Horizontal 2005 1,433 2 185 9.625 1,335 7.00 N/A
WB-15 Horizontal 2001 1,459 2 156 9.625 1,351 7.00 N/A
WB-16 Vertical 1957 1,439 2 92 10.75 1,442 5.50 880*
WB-17 Vertical 1957 1,515 2 142 8.625 1,484 6 N/A
*estimated Note: TD = total depth; TVD = true vertical depth.
Figure 5-6. Williston Basin wellbore diagrams.

Pressure monitoring started with a job safety analysis that was performed at each well before testing for
SCP. The surface temperature was noted along with the pressure observed via pre-existing gauges on the
wellhead to ensure safe conditions. Then, a visual inspection of testing equipment was performed before
any equipment was installed. The monitoring was performed by recording the temperature and pressure to
have a record of pre-test conditions. The process was continued by attaching the apparatus (Figure 5-7)
and depleting the pressure in the annulus to atmospheric conditions. In the cases where gas chamber
volume was needed, the testing apparatus with an orifice was attached to the correct valve, and the flow
out of the annulus space would have been held constant through the orifice. Later, these data were
converted from atmospheric conditions (pressure and volume coming out of the wellhead into the
atmosphere) using the time of bleed down, the annulus pressure, and a simplified form of the Ideal Gas
Law formula to calculate the gas volume released.
Figure 5-7. Well testing apparatus at the Williston Basin site.
After this process, the gauge was removed, and the pressure buildup data were collected. The data
revealed how long it took for the pressure to return to previous levels, as well as how much influence
thermal changes throughout the course of the day had upon the gas in the annulus. Ideally, if there was
SCP in the annulus of the well, then a gas sample would have been collected after the initial drawdown of
the well to ensure that the sample was representative of the gas making its way into the annulus. Each
sample would have been tracked with a tag that included the well name, API, date/time, and all other
pertinent identifiers. These samples would then be sent to a laboratory for composition testing. However,
for this site, gas samples could not be collected due to low pressure and flow at the wellhead.
Consequently, historical operator records were examined to determine the gas chemistry and the source of
the gas. These operator records were compared to available simple gas samples taken by the operator,
which were used for a rough indication of where the gas was coming from. Low amounts of CO2 indicate
that CO2 is not originating from the injection zone.
Seventeen (17) of the 60 wells considered were tested with the apparatus containing the pressure and
temperature gauge. The 17 wells were ultimately chosen by the operator because of the operator’s
extensive knowledge of the field and well characteristics. Battelle was informed by the operator that a gas

bagging technique was used to obtain a physical measurement of the gas that built up in the annulus.
Also, simple, qualitative gas samples were taken (Table 5-5) in order to get a rough estimation of gas
composition, but this is not to be interpreted as a replacement for laboratory-quality gas sampling. In
many cases, there was not enough of a buildup to justify performing a full test with the pressure and
temperature memory gauge. Initially, 60 wells were selected for the study, and for the first few weeks of
testing, only one testing apparatus was available, with test duration lasting 7 days. Later in the testing
period, two additional testing apparatuses were added, with the tests concluding in October 2017 due to
inclement weather.
Table 5-5. Gas testing records for the Williston Basin wells.
Orifice Size
Well Number H2 (ppm) CH4 (ppm) CO2 (ppm)
(in)
WB-1 N/A N/A N/A 1/8
WB-2 19,375 460,000 4,000 1/8
WB-3 2,660,000 710,000 0 1/32
WB-4 105,000 110,000 0 1/32
WB-5 N/A N/A 0 1/32
WB-6 403,000 200,000 0 1/32
WB-7 41,000 4,000 14,000 1/8
WB-8 N/A N/A 0 1/32
WB-9 85,000 190,000 0 1/32
WB-10 N/A N/A N/A 1/8
WB-11 N/A N/A 18,000 1/8
WB-12 499,000 165,000 0 1/32
WB-13 N/A N/A N/A 1/8
WB-14 2,975 0 42,000 1/8
WB-15 N/A N/A 2,000 1/8
WB-16 45,000 60,000 0 1/32
WB-17 N/A N/A 0 1/32
Note: ppm = parts per million.
A pressure bleed-down/buildup test was performed on each of the 17 wells involved in this study to
evaluate the pressure-response curve related to the SCP. The casing valve associated with the annulus
being tested was then opened to allow the pressure to bleed down to near-atmospheric conditions. Next, a
data-recording pressure/temperature gauge was connected to the annular space to monitor the pressure
recovery curve.
Figures 5-8 and 5-9 show the comparison of pressure and temperature versus time. For WB-1, WB-4,
WB-5, WB-10, WB-13, and WB-15, buildup can be seen. WB-1 shows a moderate dip in the pressure
during the middle of the test period, which may have been a device error. For WB-1, the test
administrator had noted that the testing apparatus ‘froze up’ during the test. WB-2 and WB-3 show no
significant pressure increase, with the temperature in WB-3 having a slight effect on the pressure. WB-5
shows a possible mechanical defect, indicated by the sharp logarithmic increase in pressure. WB-10
shows an unexplained dip in the pressure buildup, perhaps due to a small release of pressure, or possibly
another factor. The other wells without buildup have varying amounts of pressure dependence on
temperature, with a handful of the wells experiencing a negligible decline in pressure.

Figure 5-8. Pressure-temperature testing data from Williston Basin wells WB-1 though WB-8.

Figure 5-9. Pressure-temperature testing data from Williston Basin wells WB-9 though WB-16.

To provide more information on the nature and severity of potential well defects for the operator, the
pressure buildup curves were analyzed with the methodology developed by Moody and Dotson (2015).
The conventional diagnostic test for SCP is the bleed-down/buildup test, in which the gas pressure is bled
off the annulus and the resulting buildup is recorded. The base SCP pattern (Figure 5-10) consists of
pressure increasing at a decreasing rate to an asymptotic pressure. Several researchers have detailed SCP
analysis methods, primarily originating with Xu and Wojtanowicz (2001), who described a method to
calculate cemented annulus permeability from SCP pressure observations. Huerta et al. (2009) proposed
the use of this method for CO2 storage well application. The method developed by Moody and Dotson
(2015) was simplified to assess the cumulative effect of all the defects in the seal represented by a
hypothetical flow restriction, quantified as a flow factor located at the top of the cement. This simpler
model has the advantage that it can indicate the character of the defect, providing diagnostic information
not available if only a permeability flow model is assumed.
Figure 5-10. General SCP buildup curve.
Eleven (11) of the 17 wells tested at the Williston Basin site showed no significant pressure rebound
pattern. However, six of the wells did exhibit a minor pressure buildup pattern that might be useful for the
operator to evaluate potential well defects present in shallow zones. Well construction information,
subsurface conditions, and pressure buildup data were input into the MATLAB script designed to
estimate well defect factors based on the methodology by Moody and Dotson (2015). This methodology
examines the gas influx rate change over time to determine the nature of a well defect based on a defect
model curve, similar to pressure transient well testing methods that examine pressure derivative curves.
There was some uncertainty on several input parameters related to height of liquid volume in the annulus,
gas volume due to low flow conditions, and gas properties. It was assumed that the calculations used
500 ft true vertical depth (TVD) of liquid length and a gas chamber length of 30 ft.
The SCP analysis for the six wells suggests mainly a pressure-limited orifice type defect factor. None of
the wells correlated to a permeable cement flow model, which would suggest cement degradation and gas
migration from the reservoir zone. Figure 5-11 shows the results for the wellbore integrity analysis. For
each well, the analysis provides a curve of pressure buildup over time and gas influx over time. The gas
influx curve shows a wellhead model curve of the type of cement defect as detailed by Moody and Dotson
(2015). For the six wells analyzed, the most suitable gas influx curve was determined to be a pressure-
limited orifice (O) or non-pressure-limited vein (V) model. The wells did not show a strong match to the
gas influx rate type curve for orifice type defect, so the understanding of defect type is uncertain.

Figure 5-11. SCP plots for Williston Basin wells.

Table 5-6 summarizes SCP analysis results. WB-5 and WB-15 had higher flow factors and potential for
sustained leakage. The other wells had small flow factors and less than 0.01 thousand standard cubic feet
per day (MSCFD) sustained leakage metric. The defect response was unclear but seemed to show an
orifice type curve. Overall, results support operator observations that there is low gas flow from a shallow
source and likely water present in the casing annulus.
Table 5-6. SCP analysis results for Williston Basin field site.
Well Asymptotic Defect Instantaneous Sustained Flow Factor
Gas
Well Depth Pressure Response Leakage Metric Leakage Metric (square
Source
(ft) (psi) Type (MSCFD) (MSCFD) microns)
Unclear
WB-1 4,777 Shallow 50 0.027 0.0076 564
(Orifice)
Unclear
WB-4 4,600 Shallow 28 0.026 0.0077 627
(Orifice)
Unclear
WB-5 4,626 Shallow 65 0.033 0.1500 >1,000
(Orifice)
Unclear
WB-10 4,826 Shallow 75 0.036 0.0082 555
(Orifice)
Unclear
WB-13 4,784 Shallow 45 0.022 0.0050 379
(Orifice)
Unclear
WB-15 4,784 Shallow 27 0.220 0.0200 >1,000
(Orifice)
5.3 Key Findings of SCP Analysis

Michigan Basin Site
A total of 23 CO2 wells were surveyed for indications of SCP at the Michigan Basin site. These wells
were present at an active CO2 EOR field. The wells were 5,600 to 6,500 ft deep and installed in the early
1970s, with some more recent wells installed between 1997 and 2012. Many of the wells were used for
primary oil production, water flooding, and CO2 EOR. CO2 EOR operations began in 1997, and the wells
have been in CO2 environments in the subsurface for 10 to 20 years. In general, the wells were
constructed with typical oil and gas materials such as carbon steel casing and class A Portland cement.
The field survey established that six wells had no pressure, four wells had subgrade valves, six wells were
inaccessible, one well produced results inconclusive with SCP after pressure monitoring, and six wells
with evidence of SCP had gas samples collected and were monitored for pressure. Out of the five wells
with conclusive gas samples, only one well, MB-3, had an amount of CO2 that could be considered above
background or of any statistical significance. Even with the molar percentage measured in the sample,
both the volume of gas in the annulus and the measured pressure were very low.
The data from the pressure and temperature versus time graphs show various results. For MB-1, MB-4,
and MB-6, it was apparent that temperature drives the changes in pressure. MB-3 seems to be building
pressure over time. MB-2 and MB-5 are inconclusive. Furthermore, only one well, MB-3, had a
significant initial pressure of 162 psia, but it did not build back up to that amount during the testing
period.

Overall, SCP was not a common problem for the tested wells at the Michigan Basin site. Well
construction practices were sufficient to isolate the injection zone and prevent any significant migration of
CO2 into the wellbore, including the following items:
• Multiple strings of casing (conductor, surface, intermediate, and deep) are present in the wells,
which reduces potential for gas migration from intermediate zones.
• Most wells were cemented across or near casing string crossovers, reducing pathways for gas
migration along the boreholes.
• More cement was used to cement in the casing strings than many other areas of the Midwest.
Casing strings were cemented in with several hundred feet (and in many cases, over 1,000 ft) of
cement.
• Cement was allowed to set and the top of cement was tagged to confirm the top of cement, which
was especially pertinent to this site because there was potential to lose cement in washout zones.
Williston Basin Site

A subsample of 60 wells with exposure to CO2 was surveyed for indications of SCP at the Williston Basin
site, with 17 being tested. These wells were present at an active CO2 EOR field. The tested wells were
between 1,381 and 1,582 meters deep (MD). The tested wells in the field are being used for production,
WAG, or water injection. Eight of the wells were drilled in the 1950s, with the rest being drilled in the
1990s or 2000s. CO2 EOR operations began in 2000, and the wells have been in CO2 environments in the
subsurface since the early 2000s, some of which have likely been exposed to CO2 since being drilled.
The field survey established that six wells had pressure buildup. The results for 11 wells showed no
evidence of SCP after pressure monitoring. Further analysis would be required to draw any meaningful
conclusions to the origin of the pressure buildup, as there are many meters of open annulus, being
exposed to multiple formations.
The data from the pressure and temperature-versus-time graphs show various results. For WB-1, WB-4,
WB-5, WB-10, WB-13, and WB-15, minor pressure buildup was observed. WB-1 and WB-10 had an
inconclusive moderate dip in the pressure during the middle of the test period. WB-2 and WB-3 showed
no significant pressure increase, with the temperature in WB-3 having a slight effect on the pressure.
WB-5 showed a possible mechanical defect, indicated by the sharp logarithmic increase in pressure. The
other wells without buildup exhibited varying amounts of pressure dependence on temperature, with a
handful of the wells experiencing a negligible decline in pressure.
Overall, SCP was not a common problem for the tested wells at the Williston Basin site. Well
construction practices appeared sufficient to isolate the injection zone and prevent any significant
migration of CO2 into the wellbore, based on the low annulus pressure buildup observed. For the cases
where moderate pressure buildup was demonstrated, the source of pressure could not be determined. It
must be taken into consideration that due to the open production casing annulus space, the slight pressure
increase observed may have originated from a shallow formation. Gas samples would need to be collected
and analyzed from the annulus gas, injection gas, and gas from shallow formations in order to identify the
source of the casing pressure.

Chapter 6.0 Field Analysis of CO2 Cement Sealing and
Well Integrity
The objective of Task 6 was to analyze field data on mineralogy, fluids, cement, hydrologic conditions,
and CO2 exposure for the three field test sites to determine cement sealing and well integrity relationship.
Data from previous tasks were analyzed for indicators of wellbore integrity factors that may predict
cement sealing conditions in relation to CO2 leakage potential in legacy oil and gas wells. These results
provide a better understanding of wellbore integrity effects for CO2 storage applications.
6.1 Analysis of Subsurface Setting for Cement Sealing

Mineralogy of the primary reservoir/CO2 storage zone and primary caprock was determined with thin
sections. The minerals present in the subsurface have the potential to interact with CO2, cement, and brine
mixtures. Table 6-1 summarizes the mineralogy for the field sites. The data on mineralogy were
synthesized for inclusion in the PHREEQC modeling.
Table 6-1. Summary of mineralogy for the field test sites.

Appalachian Basin Site Michigan Basin Site Williston Basin Sitee
Mineral
Caprocka Reservoirb Caprockc Reservoird Caprock Reservoir
Dolomite --- --- 39.7-99.7% 9-97.2% 1% 2-62%
Calcite --- --- --- 0.8-82.5% 39% 0-90%
Quartz 29.8-45.8% >90% 0.3-2.1% 0-1.5% 1% 0-11%
Anhydrite --- --- 0.0-56.0% 1.3-4% 59% 1-14%
K-Feldspar 0-0.5%f <1% 0.0-0.6% 0.5% 0% 0-11%
Plagioclase --- <1% --- --- 0% 0-5%
Illite 36-60.8% --- 0.0-2.6%g 0-2% 0% 0-7%
Fluorite --- --- --- --- 0% 0-2%
Chlorite 5.9-12.2% --- --- --- --- ---
Pyrite --- --- --- 0-0.5% --- ---
Halite --- --- --- 0-0.8% --- ---
Iron Oxide 0-2.1% --- --- --- --- ---
Carbonate 0-7.8% --- --- --- --- ---
Chert RF --- <1-1% --- --- --- ---
Heavy Min. --- <1-3% --- --- --- ---
a. From Jin et al. (2010).
b. Based on Kanawha County, West Virginia, data provided by WVGES Pipeline-Plus Oil and Gas Database.
c. From core XRD analysis of three samples of the A2 carbonate from the Dover 33 Reef.
d. Includes State Chester (limestone) and Dover 33 (dolomite) reefs.
e. Based on data from Durocher et al. (2005), Braunberger et al. (2012), and Hutcheon et al. (2008).
f. Listed as feldspar in the original database.
g. Total clay.
Appalachian Basin. Mineralogy of reservoir and caprocks for the Appalachian Basin Indian Creek
field was characterized by the WVGES based on analysis of available core samples from the Tuscarora
sandstone. Cores were inspected, and 45 thin sections were prepared for core samples from Clay
County well 4701500513 and Kanawha County well 4703903914. Rock core from the well was in
poor condition, and exact sample depths were not available. The Tuscarora sandstone samples were
>90% quartz with a small fraction of accessory minerals. Rose Hill shale caprock samples were not
readily available, so mineralogy was based on research by Jin et al. (2010). The Rose Hill is mostly

siltstone and shale consisting of clay minerals, quartz, and a minor amount of iron oxide minerals and
feldspars.
Michigan Basin. Thin-section mineralogical data are available for reservoir rock in the State Chester
reef and the Dover 33 reef from the Midwest Regional Carbon Sequestration Partnership (MRCSP).
Thin sections were also collected from the caprock (A2 carbonate) from the Lawnichak-Myszker
#9-33 well. Data from the State Chester well and the Dover 33 reef indicate that the Brown Niagaran
formation (the CO2 storage/EOR zone) is a marine carbonate dominated by calcite and/or dolomite,
depending on the degree of dolomitization that the reef has undergone. Typically, calcite and dolomite
make up more than 90% of the mineral phases present in the Brown Niagaran. Minor amounts of
pyrite, clay minerals, feldspars, and quartz are present in the storage reservoir. The caprock (A2
carbonate) is a marine evaporite deposit that also has been dolomitized. Primary minerals in the A2
carbonate are dolomite and anhydrite. Small amounts of illite, feldspars, and quartz are also present in
the A2 carbonate.
Williston Basin. General mineralogy of the storage zone and caprock for the Williston Basin field site
was compiled from several previous projects to examine CO2 storage processes in the basin. Table 6-1
summarizes general mineralogy as listed by Durocher et al. (2005), Braunberger et al. (2012), and
Hutcheon et al. (2012). The storage interval at the Williston Basin site consists mainly of carbonate
with minor amounts of anhydrite, quartz, and potassium feldspar. The caprock at the field site is
anhydrite and carbonate. Since the field site covers 45,000 acres, variations in lithology and
mineralogy are likely to occur, and the summary mineralogy should be considered general in nature.
Brine geochemistry is an important component of the subsurface system, because the brine will contact
the cement and other well materials. Table 6-2 summarizes the brine chemistry for the three field sites.
Table 6-2. Summary of brine geochemistry for reservoir samples from the field test sites.
Appalachian Basina Michigan Basin Siteb Williston Basinc
Constituent
(mg/L) Baseline Post-CO2 Injection
Low High Low High
Low High Low High
Ca 10,000 40,000 67,500 110,000 84,900 99,400 1,100 2,000
Mg 800 8,000 7,985 12,100 8,060 11,200 320 490
Na 40,000 110,000 15,300 22,500 14,400 21,300 12,000 26,000
K 400 1,700 11,000 16,600 16,200 18,400 120 600
SO4 300 1,300 81 97.1 44 150 3,180 3,620
Cl 73,000 183,000 251,000 274,000 188,000 270,000 18,000 43,000
HCO3- 10 160 361 468 NDd 956 NA NA
Br 879 4,650 2,280 3,030 NDe 3,250 NA NA
Al 0.54 3.51 ND 1.0 NDf 1.0 NA NA
Fe 186 --- 10 129 52.4 654 NA NA
SiO2 0.005 0.006 6.90x10-4 1.46x10-2 2.04x10-3 2.35x10-2
pH 5.1 6.4 4.83 5.88 4.10 4.87 6.3 --
Alkalinity NA NA
7 104 296 384 NDg 785
(as HCO3-)
TOC NA NA NDf 79 27 343 NA NA
DOC NA NA NDf 66 16 295 NA NA
Salinity 120,000 300,000 348,000 450,000 380,000 450,000 35,000 110,000
Water Sat. 40% 50% 30% 40%
a: Based on produced water for Medina-Tuscarora sandstones in WV & southeast OH from Battelle (2015) and Breen et al. (1985).
b: Baseline (pre-injection) data obtained from samples from five wells in fields (or lobes) that have not seen CO2.
c: Based on data from Cantucci et al. (2009); Mills et al. (2011); and Hassani et al. (2014).
d: Minimum detected concentration: 316 mg/L.
e: Minimum detected concentration: 2,720 mg/L.
f: One detected sample.
g: Minimum detected concentration: 259 mg/L.

Appalachian Basin. Brine geochemistry for the Appalachian Basin was based on chemical analysis of
produced water from the Tuscarora sandstone in the northwest West Virginia and southeast Ohio
portions of the Appalachian Basin (Battelle, 2015). Data from Breen et al. (1985) were used as an
additional source of data. These data were necessary to obtain measurements of aluminum and silica,
which were not available in the Battelle (2015) database.
Michigan Basin. Brine geochemistry for the Michigan Basin field site was based on fluid chemistry
sampling and analysis performed under the MRCSP Phase III field demonstration project. Under this
project, fluid samples were collected prior to and after CO2 injection for EOR operations. The
geochemistry for the site reflects carbonate reservoirs and the anhydrite/salt caprock layers. The brine
has very high salinity, high calcium, and moderate pH. Post-CO2 injection samples show similar
chemistry but with lower pH. These conditions show the range of in-situ chemistry in the reservoir
during CO2 injection.
Williston Basin. Brine geochemistry for the Williston Basin field site was based on research presented
by Mills et al. (2011) and Hassani et al. (2014). The brines at the field have salinity of approximately
30,000 to 100,000 milligrams per liter (mg/L) with minor constituents of calcium, magnesium, and
sulfate. Hydrogen sulfide (Laumb et al., 2017) is present in portions of the field. The field has been
subjected to both water floods and water-alternating CO2 gas floods, which may result in lower salinity
than initial in-situ conditions.
Cement composition is also a factor for potential degradation of well materials and cement sealing. Raw
materials of Portland cement include limestone and clay (shale), which are heated in a kiln to around
1,450 to 1,550 oC (Taylor, 1997; Michaux et al., 1989; Chamberlain et al., 1995). The resulting mixture,
referred to as clinker, is ground to a powder to form Portland cement. Most of the clinker (up to 80%) is
calcium silicates: alite (tricalcium silicate) and belite (dicalcium silicate) (Table 6-3). Most of the rest of
the clinker consists of celite (tricalcium aluminate) and brownmillerite (tetracalcium aluminoferrite).
Gypsum is also added to the mixture to prevent rapid stiffening.
Table 6-3. Composition of typical Portland cement clinkers.

Weight %
Cement
Michaux et al. (1989)
Name Formula Short- Chamberlain Taylor
Class Class Class Class Class
hand et al. (1995) (1997)
A B C D&E G&H
Alite Ca3SiO5 C3S 50 53 47 58 26 50 60
Belite Ca2SiO4 C2S 25 24 32 16 54 30 20
Celite Ca3Al2O6 C3A 10 8 5 8 2 5 7.5
Brownmillerite Ca4Al2Fe2O10 C4AF 10 8 12 8 12 12 10
Gypsum CaSO4●2H2O CS-H2 5 (a) (a) (a) (a) (a) (a)
Fineness - - - 1500- 1500- 2000- 1200- 1400- -

(cm2/g) 1900 1900 2800 1600 1700
Application - - - None Sulfate Early Retarded Stringent -
resistant setting Specs
a. Weight % of gypsum not included
Appalachian Basin. Well completion records were reviewed for 53 wells in the Appalachian Basin
field site to determine cement composition. The majority of records did not include details on cement
composition because the well completion forms did not require the information. Records with cement
information show generally a neat Portland Class A cement for the production casing. Records indicate
a 50/50 Pozmix with calcium chloride (CaCl2) and latex additive in the shallow and intermediate
casing strings to approximately 2,300 ft and 5,000 ft. Since the same operator installed most of the

wells in the field, it is likely that similar cement was used for all wells. Records suggest that carbon
steel (Grade J-55) casing was used for the casing materials, which is typical for the wells in the region.
Michigan Basin. Cement materials and additives were evaluated for wells in the Michigan field site as
part of research performed under a previous project (Battelle, 2016). Review of materials used for well
completion and plugging suggests that mostly Portland Class A cement was used for nearly all wells.
Cement additives were mainly CaCl2 accelerant, gel, salt, lost circulation material, ‘Baroco’ clay,
sulf-x, latex, gasblock, and Pozmix. Records suggest that standard Grade J-55 carbon steel was used
for casing in the CO2 storage-EOR zones in these wells.
Williston Basin. Rochelle et al. (2004) list well cements for the Williston Basin site as Portland
Class A 50-50 Pozmix with 0.5% friction reducer additive. A tail mix was listed as Portland Type G
thixotropic mix with 2% CaCl2 accelerator and 0.4% flocculant. The cement mixes were prepared and
set under reservoir conditions. The set cement showed Ca-MgAl-silicate hydrate± S and Fe matrix.
Casing used for Williston Basin wells was typical Grade H-40 for the surface casing and Grade J-55
for production casing (Laumb et al., 2017).
General hydrologic conditions were compiled from regional datasets, well records, and previous site
description tasks under this project. Table 6-4 lists general hydrologic conditions for the three field sites.
All reservoirs are deep enough to sustain supercritical CO2 temperature and pressure conditions. The
Michigan Basin field site has an additional zone at 800 to 1,200 ft that has natural CO2 mixed with natural
gas. Otherwise, the sites have similar hydrologic conditions.
Table 6-4. General hydrologic conditions for three field sites.

Appalachian Basin Michigan Basin Williston Basin
Parameter
Reservoir Caprock Reservoir Caprock Reservoir Caprock
Depth (ft bgs) 6,732 6,350 6,000 5,675 4,750 4,700
Thickness (ft) 50-75 350-400 18-280 75-160 30-100 6-30
Temperature (°F) 140 137 102 97a 145 144
Pressure (psi) 2,900 2,800 3,000 2,455b 2,100 2,020
Water Saturation (%) 50 50 20-30 0 30-40 0-5
Fluid Density 1.18 1.18 1.12-1.29 1.12-1.29 1.02-1.10 1.10+
Porosity (%) 8-16 1-5 3-11 0-1 10-40 1-3
a. Thermal gradient (depth/100 + 40oF)
b. Pressure gradient (depth*0.43 + 14.7 psi)
6.2 Geochemical Analysis to Predict Cement Sealing Conditions

Geochemical modeling was performed to evaluate interactions of well cements, reservoir/caprock
minerals, brine mixtures, and CO2 in relation to cement sealing conditions and well defects for the three
field sites. Previous work by DOE-NETL was reviewed to provide guidance on geochemical processes
for well cements and CO2 environments in the subsurface. The objective of this review was to provide
perspective on the conditions present at the Appalachian Basin, Michigan Basin, and Williston Basin
sites. The review found that a large amount of research on wellbore integrity for CO2 environments has
been completed, including laboratory tests on cement cores and synthetic brines, analysis of sidewall
cores, and modeling studies. Many of the tests were based on Williston Basin field sites, and these studies
were used as the basis for this project. Consequently, the Williston Basin site was based on previous
research. The Appalachian Basin and Michigan Basin sites were analyzed with geochemical models.
Williston Basin. Core flooding experiments and numerical modeling using PHREEQC code were
completed by Azaroual et al. (2004) and Riding and Rochelle (2005) for the Weyburn rock formation
to evaluate potential for geochemical reactions. PHREEQC geochemical equilibrium code was run for

Midale mineralogy under CO2 injection conditions at 150 bar and 54°C. Results suggest calcite and
feldspar dissolution with anhydrite, gypsum, and dawsonite precipitation. Clays dissolved near the
injection zone and precipitated farther away from the injection area. Overall, the simulations showed a
decrease in porosity in the storage formation. Cement interactions for Weyburn were studied by
Rochelle et al. (2004) for fill cement and tail cement samples. The laboratory batch tests on fill cement
showed development of calcite crystals 5 to 10 micrometers (μm) long, forming a crust approximately
40 μm thick. The researchers concluded that “The carbonation reaction produced a probable calcite-
rich front with significantly reduced porosity that varied up to 50-100 μm in thickness. In contrast to
the fill cement, the tail cement reacted extensively with the CO2-rich synthetic marly porewater to
produce a series of precipitates from a probable calcite and CSH gel, to ettringite and Ca-sulphate,
chloride.”
Appalachian Basin and Michigan Basin. PHREEQC Interactive, v. 3.4.0.12927, was used for
geochemical speciation modeling of the Appalachian Basin and Michigan Basin sites. A series of
model simulations were constructed to answer the questions listed in Table 6-5: (1) equilibrium model,
(2) equilibrium phases model, (3) CO2 batch model, (4) solid equilibrium phases model, and (5) batch
model. For all models, the Lawrence Livermore National Laboratory (LLNL) Geochemical database
was used (Johnson, 2010), supplemented with five cement mineral phases from the ThermoChimie
v.8.0 (September 2011) database (sit.db): CSH0.8, CSH1.2, CSH1.6, Friedel’s salt, and vaterite.
Table 6-5. Questions to answer with model.
No. Question Method
1 What minerals are supersaturated, saturated, and Equilibrium Model
undersaturated?
2 What minerals are precipitating from initial brine solution? Equilibrium Phases
3 What minerals dissolve or precipitate with the addition of CO2 to Equilibrium Model/CO2 batch
the initial brine solutions?
4 Are the formations (reservoir and caprocks of interest or Solid Equilibrium Phases
cement) affected by brine/CO2 interactions? Do they provide a
buffering capacity?
5 What roles do pressure, temp, pH, and pe play in resultsa? Equilibrium Model w. pH, pe, P, and T
a. Appalachian Basin is reported due to redundancy. Modeling results for Michigan Basin also available.
The geochemical model is an equilibrium speciation model. Kinetic reactions for cement are not well-
defined, so the model cannot demonstrate the timing of reactions. The LLNL geochemical model is
calculated based on extended Debye-Hückel, which is not typically used for waters with high ion activity
(e.g., brine). Pitzer calculations are used for slightly higher activity brines; however, the activities of the
brines of interest in this study are beyond the conditions that dictate when Pitzer calculations are usually
used. The specific model setup is based on brine chemistry across each field test site combined with solid
mineral complexes from different locations. Site-specific cement mineral phases were not available, so a
cement value from another site was used.
The midpoint brine sample was used in the model to determine the SIs, mineral precipitation, and
matrix/brine/CO2 interactions (Tables 6-6 and 6-7 for the Appalachian and Michigan Basins,
respectively). Batch reaction modeling was done using a range of parameters guided by the minimum and
maximum values for each parameter for the Michigan Basin and Appalachian Basin brines. Furthermore,
each parameter was extended beyond the maximum value to determine the range of possibilities for other
potential brines. The batch reaction parameters are defined in Table 6-8.

Table 6-6. Appalachian Basin chemistry summary data.
Midpoint values were used for geochemical modeling.
Depth pH Density TDS Ca Mg Na K
Sample
(ft) (S.U.) (kg/L) (mg/kg) (mol/kgw)
Average 4,473 5.59 1.161 212,979 0.59 0.11 2.3 0.03
Minimum 2,545 4.30 1.085 118,000 0.23 0.03 1.3 0.01
Midpoint 4,123 5.40 1.157 198,500 0.67 0.17 2.3 0.03
Maximum 5,700 6.50 1.229 279,000 1.10 0.30 3.2 0.06
HCO3- Alkalinity SO4-2 Cl- Br- SiO2

Sample
(mol/kgw) (mg/L as HCO3-) (mol/kgw)
Average 7.3E-04 49.4 0.006 3.7 0.02 9.0E-05
Minimum 1.0E-04 7.4 0.002 1.9 0.01 8.0E-05
Midpoint 1.4E-03 55.8 0.008 3.9 0.03 9.5E-05
Maximum 2.7E-03 104.2 0.014 5.8 0.06 1.1E-04
Table 6-7. Michigan Basin chemistry summary data.

Midpoint values were used for geochemical modeling.
pH Density Ca Mg Na K HCO3- Alkalinity
Sample
(S.U.) (kg/L) (mol/kgw) (mg/L as HCO3-)
Average 4.89 1.253 1.73 0.34 0.6 0.31 354.0
Minimum 4.10 1.118 1.05 0.26 0.1 0.15 0.0047 3.5
Midpoint 5.10 1.204 1.60 0.32 0.5 0.26 0.011 427.1
Maximum 6.09 1.289 2.15 0.39 0.9 0.37 0.173 850.6
SO4-2 Cl- Br- SiO2 Al Fe

Sample
mol/kgw
Average 0.001 3.7 0.03 9.0E-5 1.3E-05 4.1E-03
Minimum 0.0004 5.8 0.02 1.1E-5 2.9E-06 1.4E-04
Midpoint 0.002 3.9 0.03 9.5E-5 1.8E-05 1.2E-02
Maximum 0.004 1.9 0.03 8.0E-5 3.3E-05 2.4E-02
Table 6-8. Batch reaction parameters.

Parameter Basin Batch
CO2 (moles) Both Basins 0.01 0.034a 0.064 0.094 0.194 0.4 0.7 1.1b 1.5 2 2.5 3 4 5
… 49
Pe Both Basins -6.0 -4.0 -2.0 0.0 2.0 4.0 6.0 8.0 10.0
Pressure (atm) Appalachian Basin 1.0 2.0 10 50 100 150 197.3c 200 250
Michigan Basin 1.0 2.0 10 50 100 150 180.3c 200 250
Temperature (oC) Appalachian Basin 25 30 35 40 45 50 55 60c 70 80 90
Michigan Basin 25 30 35 40 43.9c 50 55 60 70 80 90
pH (S.U.) Appalachian Basin 5.1c 5.5 6.0 6.4c 7.0 9.0 11.0 13.0
Michigan Basin 4.1c 4.5 5.0 5.5 5.88c 6.0 6.5 7.0 8.0 9.0
a. Saturation at standard temperature and pressure;
b. Approximate saturation at reservoir conditions;
c. From site data.

The mineralogy inputs for the solids in the Appalachian Basin and Michigan Basins are shown in
Table 6-9 and 6-10, respectively. The reservoir rock, caprock, and cement were used as inputs for the
solid equilibrium phase model and the batch reactions model. In addition, the mineral phases were used
as inputs in the equilibrium phases model, although the molar mass for the phases was not indicated in
these simulations. As shown above, site-specific data were available for the Appalachian Basin reservoir
(Tuscarora sandstone) and caprock (Rose Hill) and for the Michigan Basin reservoirs (State Chester
[limestone] and Dover 33 [dolomite] Brown Niagaran) and caprock (A2 carbonate). However, because
cement mineralogical data were not available for the specific sites, cement equilibrium phases input is
based on cement mineralogy reported by Koukouzas et al. (2017). The percent composition for the
mineral phases was used to adjust the number of moles of each phase so the molar volume of the solid
was 9,000 cubic centimeters (cm3). This was reacted with 1.0 liter of water in the solid equilibrium phases
and batch reactions modeling efforts.
Table 6-9. Appalachian Basin solid equilibrium phases.

Vmin Tuscarora Rose Hill Cement
Mineral
cm3/mol % Comp. Moles % Comp. Moles % Comp.a Moles
Quartz 23.1 0.951 296.0 0.336 18.1 - -
K-Feldspar 108.7 0.005 1.5 0.008 0.4 - -
Albite 99.9 0.005 1.5 - - - -
Illite 256.2 0.014 4.5 0.540 29.1 - -
Pyrite 24.5 0.010 3.1 - - - -
Siderite 29.3 0.004 1.3 0.005 0,2 - -
Hematite 159.7 0.012 3.6 0.005 0.2 - -
Chamosite 7A 207.6 - - 0.088 4.7 - -
Calcite 36.9 - - 0.019 1.0 0.0777 4.2
Aragonite 34.2 - - - - 0.118 6.4
Vaterite 39.4 - - - - 0.043 2.3
Wollastonite 40.9 - - - - 0.035 1.9
Portlandite 33.2 - - - - 0.080 4.4
Ca2SiO4 52.5 - - - - 0.0.138 7.5
Hatrurite 89.2 - - - - 0.028 1.5
Ca4Al2Fe2O10 52.8 - - - - 0.085 4.6
Ettringite 130.3 - - - - 0.005 0.3
CSH0.8 356.2 - - - - 0.129 7.0
CSH1.2 356.2 - - - - 0.129 7.0
CSH1.6 356.2 - - - - 0.129 7.0
Friedel's salt 123.1 - - - - 0.000 0.0
b. From Koukouzas et al. (2017), reference hydrated cement.

Table 6-10. Michigan Basin solid equilibrium phases.
Vmin Limestone Dolomite A2 Carbonate Cement
Mineral % % %
cm3/mol Moles Moles % Comp. Moles Moles
Comp.a Comp. Comp.a
Calcite 36.9 0.825 153.3 0.008 1.1 - - 0.0777 4.2
Dolomite 64.7 0.090 16.7 0.972 136.7 0.598 1.9 - -
Pyrite 24.5 0.005 0.9 - - - - - -
Illite 256.2 0.040 7.4 - - 0.015 0.5 - -
K-Feldspar 108.7 0.010 1.9 - - 0.003 89.6 - -
Anhydrite 45.8 - - 0.013 1.8 0.371 55.5 - -
Quartz 23.1 0.030 5.6 - - 0.013 5.5 - -
Halite 27.0 - - 0.008 1.1 - - - -
Aragonite 34.2 - - - - - - 0.118 6.4
Vaterite 39.4 - - - - - - 0.043 2.3
Wollastonite 40.9 - - - - - - 0.035 1.9
Portlandite 33.2 - - - - - - 0.080 4.4
Ca2SiO4 52.5 - - - - - - 0.0.138 7.5
Hatrurite 89.2 - - - - - - 0.028 1.5
Ca4Al2Fe2O10 52.8 - - - - - - 0.085 4.6
Ettringite 130.3 - - - - - - 0.005 0.3
CSH0.8 356.2 - - - - - - 0.129 7.0
CSH1.2 356.2 - - - - - - 0.129 7.0
CSH1.6 356.2 - - - - - - 0.129 7.0
Friedel's Salt 123.1 - - - - - - 0.000 0.0
a. From Koukouzas et al. (2017), reference hydrated cement.
Equilibrium Model
The results of the Appalachian Basin equilibrium model are shown in Figure 6-1. The stability of most
cement minerals, as measured by log SIs, increases from the minimum brine value to the midpoint brine
value. The increase in the log SI value from the midpoint to the maximum value is not as large, even
though the change in parameters was the same between the minimum and midpoint values as the midpoint
and maximum values. This indicates that the change begins to level off as the parameters increase in
value. The most stable cement mineral phases are calcite, aragonite, and vaterite, which are at or near
saturation in the midpoint and maximum values and have log SI values less than -3 in the minimum
sample. The cement mineral SIs calculated using the summary data for the Michigan Basin are presented
in Figure 6-2. Calcite, aragonite, and vaterite are all saturated in the minimum, midpoint, and maximum
values. This is likely due to the buffering capacity provided by the defined alkalinity.

Minimum Midpoint Maximum
20
10
0
Log Saturation Indices
-10
-20
-30
-40
-50
-60
-70 Calcite Aragonite Vaterite Wollastonite Portlandite

Ca2SiO4 Hatrurite Ca4Al2Fe2O10 Ettringite CSH0.8
CSH1.2 CSH1.6 Friedel's Salt
Figure 6-1. Log SI values for cement mineral phases in

initial Appalachian Basin summary solutions.
Minimum Midpoint Maximum

20
10
Log Saturation Indices
-10
-20
-30
-40
Calcite Aragonite Vaterite Wollastonite Portlandite

-50
Ca2SiO4 Hatrurite Ca4Al2Fe2O10 Ettringite CSH0.8

initial Michigan Basin summary solutions.
Equilibrium Phases Model

SIs provide a measure of whether a mineral is saturated with respect to the activity of its components and
the overall ionic strength of the solution. It does not, however, provide the amount of the mineral that will
precipitate from a solution. In some cases, an SI that indicates that a mineral is saturated does not equate
to mineral precipitation because other minerals may precipitate first. For instance, aragonite and calcite
have the same chemical formula (CaCO3) and often have similar SIs in a given solution; however, calcite
may precipitate preferentially and pull Ca2+ and CO32- out of solution. After calcite precipitates to achieve
equilibrium, aragonite may be undersaturated. The equilibrium phases model was designed to determine
the amount of the specific mineral species that would precipitate in each sample solution.

The mineral phases that precipitated in the Appalachian Basin summary solutions are listed in Table 6-11.
Four mineral phases precipitate from the maximum sample solution: anhydrite (CaSO4), diaspore
(α-AlO(OH)), dolomite (CaMg(CO3)2), and hematite (Fe2O3). Diaspore and hematite also precipitate from
the minimum and midpoint solutions. When 1.1 mol of CO2 is added to the solutions, the amount of each
of these minerals that precipitates from each solution decreases, except for hematite in the midpoint
solution, which remains constant. Dolomite in the maximum solution and diaspore in each summary
solution do not precipitate when CO2 is added. In contrast, hematite and anhydrite are not greatly affected
by the addition of CO2.
Table 6-11. Amount of each mineral phase that precipitates from Appalachian Basin summary
solutions, with and without 1.1 mol of CO2. Precipitates are reported in mol/kg water. Columns
labeled “Δ with CO2” show the change in precipitation when the solution is reacted with CO2.
Dolomite-
Solution Anhydrite Diaspore Hematite
Anhydrite Diaspore ordered Dol-ord Hematite
Minimum --- Δ with 2.00E-5 Δ with --- Δ with 3.34E-8 Δ with
Midpoint --- CO2 7.83E-5 CO2 --- CO2 8.51E-4 CO2
Maximum 3.64E-3 1.30E-4 3.91E-4 8.15E-5
Min. w. CO2 --- --- --- -2.00E-5 --- --- 1.19E-8 -2.15E-8
Mid. w. CO2 --- --- --- -7.83E-5 --- --- 8.51E-4 -0-
Max. w. CO2 3.57E-3 -6.5E-5 --- -1.30E-4 --- -3.91E-4 8.14E-5 -1.00E-7
The mineral phases that precipitated in the Michigan Basin summary solutions are listed in Table 6-12.
Three mineral phases precipitate from all three summary solutions: diaspore (α-AlO(OH)), dolomite
(CaMg(CO3)2), and hematite (Fe2O3). When 1.1 mol of CO2 is added to the solutions, the amount of each
of diaspore and dolomite that precipitates from each solution decreases. The amount of hematite that
precipitates did not change when CO2 was added. The amount of diaspore that precipitated after the
addition of CO2 decreased by about a third in the minimum solution and by less than 3% in the midpoint
and maximum samples. Dolomite did not precipitate in the minimum solution after the addition of CO2
and decreased by around an order of magnitude in the midpoint and maximum samples.
Table 6-12. Amount of each mineral phase that precipitates from Michigan Basin summary
solutions, with and without 1.1 mol of CO2. Precipitates are reported in mol/kg water. Columns
labeled “Δ with CO2” show the change in precipitation when the solution is reacted with CO2.
Dolomite-
Solution Diaspore Hematite
Diaspore ordered Dol-ord Hematite
Minimum 2.92E-6 Δ with 1.22E-3 Δ with 6.79E-5 Δ with
Midpoint 1.80E-5 CO2 2.24E-3 CO2 2.33E-3 CO2
Maximum 3.31E-5 3.79E-3 3.90E-3
Min. w. CO2 --- -2.92E-6 --- -1.22E-3 6.77E-5 -2.00E-7
Mid. w. CO2 --- -1.80E-5 --- -2.24E-3 2.33E-3 -0-
Max. w. CO2 --- -3.31E-5 --- -3.79E-3 3.90E-3 -0-
6.3 Review of Well Defects and Pore Network Dimensions for Field Sites
Field data from the Michigan Basin site were evaluated to determine the range of well defects and pore
network dimensions in relation to potential for CO2 migration. Several researchers have analyzed the
relationship of fracture aperture diameter and cement sealing potential (Carroll et al., 2016). Pore throat
radius of rock core samples, fracture width of reservoir/caprock, and cement features were evaluated to
determine potential for gas migration.

MICP Pore Throat Radius. Mercury injection capillary pressure (MICP) analysis data were evaluated to
estimate pore throat radius for the reservoir and caprock zones. These features may be pathways for
upward CO2 migration around the borehole. MICP is a method of characterizing the capillary-pressure
behavior of subject rock formations for determining caprock sealing potential, fluid saturations,
permeability, wettability, and capillarity properties. This analysis is based on capillary law in which liquid
penetration into small pore systems is a function of surface and interfacial liquid tensions, pore throat size
and shape, and the wetting properties of the rock, which may be described for a nonwetting liquid by the
Washburn (1921) equation below:
𝑟𝑐 = −2𝛾𝑐𝑜𝑠𝜃/𝑃𝑐
Where Pc is capillary pressure (dynes/cm2), ϒ is surface tension of Hg, θ is the contact angle of mercury
in air, and rc is the radius of the pore throat aperture (µm) for a cylindrical pore.
The Dover 9-33 well at the Michigan Basin site was evaluated based on 16 MICP experiments conducted
at various depths within the well. Comprehensive elements of pore architecture vary according to the
depositional environment and facies-specific diagenetic modifications; however, these elements generally
remain consistent within lithologies for most of the cored intervals observed, rendering lithology-based
MICP results generally consistent. Samples described were tight mudstone, vuggy mudstone, skeletal
wackestone, skeletal packstone, stromatoporoid framestone, and stromatoporoid-tabulate coral
framestone. Table 6-13 summarizes descriptive statistics performed on pore throat radii distributions for
each lithology. Pore throat distributions between major representative lithologies vary between three
orders of magnitude but statistically are not much different.
Table 6-13. Descriptive statistics of MICP-derived pore throat radius distribution
for lithologies at the Dover 33 reef field.
Pore Throat Radius Mudstone Skeletal Skeletal Stromatoporoid Stromatoporoid
Statistic Vuggy Tight Wackestone Packstone Framestone Coral Framestone
Mean Radius (µm) 8.83 8.81 8.87 8.84 8.84 8.74
Standard Dev. (µm) 21.90 21.80 22.01 21.89 21.89 21.47
Range (µm) 133.00 133.00 135.00 135.00 134.998 131.00
Minimum (µm) 0.00196 0.00196 0.00196 0.00196 0.00196 0.00196
Maximum (µm) 133 133 135 135 135 131
Kurtosis 15.64 15.67 15.96 16.07 16.07 15.50
Skewness 3.74 3.74 3.76 3.77 3.77 3.71
n 85 85 85 85 85 85
Image Log Fracture Characterization. Image logs were evaluated to estimate the dimensions of induced
and natural fractures around the borehole at the Michigan Basin site. These fractures may be pathways for
CO2 migration and carbonation. Electrical and acoustic image logs are geophysical wireline tools for
describing features around the borehole. Fracture width was measured by hand from a physical copy of
the resistivity formation image log (Baker Hughes, Star tool) from the Lawnichak 9-33 well of the Dover
33 reef field. When fractures with well-defined dimensions were identified, the apparent width of the
fractures was measured by hand in centimeters and converted into degrees to calculate the scaled apparent
aperture. With track width spanning the full 360 degrees of the borehole, features may be measured by
hand or electronically and converted into degrees. Degrees may then be converted into apparent aperture
using the following equation:
𝐴⁰(2𝜋𝑟)
𝐴′ = 360
Where A' is the apparent fracture aperture, A⁰ is the measured feature width in degrees, and r is the
borehole radius (obtained from caliper).

Fractures were identified within both caprock and reservoir intervals. Ten fractures were determined to be
suitable for aperture analysis; seven were identified within the caprock interval, and the remaining three
were within the reservoir interval. Fracture aperture data from these features are summarized in
Tables 6-14. The geometric mean of caprock fracture aperture was 46.1 mm and the arithmetic mean was
51 mm, with 95% confidence lower and upper bounds of 31 mm and 71 mm, respectively. The geometric
mean of reservoir fracture aperture was found to 110.9 mm and the arithmetic mean was 119 mm, with
95% confidence lower and upper bounds of 16.5 and 254.5 mm, respectively.
Table 6-14. Summary of aperture data derived from fractures located

within the caprock interval of the Dover 33 FMI log.
Width
Caprock fracture ID
Measured (cm) Circumferential (deg) Calculated (mm)
1 0.1 9.5 35.7
2 0.2 19.0 71.4
3 0.1 9.5 35.7
4 0.15 14.2 53.6
5 0.05 4.7 17.9
6 0.2 18.5 71.4
7 0.2 18.5 71.4
Geometric mean - - 46.1
Arithmetic mean - - 51.0
Confidence Lower bound - - 31.0
level (95%) Upper bound - - 71.0
Width
Reservoir fracture ID
Measured (cm) Circumferential (deg) Calculated (mm)
8 0.2 18.9 71.4
9 0.3 28.4 107.1
10 0.5 47.4 178.5
Geometric mean - - 110.9
Arithmetic mean - - 119.0
Confidence Lower bound - - 16.5
level (95%) Upper bound - - 254.5
CBL Features. Defects in cement were also evaluated as pathways for CO2 migration. In 2009, a CBL
mapping tool (the Schlumberger Isolation Scanner) was run through the intermediate casing section of the
State Charlton 4-30 well of the Charlton 30/31 fields in Otsego County, Michigan. The isolation scanner
utilizes ultrasonic imaging technologies in combination with flexural wave imaging to yield a full
azimuthal profile of the borehole. This profile can be used to identify cement bond integrity, possible
channels or fractures within cement, casing integrity, and degree of zonal isolation between the formation
and a wellbore. This study utilized the Isolation Scanner to measure and calculate the apparent aperture of
several cement channels observed within the State Charlton 4-30 well.
Cement features like channels or fractures were identified within the Charlton 4-30 well by noting sharp
changes in flexural attenuation where a low value of attenuation appeared. The presence of channels or
fractures was verified using the Solid Liquid Gas Map, where interpreted cement features would be
saturated with either gas or water. The apparent width of the features was measured by hand using a ruler
and a physical copy of the log. With track width spanning the full 360 degrees of the borehole, features

may be measured by hand or electronically and converted into degrees. Degrees may then be converted
into apparent width using the following equation:
𝐴⁰(2𝜋𝑟)
𝐴′ = 360
Where A' is the apparent width, A⁰ is the measured feature width in degrees, and r is the production
casing radius.
Four channels/fractures were interpreted to be present within the Charlton 4-30 well, located at 2,552 ft
MD, 2,802 ft MD, 3,014 ft MD, and 3,052 ft MD. A cement channel was identified by its low attenuation
and borehole-parallel shape from 2,552 ft to 2,562 ft. This feature has a measured maximum width of
61 degrees and a calculated maximum width of 106 mm. A large irregular feature interpreted to be a
cement channel was identified from 2,802 ft to 2,828 ft based on its irregular shape, low attenuation, and
gas saturation. The feature’s maximum width was measured as 247 degrees, and its maximum width was
calculated to be 425 mm. A small cement channel/feature was identified from 3,014 ft to 3,017 ft based
on its low attenuation, partial gas-water saturation, and distinct ellipsoid shape elongated perpendicular to
the borehole. The feature’s measured maximum width was 154 degrees, and its calculated maximum
aperture was 270 mm. The fourth and final cement fracture/channel was identified from 3,052 ft to
3,080 ft; it was identified by its low flexural attenuation, distinct shape, and gas saturation (Figure 6-3).
The feature’s measured maximum width was 123 degrees, and its calculated maximum aperture was
210 mm. Over the entire intermediate casing interval, the mean aperture size of cement features was
calculated to be 252 mm, with a standard deviation of 6.65 mm and 95% confidence upper and lower
bounds of 464 mm and 41 mm, respectively (Figure 6-4).
Figure 6-3. Cement channel/fracture identified within the Charlton 4-30 well from
3,052 to 3,080 ft, highlighted by the yellow circle. Note the low attenuation (blue color)
and its gas saturation (red) illustrated on the Solid Liquid Gas Map (right-most track).

Figure 6-4. Distribution of measured calculated defects (mm) within the
intermediate casing interval of the Charlton 4-30 well.
6.4 Key Findings of Field Analysis of CO2 Cement Sealing and Well Integrity
Appalachian Basin, Michigan Basin, and Williston Basin field test sites were evaluated for cement sealing
conditions based on mineralogy, brine geochemistry, cement chemistry, hydrologic conditions, and CO2
exposure. The analysis was completed because SCP testing and well history review on CO2 wells
suggested that well defects were not a common problem at these field sites. Therefore, the sites were
investigated to determine if cement carbonation would potentially seal well defects, limiting gas
migration and casing pressure buildup in the CO2 wells. The results provide a better understanding of
wellbore integrity effects for CO2 storage applications as listed below:
Analysis of Subsurface Setting for Cement Sealing. Thin sections, well materials, cement properties,
and hydrologic, temperature, pressure, brine chemistry, and saturation conditions were summarized for
the pertinent CO2 reservoir zones and caprocks to provide a basis for further analysis of CO2 sealing
conditions in the Appalachian Basin, Michigan Basin, and Williston Basin field sites. Key findings of the
subsurface parameters are as follows:
• The field sites have wells exposed to CO2 at depths of 1,000 to 7,000 ft, pressures of 2,000 to
3,000 psi, temperatures of 105 °F to 145 °F, and CO2 exposure durations of 5 to 50 years.
• Reservoirs were sandstone and carbonates with shale and evaporite caprocks. The reservoir zones
contain a fairly large portion of quartz in the sandstone and dolomite in the carbonate reservoirs,
with a small amount of trace minerals.
• Brine geochemistry includes high salinity at the Appalachian Basin and Michigan Basin sites.
• Standard Portland Class A cement and carbon steel well casing were used for well materials at all
three field sites.
Together, these findings help define the subsurface system for the field sites in terms of interactions
between cement, brine, minerals, and CO2 in the subsurface.
Geochemical Analysis to Predict Cement Sealing Conditions. Information compiled on the field sites
was utilized to analyze geochemical interactions in relation to cement sealing conditions and well defects
for the three field sites with geochemical model PHREEQ-C. Key conclusions of the geochemical
analysis are as follows:

• Geochemical modeling of the subsurface systems was completed using a combination of
equilibrium, equilibrium phase, and batch reaction models.
• Overall, it appears that the cement would be the most reactive component of the system, similar
to results observed in other research on cement-CO2 interactions in the subsurface.
• Model results suggest that carbonation reactions occur between the cement and brine that
potentially seal defects.
Analysis of MICP and Wireline Data to Evaluate Potential Gas Migration. Potential gas migration
features in the Michigan Basin field site were analyzed to estimate potential for gas migration through
pore network, natural fractures, and cement around the borehole.
• MICP tests on rock core had mean pore throat radius of 8.8 μm and maximum of 135 μm.
• Natural features observed in image logs from reef wells had mean aperture of 46 mm in the
reservoir and 110 mm in the caprock.
• Features observed in cement logs from the Michigan Basin site were irregular voids, partial
channels, and gaps. These features appeared to be 200 to 500 mm in size, but they were isolated
features that were not connected along the borehole.
• Overall, it appears that the features observed at the Michigan Basin field site would not be
pathways for CO2 migration. Pore throat network dimensions would fall into the category that
would be sealed based on research by Carroll et al. (2016), which suggests that fractures less than
~200 μm would be sealed by cement mineralization. The natural fractures and cement channels
observed in the Michigan Basin field site were in the range of 46 to 500 mm, but they appeared to
be isolated features that would not be gas migration pathways.

Chapter 7.0 Wellbore Integrity Sealing Factor
Uncertainty Analysis
Ranges of formation geothermal gradients, water chemistry, subsurface pressure, and mineralogy were
evaluated with geochemical models to determine if there are suitable indicators for cement sealing
conditions in the subsurface. Potential indicators were analyzed using a CO2 batch reaction model and the
solid equilibrium phase model. Geochemical sealing conditions in the subsurface were also evaluated for
four test study areas with meta-modeling techniques.
7.1 CO2 Batch Model Indicator Geochemical Analysis

CO2 batch model scenarios were run with different CO2 levels to evaluate the effect of the addition of
CO2 in relation to various parameters. The changes in pH and pe when CO2 is added to the Appalachian
Basin midpoint summary solution are shown in Figure 7-1. The reduction in pH due to the addition of
CO2 is initially large (i.e., a 32% decrease with the addition of 0.01 mol of CO2) and continues to decrease
at slower rates with the addition of more CO2 until it reaches a minimum of 2.27 after 27 moles of CO2
are added. At this point, the pH begins to increase, slowly at first (0.02% per mole added) until the model
ends when the pH has reached a pH of 2.62 after 49 moles of CO2 is added.
The SIs of cement minerals when CO2 is added to the Appalachian Basin midpoint summary solution are
shown in Figure 7-2. The addition of CO2 decreases the SIs of all cement mineral phases, although at
different scales. For instance, the most stable mineral phases found in the equilibrium model, calcite,
aragonite, and vaterite, decreases from near saturation to around -2.5 to around -3.1 in brine that is 48.5
molal CO2. Other mineral phases, on the other hand decrease more dramatically, similar to the pH batch
reactions.
Figure 7-1. pH and pe of the Appalachian Basin summary solutions, CO2 batch reactions.
Reference pH and pe are simulated for CO2 added to pure water for comparison.

Appalachian Basin summary solutions, CO2 batch reactions.
Solid Equilibrium Phases Model

The change in precipitation of mineral phases with the addition of CO2 to the Appalachian Basin midpoint
brine summary and solid equilibrium phases is shown in Figure 7-3. With the addition of 0.01 moles of
CO2, quartz, illite, and hematite precipitate and albite and siderite dissolve. Pyrite and K-feldspar are at
equilibrium. The amount of these minerals that precipitate with the addition of more CO2 continue to
increase the amount of these minerals that precipitate and dilute until four moles of CO2 are added, at
which point quartz, illite, and albite do not change with the addition of CO2. Prior to the addition of
4 moles of CO2, the pH and pe were buffered from the effects of the CO2. Like the Tuscarora, the Rose
Hill also provides a buffer for changes in pH and pe. In the cement model, Calcite, wollastonite, Friedel’s
salt, and portlandite with the addition of 0.01 moles CO2 while vaterite, dicalcium silicate, and aragonite,
hatrurite, CSH0.8, CSH1.2, CSH1.6 all dissolve completely. With the addition of CO2, portlandite begins to
dissolve at a relatively constant rate (around 1 mole per mole of CO2 added) until it is depleted after 20
moles of CO2 are added, after portlandite is consumed, Ca4Al2Fe2O10 begins to dissolve at a rate of around
0.3 moles per mole of CO2 added until it is consumed after 29 moles of CO2 are added. Only two solid
phases remaining after the batch reactions: calcite 61.4 moles, a net increase of 48.5 moles) and
wollastonite (23.2 moles, a net increase of 21.3 moles).

Tuscarora Sandstone
Rose Hill Formation
Cement Phases
from previous CO2 batch step) and moles of mineral phases with midpoint summary solution
for Tuscarora sandstone, Rose Hill formation, and cement phases.
Michigan Basin
The change in precipitation of mineral phases with the addition of CO2 to the Michigan Basin midpoint
brine summary for the State Chester limestone, Dover 33 dolomite, Dover 33 A2 Carbonate, and cement
mineral solid equilibrium phases is shown in Figure 7-4. K-feldspar dissolves and quartz precipitates as
CO2 is added. The remaining minerals are at or near equilibrium, although, at a lower rate, dolomite
dissolves and calcite precipitates. In the Dover 33 dolomite, the dolomite precipitates and calcite and
halite dissolve with the addition of CO2. The minerals are then at or near equilibrium as CO2 nears
saturation. The remaining calcite then begins to dissolve (around 0.2 mol calcite dissolve in the final step
of the batch reaction). The Michigan Basin midpoint brine summary and Dover 33 A2 Carbonate solid
equilibrium phases shows that illite and quartz precipitate and K-feldspar dissolves. All minerals then
reach a state of equilibrium when 6.0 moles of CO2 added. The changes in the Michigan Basin cement
with the addition of CO2 are similar to the cement solid equilibrium cement phases analysis for the
Appalachian Basin, indicating that the initial brine solution is not an important factor when comparing the
midpoint values of the Michigan and Appalachian Basins.

State Chester Limestone
Dover 33 Limestone
Dover 33 A2 Carbonate
Cement Phases
from previous CO2 batch step) (top) and moles of mineral phases (bottom), State Chester
limestone, Dover 33 limestone, Dover 33 A2 carbonate, and cement mineral phases
with midpoint summary solution. The reference pH and pe (Ref. pH and Ref. pe)
were found by reacting the CO2 batch with pure water alone.

7.2 CO2 Reaction Batch Model Indicator Geochemical Analysis
The Appalachian Basin was used as the example for the batch reactions model. Results from model
scenarios were used to evaluate variation in pH, pe, pressure, and temperature on geochemical reactions.
Figure 7-5 shows the SIs for cement minerals in the Appalachian Basin when the minerals are exposed to
varied pH, temperature, and pressure conditions. For pH, the SIs for cement minerals in the midpoint
summary solution show that the most stable cement mineral is calcite, which is above saturation until
under pH 9. For temperature, the SIs for cement minerals in the midpoint summary solution show that all
cement minerals become more stable (move closer to saturation) as the temperature increases, with one
exception: ettringite, which has a log SI value of -13.9 at 35oC and decreases to -16.5 at 90oC. The most
stable minerals are calcite and aragonite; their log SI values increase from -0.4 at 25oC and become
supersaturated at 80oC. Finally, the SIs for cement minerals exposed to the pressure batch reaction show
that pressure does not affect the log SI values or precipitation of cement mineral phases in any
appreciable way. The addition of CO2 does not affect these results.
Figure 7-5. Log SI values for cement minerals batch reaction for
pH, temperature, and pressure, Appalachian Basin.

7.3 Geochemical Analysis Meta-Modeling
Subsurface conditions, well construction specifications, and storage zone/caprock mineralogy were
assessed for four test study areas. This information was identified to aid in the meta-modeling of cement
sealing conditions across actual oil and gas fields. Four test study areas were selected for further analysis
of cement sealing conditions (Figure 7-6):
• a 7- x 7-kilometer (km) area in the vicinity of Calhoun County, Michigan,
• a 6- x 6-km area in the vicinity of St. Clair County, Michigan,
• a 15- x 15-km area in the vicinity of Muskingum-Coshocton County, Ohio, and
• a 15- x 15-km area in the vicinity of Trumbull County, Ohio.
The objective of this analysis was to examine the uncertainty and variations in subsurface well integrity
and cement sealing conditions based on actual well materials, hydrologic conditions, and geologic setting.
The test study areas were selected because they have many legacy oil and gas wells and are located near
large CO2 sources. Therefore, they present useful and realistic test study areas to examine the
ramifications of wellbore integrity for CO2 storage applications.
Table 7-1 summarizes the test study area parameters. The size of each test study was determined by
estimating the size of the CO2 plume after injection of 70 million metric tons of CO2 into the most
suitable storage zone using a simple volumetric calculation. 70 million metric tons equates to 3.5 million
metric tons of CO2 per year emitted from a typical 500-megawatt coal-fired power plant for 20 years.
Well construction specifications were described for these sites in a previous DOE-NETL project
(FE0009367) examining wellbore integrity (Sminchak et al., 2016). This database was used as a starting
point for the cement sealing uncertainty analysis. Test study areas are described in more detail below.
Figure 7-6. Map showing four test study areas.

Table 7-1. Summary of four test study areas.
Test Study Area
Parameter
1 2 3 4
Location Calhoun Co., MI St. Clair Co., MI Muskingum Co., OH Trumbull Co., OH
Size (km) 7x7 6x6 15 x 15 15 x 15
Reservoir Depth (m) 1,707 975 2,150 2,200
Thickness (m) 100 122 35 35
Reservoir Type Sandstone Niagaran Reefs Carbonates Carbonates
Caprock Shale Evaporites Shale Shale
Porosity (%) 12 12 6.5 6.5
Temperature (°C) 44 31 53 57
Pressure (MPa) 17.2 10.1 23.3 23.8
Salinity (mg/L) 225,000 350,000 250,000 300,000
# Wells 22 155 1,221 357
Test Study Area 1. Test study area 1, located in the vicinity of Calhoun County, Michigan, is 7 x 7
km based on the parameters listed in Table 7-1, which were selected using wireline data and literature.
The study area encompasses 22 oil and gas wells that primarily target the Trenton-Black River
formations in the Albion-Scipio play. The potential storage zone in this study area is the Mt. Simon
sandstone, with the Eau Claire shale as the primary confining layer.
Test Study Area 2. Test study area 2, located in the vicinity of Saint Clair County, Michigan, was
estimated as 6 x 6 km based on the parameters listed in Table 7-1, which were selected using wireline
data and literature. The test study area encompasses 155 oil and gas wells that primarily target the
Niagara reef system. These Niagaran Reefs are the primary target for CO2, and this site would be a
CO2 EOR scenario where the depleted reefs would be filled with CO2. The reefs are fairly shallow at
approximately 975 m depth.
Test Study Area 3. Test study area 3, located in the vicinity of Muskingum County, Ohio, was
estimated as 15 x 15 km based on the parameters listed in Table 7-1, which were selected using
wireline data and literature. The test study area has 1,221 oil and gas wells that primarily target the
Clinton-Cataract group. There are 12 large CO2-emitting facilities nearby. The targeted storage
formations are the Copper Ridge dolomite down to the basal sandstone, with the Queenston shale to
the Black River group as the confining layers.
Test Study Area 4. Test study area four, located in the vicinity of Trumbull County, Ohio, was
estimated as 15 x 15 km based on the parameters listed in Table 7-1, which were selected using
wireline data and literature. The study area encompasses 357 oil and gas wells that primarily target the
Cataract-“Clinton” sandstone group. The potential storage zone in this study area is the Copper Ridge
dolomite down to the basal sandstone, with the Queenston shale to Black River group as the confining
layers.

Well materials were reviewed for each test study area to determine the general well cement, casing, and
additives used to complete wells in each study area. Table 7-2 summarizes well materials for the four
study areas. As shown, there are 22 wells in test study area 1, 155 wells in test study area 2, 1,221 wells
test study area 3, and 357 wells in test study area 4. Some of the wells date back to the 1920s, but most of
the wells are 1970-1980s vintage. Standard casing programs and class A Portland cement were used for
well construction, as is the typical practice in the Midwest United States. Additives include Pozmix,
CaCl2, and lost circulation material (LCM).
Table 7-2. Summary of test study area well construction specifications.

Test Study Area
Parameter
1 2 3 4
Location Calhoun Co., MI St. Clair Co., MI Muskingum Co., OH Trumbull Co., OH
# Wells 22 155 1,221 357
Storage reservoir 1,707 975 2,150 2,200
depth (m)
Oil and gas Albion-Scipio Niagaran Reefs Berea, Clinton-Medina, Clinton-Medina
reservoirs (Trenton) Rose Run
General casing Shallow 8 5/8” Shallow 8 5/8” Shallow 8 5/8” Shallow 8 5/8”
program Deep 5 1/2” Deep 5 ½” Deep 4 ½” Deep 4 ½”
Cement Class A Portland Class A Class A Class A
(shallow)
Class A Portland
50/50 Pozmix (deep)
Cement additives 2% CaCl2, LCM Pozmix, 2-4% Pozmix Pozmix, CaCl2,
CaCl2 gilsonite
Well construction specifications were also tabulated for wells at the four test study areas. Well locations,
completion date, total depth, casing schedule, cementing information, plugging details, and well status
details were compiled. The thickness of the cement in the production casing was also mapped out for test
study areas, because this directly relates to potential CO2 migration from the storage zone along the
boring. In addition, the thickness of the plugs was mapped out for plugged and abandoned wells.
Figure 7-7 shows maps of cement thickness outside of the production casing string. As shown, all sites
generally have 100 to 200 ft production casing cement, and many wells were cemented more than 500 ft
into overlying casing string. Production zone plugs are less thick and have a greater variation.
Subsurface Conditions. Subsurface conditions for the test study areas were determined for
caprock/reservoir mineralogy, pressure, temperature, salinity, and pH. This provides a range of conditions
to be examined with the uncertainty analysis and meta-modeling. Mineralogy was based on regional
trends in oil and gas reservoirs (Roan et al. 1996), oil and gas drilling records, and research on CO2
storage. Table 7-3 summarizes the general mineralogy for the caprock and reservoir zones for each study
area. Overall, the mineralogy would be expected to be fairly consistent across these local test study areas.
Brine geochemistry was based on regional studies on brine samples from the reservoir zones for each site
(Table 7-4). As shown, the sites have high salinity and moderate pH, which is typical for Paleozoic-age
rocks in the Midwest United States.

Figure 7-7. Map of production casing cement thickness across test study areas.

Table 7-3. Summary of general mineralogy for test study areas.
Test Study Area 1 Test Study Area 2 Test Study Area 3 Test Study Area 4
Mineral Caprock Reservoir Caprock Reservoir Caprock Reservoir Caprock Reservoir
Shale Sandstone Evaporites Carbonates Shales Dolomites Shales Dolomites
Dolomite 12% <0.1% 40-99% 9-97% 3% 50% 3% 50%
Calcite 8% <0.1% <0.1% 0.8-82% 95% 0.3% 95% 0.3%
Quartz 32% 72% 0.3-2% 0-1.5% 1% 43% 1% 43%
Anhydrite <1% 1% 0-56% 1.3-4% <0.1% <1% <0.1% <1%
K-Feldspar 36% 22% 0-0.6% 0.5% <0.1% 0.5% <0.1% 0.5%
Plagioclase <1% <0.1% <0.1% <0.1% 0.6% 0.6 0.6% 0.6
Clays 9% 4% <0.1% 0-2.6% 0.7% 2.5% 0.7% 2.5%
Other <0.1% <0.1% 0.3%pyrite 0.3%pyrite
Source Gupta et al., 2004 MRCSP, 2004 Wickstrom et al., 2011 Wickstrom et al., 2011
Table 7-4. Summary of brine geochemistry for test study areas.

Constituent Test Study Area 1 Test Study Area 2 Test Study Area 3 Test Study Area 4
(mg/L) Low High Low High Low High Low High
Na 28,000 65,000 15,300 22,500 45,000 82,000 45,000 82,000
Cl 57,000 83,000 251,000 274,000 150,000 200,000 150,000 200,000
Ca 7,200 14,000 67,500 110,000 22,000 45,000 22,000 45,000
K 975 975 11,000 16,600 3,500 7,000 3,500 7,000
Fe 92 92 10 129 30 600 30 600
Mg 1,400 4,500 7,985 12,100 2,400 9,100 2,400 9,100
SO4 277 1,450 81 97 90 800 90 800
pH 5.6 7.9 4.8 5.9 5.6 7.5 5.6 7.5
HCO3- 23 23 361 468 25 260 25 260
Salinity 40,000 110,000 348,000 450,000 200,000 300,000 200,000 300,000
Source Sass et al., 1998 MRCSP sampling Gupta et al., 2004; Sass et al., 1998
To help visualize the variations in subsurface conditions for wells at the test study areas, maps were
prepared for temperature and pressure. Temperature and pressure were based on well depths in each study
area. Subsurface temperatures were estimated with mean ambient temperature and 1° F per 100-ft
gradient. Pressures were based on a 0.444 psi/ft pressure gradient. Figure 7-8 shows the estimated
subsurface pressure distribution in wells. Since the pressure and temperatures were based on well depth,
the maps mainly reflect depths of the wells.

Figure 7-8. Baseline pressure conditions at test study areas.
Meta-Model Development
To evaluate the cement sealing potential in wells at the four test study areas, meta-models were applied to
the sites based on subsurface pressure, temperature, and pH conditions present at the sites. The meta-
models portray how cement sealing conditions in the subsurface may vary across CO2 storage areas with
legacy wells. A meta-model is a model of a model, or a systematic method to portray a problem based on
metadata or input parameters. In this case, the metadata are the pressure, temperature, pH, fluid
chemistry, mineralogy, and cement makeup for the legacy wells. The meta-model approach assumes that
CO2 would be injected, contact the legacy wells near the well total depth, and introduce changes in
pressure, temperature, and pH (Figure 7-9). The meta-modeling predictions were constrained to
parameters in the space of interest. For example, the pressure ranges were constrained to initial pressure
and maximum injection pressure anticipated in the subsurface for each test study area. In this case, the
PHREEQ-C model output was represented by a proxy-model estimating CaCO3 saturation indices (SIs) at
various pressure, depth, and pH conditions. These scenarios represent conditions likely to be present in
the subsurface due to the injection of CO2.

Figure 7-9. Flow chart illustrating the meta-modeling process for evaluating
geochemical cement sealing conditions across the four test study areas.
The PHREEQC geochemical simulations results were processed to develop a proxy model to estimate
geochemical changes as a function of the depth, temperature, pressure, and introduced pH. The results
focused on the calcite SIs as an indicator of cement sealing potential. The proxy model was used to
develop a database of CaCO3 SIs for the wells at the four test study areas for different combinations of
pressure, temperature, and pH. For example, Figure 7-10 shows the meta-model output for the 357 wells
at test study area 4 at temperature +10 °C, pressure + 0%, and pH -2 from baseline conditions. The user
may adjust the ‘sliders’ for pressure, temperature, and pH to see how they affect the SI for the wells.
Thus, the meta-model is a way to visualize how differences in subsurface conditions may affect cement
sealing potential. In general, the meta-models showed only minor amount of variation in SI across the test
study areas. For example, the scenario shown in Figure 7-10 ranged from -4.9 to -5.2, a fairly small range.
Meta-models for the other test study areas show small changes for a combination of metadata. This
suggests that CO2-related cement sealing processes would not be sensitive to subsurface conditions.

Figure 7-10. Meta-modeling output for test study area 4
at pressure +0, temperature +10 °C, and pH -2.
7.4 Key Findings of Wellbore Integrity Sealing Factor Uncertainty Analysis

The sensitivity of the subsurface system was examined with geochemical models run over a range of
conditions (temperature, CO2 saturations, pressure, mineralogy, brine chemistry). Results of the modeling
were processed to identify key indicator parameters related to well-sealing conditions in the regional
basins (Appalachian, Michigan, and Williston). Key conclusions of the indicator analysis are as follows:
• Model results were not especially sensitive to the range of subsurface conditions. Temperature
affected the SIs of most cement mineral phases, but the pH and, to a lesser extent, pe were the
main factors investigated that influenced cement mineral phase reactions.
• The model showed that the largest effect of the addition of CO2 is a decrease in pH. This, in turn,
drives dissolution reactions of cement mineral phases.
• The addition of CO2 also increased the pe, which the batch reactions suggest makes cement
mineral phases more unstable (more undersaturated). The change in pe could also cause redox-
sensitive metals, particularly iron, to precipitate and potentially affect the porosity and
permeability of the cement or surrounding formation.
• The solid equilibrium phases model showed a logical progression of cement mineral dissolution,
beginning with portlandite and continuing with (in order) Ca4Al2Fe2O10, ettringite, and
wollastonite. Calcite precipitated in conjunction with these reactions (i.e., cement carbonation
was occurring). In addition, portlandite provided a pH buffer with the release of hydroxide ions
until it was fully dissolved, an effect suggested by Kutchko et al. (2007).
Meta-models were applied to the four test study areas in Michigan and Ohio to assess the relationship
between cement carbonation and subsurface conditions in typical wells in the region. These test study
areas had 22-1,221 legacy oil and gas wells. Thus, they provide a real-world example of conditions at

potential carbon storage applications. PHREEQC geochemical simulation results were processed to
develop a proxy model to estimate geochemical changes as a function of the depth, temperature, pressure,
and introduced CO2. Results used calcite SIs as an indicator of cement-sealing potential. The proxy model
was used to develop a database of CaCO3 SIs for the wells at the four test study areas for different
combinations of pressure, temperature, and pH. Results suggest that subsurface conditions would not
have a large effect on the carbonation of well cements and potential cement sealing.
Like many other studies, the results suggested that carbonation of well cements would occur in these
subsurface environments. It appears that mineralogy, hydrologic conditions, cement blends, and brine
geochemistry were not especially critical factors to the cement carbonation process, as they had minor
effect on the modeled saturation index. The largest effect on the subsurface was pH change triggered by
the dissolution of the injected CO2 into the subsurface and subsequent production of carbonic acid.
Results indicate that life-cycle effects of CO2 on well integrity are challenging to evaluate. It appears that
well construction procedures, well design, and well logging/testing for defects are important
considerations for wellbore integrity in CO2 environments in the subsurface.

Chapter 8.0 Reporting and Technology Transfer
The objective of the reporting and technology transfer task was to document project results and provide
project data to other CO2 research projects. Reporting and technology transfer activities included
preparation of task reports, technical meetings, project review meetings, and synergistic activities with
other carbon storage research projects.
8.1 Reporting
Summary reports were prepared for the major technical tasks and field work. Table 8-1 summarizes
reports generated and submitted to DOE-NETL. In addition, routine quarterly research performance
progress reports and financial reports were submitted to document technical and financial progress.
Table 8-1. Summary of deliverables.
Planned Submission
Task Milestone Description Deliverable
Due Date Date
1 Update Project Management Project Management Plan 30 days
October 6,
Plan after initial
2015
award
2 Complete Wellbore Integrity Well Integrity Registry Summary
June 2016 June 30, 2016
Registry Report
3 Collect Well Record Data Well Record Data Summary Report
June 2017 June 29, 2017
4 Complete Log & Testing Log & Testing Based Well Integrity
September November 1,
Based Well Integrity Assessment Summary Report
2017 2017
Assessment
5 Collect All SCP Analysis Compiled database of SCP
Data Analysis Data (uploaded to EDX
September 2018)
December
Appalachian Basin Field Testing March 2018
2017
Summary Report (March 2017)
Williston Basin Field Testing
Summary Report (February 2018)
6-7 Wellbore Integrity Sealing Wellbore Integrity Sealing Factor
June 2018 June 29, 2018
Factor Analysis Summary Report
8 Final Technical Report Final report with all project September September 27,
methods, results, and conclusions 2018 2018
8.2 Technology Transfer

Several technical presentations, posters, and papers were prepared under the project to communicate
project results to scientific community, stakeholders, and industry. A project overview presentation was
given by J.R. Sminchak at the DOE-NETL Carbon Storage Program Review Meeting in Pittsburgh,
Pennsylvania, August 16-18, 2016, August 12-15, 2017, and August 13-17, 2018. Summary Excel data
sheets were prepared for the well SCP tests. The data sheets contain the pressure-temperature time series,
well description, and well diagrams. The data sheets were uploaded to the DOE-NETL Energy Data
Exchange (EDX) website on September 5, 2018 (Figure 8-1). The project team also engaged with
researchers on the DOE-NETL National Risk Assessment Partnership (NRAP) program to provide

information on field data for NRAP tool validation for CarbonSAFE projects. Technology transfer items
are summarized as follows:
• Sustained Casing Pressure Testing of CO2 Wells for Wellbore Integrity Defects: Environmental
Risk Implications for Carbon Storage Projects, J.R. Sminchak and Matt Place, IEAGHG
Modelling and Risk Network Meeting, 19-22 June 2018, Grand Forks, North Dakota, USA.
• Battelle Progress on CO2 injection projects and legacy oil and gas wells, provided to IEAGHG
working group to prepare a Report on Well Engineering and Injection Regularity in CO2 Storage
Wells, January 2018.
• DOE-NETL Web Meeting- Project Update: Integrated Wellbore Integrity Analysis Program for
CO2 Storage Applications, J.R. Sminchak, A. Duguid, A. Haagsma, and M. Place, 2 March 2018.
• Case Study on Wellbore Integrity for Two Fields with Wells Exposed to CO2 in the Subsurface in
the Midwest U.S. Jacob Markiewicz, J.R. Sminchak, and Mark Moody. SPE Eastern Section
Regional Meeting, 4-6 October 2017, Lexington, Kentucky.
• Is your well flat or carbonated? What sustained casing pressure testing and beer have in
common. J.R. Sminchak. 11th IEAGHG Monitoring Network Meeting. June 13-15, 2017,
Traverse City, Michigan.
• Field Testing and Well History Analysis on Wells Exposed to CO2 in the Subsurface in the
Midwest U.S., J.R. Sminchak, Mark Moody, Autumn Haagsma, Andrew Duguid, Matt Place, and
Neeraj Gupta. Carbon Capture, Utilization, and Storage Conference, June 14-16, 2016, Tysons,
Virginia, USA.
• Sustained Casing Pressure Diagnosis with Extended Data Collection, Matt Place, Glenn Larsen,
Bryan Dotson, Nigel Jenvey, and Mark Moody, SPE Eastern Regional Meeting, 13-15 October
2015, Morgantown, West Virginia.
Figure 8-1. EDX data upload submission summary.

Chapter 9.0 Conclusions
Subsurface exposure to CO2 is a concern for wellbore integrity at CO2 storage sites, because CO2 can
corrode well materials and migrate along defects around the borehole. These processes may affect new
wells and legacy oil and gas wells. Consequently, this project completed a program to evaluate well
integrity in CO2 wells with a combination of direct field testing and analysis of well records. Project
results were used with geochemical analysis to identify trends that lead to better understanding and
prediction of well integrity issues for CO2 storage applications.
In this project, approximately 1,500 wells at three sites were reviewed in terms of well construction,
history of exposure to CO2, geochemistry, mineralogy, and well materials. The field sites included the
following locations:
• Appalachian Basin. The Appalachian Basin Indian Creek site is a natural CO2 and methane field
located in Kanawha County, West Virginia. The field contains approximately 58 wells at total
depths between 6,200 ft and 6,700 ft. The Indian Creek field produces in the Tuscarora sandstone.
The field has CO2 levels up to 60% in some areas.
• Michigan Basin. The Michigan Basin study site is located in the northern portion of the Niagaran
reef trend in Otsego County, Michigan. The fields were developed since the 1960s in the region,
and selected reefs have been subject to CO2 EOR since the 1990s. CO2 is also present in the
Antrim gas wells between 5% and 30% at depths between 1,000 ft and 1,500 ft. The Niagaran
reef EOR wells are completed at depths ranging from 5,000 ft to 7,000 ft.
• Williston Basin. The Williston Basin field is located on the northwestern edge of the Williston
Basin geologic feature and was a major oil play which is used for CO2 EOR. Approximately
3,000 wells are located in the Williston Basin testing site, completed at depth of 6,000 to 7,000 ft.
The Williston Basin testing site is a mature oil field that began production in 1954. CO2 EOR was
started in 2000 at the site, expanding to additional areas over time.
A total of 83 CO2 wells were surveyed at the Michigan Basin (23 wells) and Williston Basin (60 wells)
sites for wellhead casing pressures that may indicate well defects. The Appalachian Basin site was not
available for field testing, because the asset was sold to a new operator. Detailed SCP testing was
completed on 23 wells that had some indication of significant SCP.
The test results did not show significant well defects, with casing pressures less than 1 MPa and minor
pressure buildup patterns. The wells demonstrated zonal isolation, with no CO2 gas migrating to the
wellhead b-annulus. Analysis of wells with minor pressure buildup was inconclusive in terms of nature
and severity of the defects. These results were surprising, given that 15-20% of typical oil and gas wells
develop casing pressure. It was expected that more defects would be present in CO2 wells, because of
carbonic acid evolution and corrosive conditions in the subsurface.
Additional geochemical modeling and meta-modeling for the three field sites and four test study areas
were completed to determine if subsurface conditions at the field sites were suitable for cement sealing of
gas migration pathways via CaCO3 precipitation. Well construction and/or geochemical conditions for
cement carbonation appear to have contributed to well integrity. Results support management of CO2
storage applications in areas with many legacy oil and gas wells.
Some key conclusions of the integrated wellbore integrity analysis program include the following:
• The wellbore integrity registry developed in this project provides a catalog of the well
component, integrity issues, causes, timing, and leakage pathways that may occur in wells. Most
wellbore integrity problems are in the casing, cement, or interface between the two components
or arise due to geological processes such as formation lithology and geomechanical stresses.

• The three field sites have wells which have been exposed to CO2 either naturally or through EOR
operations for 5 to 50 years. The sites have different geologic settings and subsurface conditions.
These datasets provide unique opportunities to study the influence of CO2 on wellbore integrity.
• Three different fields were assessed using a risk ranking methodology. The fields, in Otsego,
Michigan, Indian Creek, West Virginia, and Alberta, Saskatchewan, varied in size, geography,
and geology. The assessment showed that Total Severity was more important in determining the
Total Risk.
• The field testing was completed on a subsample of wells and does not mean all CO2 wells would
be free of defects. In addition, the SCP testing methodology requires defects that would lead to
gas migration to the wellhead. Therefore, there may be existing downhole defects not revealed by
the testing.
• SCP testing is an effective, quick, direct, and low-cost method to test and monitor wellbore
integrity. It is a useful option for CO2 storage areas with many legacy oil and gas wells in lieu of
more expensive down-hole logging or well plugging.
• Geochemical analysis and modeling results suggested that carbonation of well cements would
occur in these subsurface environments. It appears that mineralogy, hydrologic conditions,
cement blends, and brine geochemistry were not especially critical factors to the cement
carbonation process.
• Results indicate that life-cycle effects of CO2 on well integrity are challenging to evaluate. It
appears that well construction procedures, well design, and well logging/testing for defects are
important considerations for wellbore integrity in CO2 environments in the subsurface.
Additional work on the life-cycle effects of CO2 would highlight changes over time due to subsurface
exposure to CO2 in wells. This life-cycle analysis may include periodic CBLs, sidewall cores, and SCP
testing. The analysis would help determine the rate, nature, and severity of well integrity effects over time
for CO2 wells.

Chapter 10.0 References
Aadnoy, B. S. (1990). In-Situ Stress Directions from Borehole Facture Traces. Journal of Petroleum
Science and Engineering, 4: 143-153
Avary, K. L. (1996). The Lower Silurian Tuscarora sandstone fractured anticlinal play. The atlas of major
Appalachian gas plays, Roen, J. B., & Walker, B. J. (Eds), West Virginia Geological and
Economic Survey Publication, 25: 151-155.
Azaroual, M., Durst, P., Czernichowski-Lauriol, I., Olsen, D., Stentoft, N., Springer, N., Rochelle, C. A.,
Pearce, J., Bateman, K., & Birchall, D. (2004). The geochemical reactions resulting from CO2
injection into the midale formation, Weyburn oilfield; A laboratory experimental and modelling
study. GHGT7: 7th International Conference on Greenhouse Gas Control Technologies,
Vancouver, Canada, 5-9 September 2004, 4 p.
Battelle. (2015). Development of Subsurface Brine Disposal Framework in the Northern Appalachian
Basin. Report prepared for Research Partnership to Secure Energy for America under contract
#11123-73, 412 p.
Battelle. (2016). Final Technical Report: Systematic Assessment of Wellbore Integrity for Geologic
Carbon Storage Projects Using Regulatory and Industry Information. Prepared for U.S.DOE-
NETL under contract DE-FE009367.
Bell, J. S., & Gough, D. I. (1979). Northeast-southwest compressive stress in Alberta: Evidence from oil
wells, Earth Planet, Sci. Lett., 45: 475-482.
Braunberger, J., Bremer, J., Liu, G., Gorecki, C., Peck, W., Steadman, E., & Harju, J. (2012).
Characterization, Petrography, and Static Modeling of an Unconventional Carbonate Reservoir:
Intervals of the Midale and Rival “Nessons” Beds in the Mississippian Madison Group, Burke
County, North Dakota. 012 AAPG Annual Convention and Exhibition, 22-25 April 2012, Long
Beach, California.
Breen, K., Angelo, C., Masters, R., & Sedam, A. (1985). Chemical and isotopic characteristics of brines
from three oil- and gas-producing sandstones in eastern Ohio, with applications to the
geochemical tracing of brine sources. Water-Resources Investigations Report 84-4314. Prepared
in cooperation with the Ohio Department of Natural Resources, Division of Oil and Gas.
Cantucci, B., Montegrossi, G., Vaselli, O., Tassi, F., Quattrocchi, F., & Perkins, E. H. (2009).
Geochemical modeling of CO2 storage in deep reservoirs: The Weyburn Project (Canada) case
study. Chemical Geology, 265: 181-197.
Carey, J. W. (2013). Geochemistry of Wellbore Integrity in CO2 Sequestration: Portland Cement-Steel-
Brine-CO2 Interactions. Reviews in Mineralogy and Geochemistry, 77(1): 505-539.
Carey, J. W., Wigand, M., Chipera, S., WoldeGabriel, G., Pawar, R., Lichtner, P., Wehner, S.,
Raines, M., & Guthrie Jr., G. D. (2007). Analysis and performance of oil well cement with
30 years of CO2 exposure from the SACROC Unit, West Texas, USA. International Journal of
Greenhouse Gas, 1(1): 75-85.
Carroll S., Carey. J. W., Dzombak, D., Huerta, N., Li, L., Richard, T., Um, W., Walsh, S., & Zhang, L.
(2016). Review: Role of Chemistry, Mechanics, and Transport on Well Integrity in CO2 Storage
Environments. International Journal of Greenhouse Gas Control, 49: 149-160.
Chamberlain, B., Chiesl, N., Day, J., Dowd, L., Overocker, B., Pape, D., Petrus, M., Swanson, M., &
Toles, J. (1995). Concrete - A Material for the New Stone Age. University of Illinois, Materials
Science and Technology, A MAST Module. Retrieved from <illinois.edu>.

Crow, W., Carey, J. W., Gasda, S., Williams, D. B., & Celia, M. (2010). Wellbore integrity analysis of a
natural CO2 producer. International Journal of Greenhouse Gas Control, 4: 186–197.
Duguid, A., Zaluski, W., El-Kaseeh, G., Lee, S., and Piercy, M.(2017), “Well integrity risk assessment to
inform containment risk monitoring for carbon capture, utilization, and storage, applied to the
Weyburn-Midale Field, Canada,” International Journal of Greenhouse Gas Control,
http://dx.doi.org/10.1016/j.ijggc.2016.10.018, 2017.
Duguid, A., Carey, J. W., & Butsch, R. (2014). Well Integrity Assessment of a 68 Year Old Well at a CO2
Injection Project. Proceedings of the 12th International Conference on Greenhouse Gas
Technologies. Austin, Texas, October 2014.
Duguid, A., Butsch, R., Carey, J. W., Celia, M., Chugunov, N., Gasda, S., Ramakrishnan, T. S., Stamp,
V., & Wang, J. (2012). Pre-injection Baseline Data Collection to Establish Existing Wellbore
Leakage Properties. Proceedings of the 11th International Conference on Greenhouse Gas
Technologies. Kyoto, Japan, September 2012.
Duguid, A., Radonjic, M., & Scherer, G. W. (2011). Degradation of cement at the reservoir/cement
interface from exposure to carbonated brine. International Journal of Greenhouse Gas Control,
5(6): 1413-1428.
Duguid, A., & Scherer, G. W. (2010). Degradation of Oilwell Cement due to Exposure to Carbonated
Brine. International Journal of Greenhouse Gas Control, 4: 546–560, 2010.
Durocher, K. E., Kotzer, T. G., & Whittaker, S. G. (2005). Physical and chemical characterization of
subterranean CO2 storage sites using synchrotron-based computed microtomography techniques.
Summary of Investigations 2005, V.1, Saskatchewan Geological Survey, Sask. Industry
Resources, Misc. Rep. 2005-4.1, CD-ROM, Paper A-3, 7 p.
Gasda, S. E., Bachu, S., & Celia, M. A. (2004). Spatial Characterization of the Location of Potentially
Leaky Wells Penetrating a Deep Saline Aquifer in a Mature Sedimentary Basin. Env. Geol.
46(6-7): 707-720.
Gupta, N., Sass, B., Chattopadhyay, S., Sminchak, J., Wang, P., and Espie, T. (2004). Geologic storage of
CO2 from refining and chemical facilities in the Midwestern US. Energy, v. 29, p. 1599-1609.
Haagsma, A., Sminchak, J. R., Moody, M., Gerst, J., Burchwell, A., & Main, J. (2015). Utilizing Cement
Bond Logs to Evaluate Wellbore Integrity for CO2 Storage. 14th Annual CCUS Conference,
Pittsburgh, Pennsylvania, 28 April – 1 May 2015.
Hamak, J. E., & Gage, B. D. (1992). Analyses of natural gases, 1991. U.S. Bureau of Mines Information
Circular IC 9318, 97 p.
Hamak, J. E., & Sigler, S. (1991). Analyses of natural gases, 1986-1990. U.S. Bureau of Mines
Information Circular IC 9301, 315 p.
Hassani, S., Vu, T. N., Rosli, N. R., Esmaeely, S. N., Choi, Y.-S., Young, D., & Nesic, S. (2014).
Wellbore integrity and corrosion of low alloy and stainless steels in high pressure CO2 geologic
storage environments: An experimental study. International Journal of Greenhouse Gas Control,
23: 30–43.
Huerta, N. J., Checkai, D., & Bryant, S. L. (2009). Utilizing Sustained Casing Pressure Analog to Provide
Parameters to Study CO2 Leakage Rates Along a Wellbore. Society of Petroleum Engineers, SPE
International Conference on CO2 Capture, Storage, and Utilization, San Diego, California, 2-4
November 2009.

Hutcheon, I., Durocher, K., Shevalier, M., Mayer, B., Block, J., & Perkins, E. (2008). Mineralogy and
CO2 Storage at the Weyburn IEA CO2-EOR Site. AAPG Search and Discovery Article #90170,
CSPG/CSEG/CWLS GeoConvention 2008, Calgary, Alberta, Canada, 12-15 May 2008.
Jenden, P. D., Drazan, D. J., & Kaplan, I. R. (1993). Mixing of thermogenic natural gases in northern
Appalachian basin. AAPG Bulletin, 77, 980-998.
Jin, L., Ravella, R., Ketchum, B., Bierman, P. R., Heaney, P., White, T., & Brantley, S. L. (2010).
Mineral weathering and elemental transport during hillslope evolution at the Susquehanna/Shale
Hills Critical Zone Observatory. Geochimica et Cosmochimica Acta, 74(13): 3669-3691.
Johnson, J. (2010). Lawrence Livermore National Laboratory (LLNL) Geochemical database.
Koukouzas, N., Kypritidou, Z., Vasilatos, C., Tsoukalas, N., Rochelle, C. A., & Purser, G. (2017).
Geochemical modeling of carbonation of hydrated oil well cement exposed to CO2-Saturated
brine solution. Applied Geochemistry, 85: 35-48.
Kutchko, B., Strazisar, B., Dzombak, D., Lowry, G., & Thaulow, N. (2007). Degradation of Well Cement
by CO2 under Geologic Sequestration Conditions. Environ Sci Technol 41(13): 4787-4792.
Laumb, J., Glazewski, K., Hamling, J., Azenkeng, A., Kalenze, N., & Watson, T. (2017). Corrosion and
Failure Assessment for CO2 EOR and Associated Storage in the Weyburn Field. Energy
Procedia, 114: 5173-5181.
Michaux, M., Nelson, E., & Vidick, B. (1989). Cement Chemistry and Additives. Oilfield Review,
1(1): 18-25.
Mills, J., Riazi, M, & Sohrabi, M. (2011). Wettability of common rock-forming minerals in a CO2-Brine
System at Reservoir Conditions. International Symposium of the Society of Core Analysts
Meeting, Texas, USA, 18-21 September 2011, 12 p.
Moody, M., & Dotson, B. (2015). Sustained Casing Pressure Diagnosis Using the Wellhead Model.
SPE/CSGM Gas Migration Challenges-Identification and Treatment Workshop, May 12-14,
2015, Banff, Canada.
MRCSP. (2004). Characterization of Geologic Sequestration Opportunities in the MRCSP Region. Report
prepared under DOE Cooperative Agreement No. DE-PS26-05NT42255. 160 p.
Orlic, B. (2009). Some Geomechanical Aspects of Geologic CO2 Sequestration. Journal of Civil
Engineering, 13: 225-232.
PCOR (Plains CO2 Reduction Partnership) (2014). Operational flexibility of CO2 transport and storage.
Riding, J., & Rochelle, C. (2005). The IEA Weyburn CO2 Monitoring and Storage project: Final report of
the European research team. British Geological Survey, Research Report RR/05/03, 54 p.
Roen, J. B., &Walker, B. J. (Eds.). (1996). The Atlas of Major Appalachian Gas Plays. West Virginia
Geological and Economic Survey, Publication V-25.
Rochelle, C., Pearce, J., Bateman, K., Birchall, D., & Turner, G. (2004). Interactions between
supercritical CO2 and borehole cements used at the Weyburn oilfield. British Geological Survey
Commissioned Report, CR/04/009. 22 p.
Sass, B., Gupta, N., Sminchak, J., and Bergman, P. (1998). Geochemical Modeling to Assess the Capacity
of a Midwestern United States Geologic Formation for CO2 Sequestration. Proceedings of the
Fourth International Conference on Greenhouse Gas Control Technologies, Interlaken,
Switzerland.

Sminchak, J. R., Moody, M., Gupta, N., & Larsen, G. (2016). Wellbore Integrity Factors for CO2 Storage
in Oil and Gas Producing Areas in the Midwest United States. Greenhouse Gases Science and
Technology, 7(5): 817-827.
Taylor, H. F. W. (1997). Cement Chemistry, 2nd ed. Thomas Telford Publishing, London. 459 p.
Washburn, E. W. (1921). The dynamics of capillary flow. Phys. Rev, 17 (3): 273-283.
Wickstorm, L., Riley, R., Spane, F., McDonald, J., Slucher, E., Baranoski, M., Zody, S., Wells, J., and
Howat, E. (2011). Geologic Assessment of the Ohio Geological Survey CO2 No. 1 Well in
Tuscarawas County and Surrounding Vicinity. Report prepared for Ohio Air Quality
Development Authority Ohio Coal Development Office and DOE-NETL. 658 p.
Wilson, M., & Monea, M. (Eds.). (2004). IEA GHG Weyburn CO2 Monitoring and Storage Project.
Summary Report 2000-2004. Regina, Canada: Petroleum Technology Research Centre.
Xu, R., & Wojtanowicz, A. K. (2001). Diagnosis of sustained casing pressure from bleed-off/buildup
testing patterns. Society of Petroleum Engineers. Paper SPE 67194 presented at SPE Production
and Operations Symposium, Oklahoma City, Oklahoma, USA. 24-27 Match.
Zaluski, W., El-Kaseeh, G., Lee, S., Piercey, M., and Duguid, A. (2016). Monitoring technology ranking
methodology for CO2-EOR sites using the Weyburn-Midale Field as a case study. International
Journal of Greenhouse Gas Control, 54 (2016): 466-478.
Zhang, M., & Bachu, S. (2011). Review of Integrity of Existing Wells in Relation to CO2 Geological
Storage: What do we know? International Journal of Greenhouse Gas Control, 5: 826-840.

Well Intergity

Uploaded by

Copyright:

Available Formats

Well Intergity

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Well Intergity

Uploaded by

Copyright:

Available Formats

Final Technical Report

Integrated Wellbore Integrity Analysis Program for CO2 Storage Applications

U.S. Department of Energy National Energy Technology Laboratory

Principal Investigator: J.R. Sminchak

September 28, 2018

This report was prepared as an account of work sponsored by an agency of

Battelle does not engage in research for advertising, sales promotion, or

Exposure to CO2 in the subsurface is a concern for wellbore integrity at

Battelle | September 28, 2018 i

Battelle | September 28, 2018 ii

Battelle | September 28, 2018 iii

Battelle | September 28, 2018 iv

Battelle | September 28, 2018 v

Battelle | September 28, 2018 vi

Battelle | September 28, 2018 vii

Battelle | September 28, 2018 viii

Battelle | September 28, 2018 ix

1.1 Project Background

Battelle | September 28, 2018 1

1.3 Research on CO2 Storage and Wellbore Integrity

Battelle | September 28, 2018 2

Table 1-2. Summary of research on wellbore integrity for CO2 storage.

Battelle | September 28, 2018 3

Battelle | September 28, 2018 4

Battelle | September 28, 2018 5

Battelle | September 28, 2018 6

Table 1-3. Summary of field sites.

Appalachian Basin Site

Battelle | September 28, 2018 7

Michigan Basin Site

Battelle | September 28, 2018 8

Figure 1-4. Location of the Williston Basin field site.

Battelle | September 28, 2018 9

2.1 Well Construction Methods

Battelle | September 28, 2018 10

2.2 Well Casing Integrity

Thermomechanical cycling. Thermomechanical cycling due to production or injection can cause

Battelle | September 28, 2018 11

2.3 Well Cement Integrity

Battelle | September 28, 2018 12

Battelle | September 28, 2018 13

Battelle | September 28, 2018 14

Battelle | September 28, 2018 15

Well Component Integrity Issue Description Causes When Leakage Pathway

casing with wellbore and during holes in casing

contamination removal before design, poor hole construction through bulk

lithology breakout and greater than to borehole wall

fractures formation stress

Battelle | September 28, 2018 16

Battelle | September 28, 2018 17

Well Cementing and Well Workover/Leakage

Figure 3-1. Summary of well record collection parameters.

3.1 Field Site Descriptions

Battelle | September 28, 2018 18

3.2 Michigan Basin Site

Battelle | September 28, 2018 19

Battelle | September 28, 2018 20

Battelle | September 28, 2018 21

3.3 Appalachian Basin Site

Battelle | September 28, 2018 22