GENERAL TECHNICAL SUPPORT DOCUMENT
FOR INJECTION AND GEOLOGIC
SEQUESTRATION OF CARBON DIOXIDE:
SUBPARTS RR AND UU
GREENHOUSE GAS
REPORTING PROGRAM
Office of Air and Radiation
U.S. Environmental Protection Agency
November 2010
-------
Disclaimer
The Environmental Protection Agency (EPA) regulations cited in this technical support
document (TSD) contain legally-binding requirements. In several chapters this TSD
offers illustrative examples for complying with the minimum requirements indicated by
the regulations. This is done to provide information that may be helpful for reporters'
implementation efforts. Such recommendations are prefaced by the words "may" or
"should" and are to be considered advisory. They are not required elements of the
regulations cited in this TSD. Therefore, this document does not substitute for the
regulations cited in this TSD, nor is it a regulation itself, so it does not impose legally-
binding requirements on EPA or the regulated community. It may not apply to a
particular situation based upon the circumstances. Mention of trade names or
commercial products does not constitute endorsement or recommendation for use.
While EPA has made every effort to ensure the accuracy of the discussion in this
document, the obligations of the regulated community are determined by statutes,
regulations or other legally binding requirements. In the event of a conflict between the
discussion in this document and any statute or regulation, this document would not be
controlling.
Note that this document only addresses issues covered by 49 CFR part 98, subpart RR
and subpart UU, which are being promulgated under EPA's authorities under the Clean
Air Act (CAA). Other statutory and regulatory requirements, such as compliance with
Safe Drinking Water Act (SDWA) and Underground Injection Control (UIC) Program
requirements1, are not within the scope of this TSD. Please see section I.D. of the
preamble to the final rule for more information on the relationship between subpart UU,
subpart RR and the UIC Program.
1 Please refer to EPA's UIC Web site for more information:
http ://water. epa. gov/tvpe/groundwater/uic/wells sequestration, cfm.
2
-------
Table of Contents
Disclaimer 2
Table of Contents 3
Acronyms and Abbreviations 6
1. Source Description 8
1.1 Overview of Source Categories 8
1.1.1 Geologic Sequestration 8
1.1.2 Enhanced Oil and Gas Recovery 8
1.1.3 Other Industries That Use C02 9
1.2 Delineation of Facility 10
2. Reporting Threshold Analysis 16
3. Monitoring Methods for C02 Received and C02 Injected 23
3.1 Determining the Amount of C02 Received Using Data from Sales, Contracts,
Invoices, or Manifests Associated with Commercial Transactions 23
3.2 Determining the Amount of C02 Received, Injected, or Produced Using Mass
Flow Meters or Volumetric Flow Meters 24
4. Overview of Monitoring Technologies for CO2 Leakage Detection and
Quantification 30
4.1 Monitoring of the Injection Well 31
4.2 Monitoring of Subsurface C02 Plume 32
4.3 Vadose Zone Testing 36
4.4 Soil Zone and Vegetation Testing 36
4.5 Atmospheric Monitoring 39
4.6 Leakage Detection and Quantification in the Offshore Environment 44
4.7 Detection Range, Accuracy and Precision of Various Monitoring Technologies .
46
5. Monitoring, Reporting, and Verification Plans 49
5.1 Delineation of the MMA and AMA 50
5.1.1 Maximum Monitoring Area (MMA) 52
5.1.2 Active Monitoring Area (AMA) 53
5.2 Assessment of the Risk of Potential Leakage of CO2 to the Surface 54
5.2.1 Assessment of the Risk of Potential Leakage of CO2 through Wells 55
5.2.2 Assessment of the Risk of Potential Leakage of CO2 through Fractures,
Faults, and Bedding Plane Partings 58
5.2.3 Assessment of the Risk of Potential Leakage of CO2 based on the
Competency, Extent, and Dip of the Confining Zone 59
5.3 Strategy for Detecting and Quantifying any CO2 Leakage to the Surface 60
5.3.1 Detecting Leakage 60
5.3.2 Verifying Leakage 63
5.3.3 Quantifying Leakage 64
5.4 Strategy for Establishing the Expected Baselines 67
5.4.1 Approaches for Establishing an Expected Baseline 68
5.4.2 Considerations in Establishing Expected Baselines 70
5.5 Site-Specific Variables for the Mass Balance Equation 73
5.6 Quality Assurance and Quality Control 76
3
-------
5.7 Missing Data 77
6. MRV Plan Approval Process Overview 78
6.1 Suggested Outline for the MRV Plan 78
6.2 Notification of Receipt of MRV Plan 78
6.3 Completeness Check 78
6.4 Technical Review 79
6.5 Issuance of EPA Decision 80
6.6 Appeals Period 80
6.7 Resubmittal of MRV Plans 80
7. Annual Monitoring Report and Records Retention 81
7.1 Narrative History of Annual Monitoring Efforts 81
7.2 Report of Non-Material Changes to the MRV Plan 81
7.3 Narrative History of Monitoring Anomalies 82
7.4 Description of Surface Leakage 82
7.5 Records Retention Requirements for MRV Plans 83
Appendix A: Glossary 84
Appendix B: U.S. Oilfields Using CO2 Injection for Enhanced Oil Recovery 87
Appendix C: Summary of Other End Uses of Captured or Produced C02 90
Appendix D: Suggested Outlines for MRV Plans, MRV Plan Resubmittals, and Annual
Monitoring Reports 93
Appendix E: Sampling Considerations for Designing a Monitoring Strategy 96
4
-------
List of Figures
Figure 1-1: Facility Example 1 11
Figure 1-2: Facility Example 2 12
Figure 1-3: Facility Example 3 13
Figure 1-4: Facility Example 4 14
Figure 1-5: Facility Example 5 14
Figure 1-6: Facility Example 6 15
Figure 2-1 Total C02 used in ER showing growth since 2004 17
Figure 5-1: Monitoring Areas as They Relate to UIC Class VI and subpart RR
Requirements 54
List of Tables
Table 2-1: Threshold Analysis Based on Amount of CO2 Received by a Facility 21
Table 4-1: Monitoring Technologies as Deployed at 10 Existing CCS Projects 30
Table 4-2: Example Leak Quantification Using Tenting and Known Test Gas Flow 38
Table 4-3: Detection Range, Accuracy, and Precision of Various Monitoring
Technologies 47
Table 5-1. Surface Components as Potential C02 Emissions Sources at Injection
Facilities 74
Table 6-1: Completeness Check Criteria for MRV Plan Submittals and Resubmittals ... 79
Table D-l: Suggested Format for Listing of Monitoring Equipment 95
5
-------
2-D
3-D
4-D
ACZ
AMA
AMR
ASTM
bbl
CCS
CFR
CH4
C02
C02n
CO2FP
CO3-
cz
DIAL
DIC
DOE
e-GGRT
EOS
EM
EOR
EPA
ER
FTIR
GHG
GPA
GPS
GS
H2C03
h2s
HCO3-
HCPV
He
InSAR
IOGCC
IRGA
IZ
Acronyms and Abbreviations
Two-Dimensional
Three-Dimensional
Four-Dimensional
Above Confining Zone
Active Monitoring Area
Annual Monitoring Report
American Society for Testing and Materials
Barrel
Carbon Dioxide Capture and Geologic Sequestration
Code of Federal Regulations
Methane
Carbon Dioxide
Total annual C02 mass emitted (metric tons) as equipment leaks or vented
emissions from equipment located on the surface between the flow meter
used to measure injection quantity and the injection wellhead.
Total annual C02 mass emitted (metric tons) as equipment leaks or vented
emissions from equipment located on the surface between the production
wellhead and the flow meter used to measure production quantity.
Carbonate
Confining Zone
Differential Absorption Light Detection and Ranging
Dissolved Inorganic Carbon
Department of Energy
Electronic Greenhouse Gas Reporting Tool
Equation of State
Electromagnetic
Enhanced Oil Recovery
Environmental Protection Agency
Enhanced Oil and Gas Recovery
Fourier Transform Infrared
Greenhouse Gas
Gas Processors Association
Global Positioning System
Geologic Sequestration
Carbonic Acid
Hydrogen Sulfide
Bicarbonate
Hydrocarbon Pore Volume
Helium
Interferometric Synthetic Aperture Radar
Interstate Oil and Gas Compact Commission
Infrared Gas Analyzer
Injection Zone
6
-------
km
Kilometer
LIDAR
Light Detection and Ranging
mcf
Thousand Cubic Feet
MIT
Mechanical Integrity Testing
MMA
Maximum Monitoring Area
MRV
Monitoring, Reporting, and Verification
NIST
National Institute of Standards and Technology
O&GJ
Oil and Gas Journal
PFC
Perfluorocarbons
ppm
Parts Per Million
ppmv
Parts Per Million by Volume
psi
Pounds Per Square Inch
SDWA
Safe Drinking Water Act
STP
Standard Temperature and Pressure
TDL
Tunable Diode Laser
TSD
Technical Support Document
UIC
Underground Injection Control
U.S.
United States
USDW
Underground Source of Drinking Water
ZERT
Zero Emissions Research and Technology
-------
1. Source Description
Preliminary estimates indicate that the amount of carbon dioxide (CO2) captured from
industrial processes, including combustion and chemical manufacturing, and produced
from naturally occurring subsurface CO2 reservoirs is approximately 44 million metric
tons carbon dioxide equivalent ,2'3 Currently more than 95 percent of this CO2 supplied
to the economy is injected underground for enhanced oil and gas recovery (ER).2 CO2
may be injected underground for geologic sequestration (GS). GS is the long-term
containment of a CO2 stream in subsurface geologic formations and is a key component
of a set of climate change mitigation technologies known as carbon dioxide capture and
geologic sequestration (CCS). CCS has the potential to enable large emitters of CO2 such
as coal fired power plants to significantly reduce greenhouse gas (GHG) emissions.
1.1 Overview of Source Categories
Three sectors were considered for inclusion in this rule: injection of CO2 underground for
GS, injection of CO2 underground for ER, and end uses of CO2 by other industries.
1.1.1 Geologic Sequestration
Underground geologic formations that can be used for GS include deep saline formations,
oil and natural gas reservoirs, and unmineable coal seams. In addition, CO2 may be
injected into other types of subsurface geologic formations, such as basalt formations.
The UIC program, which is authorized by Part C of the SDWA, regulates underground
C02 injection.
Geologic sequestration occurs through a combination of structural and stratigraphic
trapping, residual CO2 trapping, solubility trapping, mineral trapping, and preferential
adsorption trapping. These mechanisms are functions of the physical and chemical
properties of CO2 and the geologic formations into which the CO2 is injected. For more
background information on GS trapping mechanisms, see the Vulnerability Evaluation
Framework for Geologic Sequestration of Carbon Dioxide.4
1.1.2 Enhanced Oil and Gas Recovery
CO2 is currently being injected into subsurface geologic formations in the United States
(U.S.) for ER. The C02 currently being used in ER is primarily produced from naturally
occurring underground CO2 reservoirs but is also captured from industrial processes,
including combustion and chemical manufacturing.
2 U.S. EPA. 2010. Inventory of U.S. Greenhouse Gas Emissions and Sinks, 1990-2008, EPA 430-R-10-
006. Available at: http://epa.gov/climatechange/emissions/usinventorvreport.html.
3 The estimated 44.2 million metric tons of C02 does not include biogenic C02.
4 Vulnerability Evaluation Framework for Geologic Sequestration of Carbon Dioxide (see docket ID No.
EPA-HQ-OAR-2009-0926)
8
-------
ER involves injecting CO2 into oil or natural gas reservoirs via injection wells for the
purposes of increasing crude oil production or to enhance recovery of natural gas. The
crude oil and CO2 mixture is produced from production wells and sent to a two-phase
separator where the crude oil is separated from the gaseous hydrocarbons and C02. The
gaseous C02-rich stream then is typically dehydrated, recompressed, and reinjected into
the oil or natural gas reservoir to further enhance recovery. If the concentration of
hydrocarbons in the CO2 stream from the dehydrator is significant, then an acid gas
recovery unit is used to separate the hydrocarbons from the C02.
Injection of C02 into unmineable coal seams may result in displacement of methane
(CH4) from the coal seam and subsequent production of the methane as a product (or
enhanced coal bed methane).
There are currently 80 ER fields operating in the United States where C02 is being
injected for the purposes of ER.5 ER projects operating in the United States range from
new pilot-scale projects with one or two injection wells to C02 floods that commenced
operation in the 1970s and that have hundreds of injection wells. Approximately 44
million metric tons of C02 was received for injection underground for ER in 2008. Of
this amount approximately 80 percent was produced from naturally occurring
underground C02 reservoirs and 20 percent was captured from industrial processes,
including combustion and chemical manufacturing.6
Natural gas processing plants and wellhead treatment units condition incoming natural
gas from the wellhead to meet sales and natural gas pipeline specifications. In some fields
the natural gas may contain a significant quantity of hydrogen sulfide (H2S) or C02,
which is separated from the natural gas by the processing plants and treatment units.
Because of the highly corrosive nature of this stream, the combination of H2S and C02
separated from the natural gas is called acid gas. The composition is quite variable and
can range from 2 percent H2S and 98 percent C02 to about 85 percent H2S and 15 percent
C02.7 Most acid gas is disposed of by underground injection under a UIC Class II permit.
These permits may allow for disposal of other oil and gas production wastes, including
brine, well completion and work-over fluids, and spent dehydration unit fluids, in
addition to the acid gas.
1.1.3 Other Industries That Use CO2
5 See Appendix B of this technical support document.
6 U.S. EPA. 2010. Inventory of U.S. Greenhouse Gas Emissions and Sinks, 1990-2008, EPA 430-R-10-
006. Available at: http://epa.gov/climatechange/emissions/usinventoryreport.html.
7 S. Bachu and W.D. Gunter. 2004. Overview of Acid Gas Operations in Western Canada. In: Proceedings
of 7th International Conference on Greenhouse Gas Control Technologies. Volume 1: Peer-Reviewed
Papers and Plenary Presentations.
9
-------
EPA identified and considered a total of 22 commercial end use sectors8 that use gaseous,
liquid, or solid C02, excluding ER and GS. These end use sectors are summarized in
Appendix C. EPA received comments to subpart PP suggesting that at least one of these
end-uses - precipitated calcium carbonate production - may be a non-emissive use. At
proposal for this rule, EPA sought comment on whether applications, such as precipitated
calcium carbonate and some cement production, permanently sequester C02 and if so,
which industries this would include; how many facilities operate in each of these
industries; how much of the C02 consumed in each industry would be sequestered;
whether a sequestration factor would be reasonable in any case; and what methodologies
could be used to verify this sequestration. These sectors were not included in this final
rule. Please refer to the Response to Comments document for further information.9
1.2 Delineation of Facility
For C02 received by pipeline, the point of measurement for reporting the amount of CO2
received at a facility will often be identified by a change in ownership (or custody) of the
CO2. The transfer of custody of the CO2 is the basis of sales contracts and revenue
reporting, and is generally measured with meters that conform to state oil and gas board
regulations and industry standards. The question arises of how to delineate a facility
when collecting data on C02 received. Subparts RR and UU rely on the definition of
"facility" in 40 Code of Federal Regulations (CFR) 98.6, which states that "a facility
means any physical property, plant, building, structure, source, or stationary equipment
located on one or more contiguous or adjacent properties in actual physical contact or
separated solely by a public roadway or other public right-of-way and under common
ownership or common control..." To illustrate how EPA applies this definition of facility
to injection wells under subparts RR and UU, six example scenarios are presented in
Figure 1-1 through Figure 1-6.
8 This count includes the various chemical, pharmaceutical, and other processes that use C02 as an end
product and excludes intermediate C02 processors. After C02 is captured from production process units or
produced from natural reservoirs for subsequent commercial application, it is generally purified,
compressed, and liquefied before being delivered for a multitude of end use commercial applications. This
processing is done so that end users can input the C02 into their processes in the necessary state; e.g., food
grade C02 used for food and beverage production. Intermediate processors often receive the produced C02
from the production process units or natural reservoirs, purify, compress, and/or liquefy it, and then deliver
it to end users. Intermediate processors are not discussed further in this TSD.
9 See docket ID No. EPA-HQ-OAR-2009-0926.
10
-------
KEY
| M 1 Custom Transfer Meier
©
Oil Production Well
(gathering I roes not shown)
CQj Injection Well
o>
Facility outline
.X.
C02 Trunk line
C02 Distribution
Lines (s«ze denotes
relative dlametw)
Figure 1-1: Facility Example 1
In the first example (Figure 1-1) an operator is conducting CO2-ER operations in a field
that is fully contained within one leased area. A large-diameter CO2 supply pipeline
(trunk line) owned by a third party supplies all the CO2 used in the field. The CO2 is
delivered to the field from the trunk line via a distribution line owned by the field
operator. The custody of the CO2 is transferred to the field operator at the meter where
the distribution line begins. Once the field operator takes possession of the CO2, it is
transported over some distance and further distributed and injected into the wells in the
field. Although not shown on the figure, there would likely be gathering lines from the
producing oil wells that lead to facilities that capture and re-process the CO2 for re-
injection downstream of the custody transfer meter. In this scenario the field is one
facility because it is under common control, and the meter reading at the custody transfer
point is the amount of CO2 received.
Field A
Trunk Line
Company XYZ Lease
11
-------
KEY
| M 1 Custom Transfer Meier
Oil Production Well
(gathering Imes not shown)
o>
Oi Field
C02 injection well
Facility outline
Company XYZ Leases
C02 Trunk line
C02 Distribution
Lines (s«ze denotes
relative diameter)
Trunk Line
Figure 1-2: Facility Example 2
In the second example, Figure 1-2, two fields under common control are fed by a
continuous CO2 distribution line that passes through Field A and terminates in Field B.
The operator has placed a meter at the upstream end of Field B to assist with state oil and
gas board reporting requirements and reservoir management. In this case, whether this
scenario is one or two facilities depends on site specific conditions. Reporters should
review the definition of facility carefully and decide whether to report as one or two
facilities in their specific situation.
12
-------
KEY
l~M"l Custody Transfer Meter
Oil ProOuction well
(gathe ring Imes not shown)
C02 Injection Well
\
C02 Trunk line
CO? Distribution
Lines (size denotes
reiative diameter)
Trunk Line
Figure 1-3: Facility Example 3
Company XYZ Lease
Company ABC Lease
An example similar to Figure 1-1 is shown in Figure 1-3. In this instance, Fields A and B
are on different leases and operated by different entities, but fed by a single CO2
distribution line from the main trunk line. Ownership of the C02 is transferred at the
meter between Fields A and B. In this example because Field A and Field B are under
different ownership and control, they represent two different facilities. The amount of
CO2 passing through meter A (at the trunk line) represents the total amount of CO2
supplied to both fields; therefore the owner of Field A should report the total amount at
meter A less the amount at meter B (the amount of CO2 provided to Field B).
13
-------
KEY
Trunk Line
Oil ProOuction well
(gathe ring lines not shown)
C02 Injection Well
CO? Distribution
Lines (size denotes
reiative diameter)
l~M"l Custody Transfer Meter
Company XYZ
Leases
Figure 1-4: Facility Example 4
Figure 1-4 shows a trunk line that splits and supplies two different fields operated by the
same entity, and thus are under common control. Custody of the C02 is transferred at the
meters that connect the distribution line to the trunk line. This example is similar to the
second example, in that the determination of whether this situation represents one or two
facilities depends on site-specific conditions. Reporters should review the definition of
facility carefully and decide whether to report as one or two facilities in their specific
situation.
KEY
©
rwi Custody Transfer Meter
Oil Production Well
(gathering lines not shown)
\
C02 Trunk line
CQ2 Distribution
Lines (see denotes
relative diameter)
Company XYZ Lease
Figure 1-5: Facility Example 5
14
-------
In large ER operations in a single field operated by the same entity, the trunk line may
pass directly through the field and connect to multiple distribution lines (Figure 1-5).
Custody of the C02 is transferred to the operator at each meter. Even though there are
many custody meters, the entire field is under common ownership and would be
considered a single facility. To determine the total amount of C02 received at the facility,
the reporter would sum the CO2 amounts from all meters within the field.
KEY
l~M~l Custody Transfer Meter
©
Oil Production Wfcli
[gathering linen not shown)
C02 Injection Well
O
\
C02 Trunk line
CO? Distribution
Lines (atze denotes
relative dlMTMttf)
XYZ Power Plant
Company XYZ Lease
Figure 1-6: Facility Example 6
In the last example, Figure 1-6, a CO2 generator supplies CCMo an ER project that it
operates. The generation facility and ER operation are located on the same property and
are connected by a dedicated transmission line. In this example the generating facility,
the transmission line, and the ER operation are under common ownership and are located
on the same property and therefore are considered a single facility under subpart RR. The
CO2 received is the amount passing through the meter at the point of generation in this
example. A third-party pipeline company that owns and operates the transmission line,
but does not take custody of the gas or operate the injection facility, is not a facility or
operator under subpart RR.
15
-------
2. Reporting Threshold Analysis
To determine the appropriate threshold for reporting, EPA considered a threshold based
on the amount of CO2 emitted, a threshold based on the amount of CO2 injected, and a
threshold based on the amount of CO2 received by the facility. EPA concluded that an
emissions-based threshold would be problematic because of the lack of data on the
incidence and scale of surface emissions and leakage from facilities that conduct GS and
all other facilities that inject CO2 underground. EPA also concluded that the amount of
CO2 injected one year is not a good indicator of the amount of CO2 injected the following
year, and that injected CO2 may be produced and recycled at some facilities. EPA
accordingly analyzed injection facilities based on the quantity of CO2 received by the
facility.
EPA conducted the threshold analysis based on the quantity of CO2 received at the
facility (not including CO2 being recycled onsite) and considered whether a threshold on
CO2 received should apply. EPA evaluated a no threshold option (i.e., all facilities that
inject CO2 would be required to report), 1,000 metric tons per year, 10,000 metric tons
per year, 25,000 metric tons per year, and 100,000 metric tons per year of CO2 received
per facility. Only facilities expected to be operating in 2011, the first year of reporting
under this proposal, were considered in the analysis. For this analysis it was assumed that
all of the CO2 received by the facility was injected for ER.
To establish a facility count, EPA relied on data reported in the Oil and Gas Journal
(O&GJ) Enhanced Oil Recovery Survey published in April 2008.10 These data come from
the results of a voluntary survey of oil producers reporting miscible and immiscible CO2
projects in the U.S. operating as of the end of 2007. The O&GJ survey asked for the
production figures "as of the end of 2007." For projects that were active throughout 2007
the value represents annual production (12 months). Several of the projects started mid-
year, or had scheduled downtime in 2007, so the reported annual 2007 production may
represent only a partial year of production. Production data were not pro-rated to come up
with an annual estimate. Some projects starting late in the year did not see any increase in
crude oil production due to ER operations, or were unable to measure it. These were
noted as "TETT" (Too Early To Tell) in the O&GJ survey. In a few cases, companies did
not report crude oil production data to the O&GJ. The O&GJ did not specify how the
production data were to be calculated by the reporting facilities, so most of the data are
based on internal oil company reporting, which is consistent with state oil and gas
commission reporting requirements.
The O&GJ left it up to the companies to define a "project" when they reported data. Most
projects are defined based on production reporting requirements to state oil and gas
commissions, based on the unit or pool. Some state oil and gas commissions finely divide
the production zones in order to manage resource conservation. In one operation in
Michigan, for example, the fields are very small (2-5 wells each) because they target
pinnacle reefs that are limited in their extent. The company reported two "projects"
separately to the O&GJ even though they are located in the same field, because it appears
10 Enhanced Oil Recovery Survey. 2008. Oil and Gas Journal, Volume 105, Issue 15.
16
-------
that they target different producing zones within the same formation. A somewhat
different situation exists for the Seminole field in West Texas, where several projects are
listed. These projects are in different geographic locations within the large Seminole
field, and target different parts of the reservoir (i.e., main oil pay or residual oil) using
different CO2 injection technologies. These are identified as "pilot tests" in the O&GJ
data, so the company considers them separate projects for evaluation purposes.
Depending on how "project" or "facility" is defined, they could be separate or combined.
The use of CO2 in ER has increased significantly since 2004 (Figure 2-1), and is expected
to increase further in the next five to 10 years. The O&GJ survey is the latest and most
complete data set of active ER projects in the U.S. EPA believes that the 2007 data
presented in the O&GJ survey are representative of the type and scope of ER projects that
will be operating in 2011.
For this analysis, EPA compiled all the projects listed for miscible and immiscible CO2
floods11 reported in the O&GJ survey. The list of active projects represents a wide cross-
section of several geologic basins in 10 states, ranging from relatively recent pilot-scale
projects with one or two injection wells to C02 floods started in the 1970s with hundreds
of injection wells. The Wasson Field in the Permian Basin is the largest user of CO2, and
has six active C02 ER projects managed by two operators.
|
i
~Jackson Dome
¦ Sheep Mtn.
~McElmoDome
¦LaBarge
~ Doe Canyon
~ Bravo Dome
~ Dakota Gasification
~Val Verde Basin
Figure 2-1: Total C02 used in ER showing growth since 2004. Graph based on data from
specified anthropogenic and natural C02 sources.
Data based on sales. Source: Kinder Morgan C02 Company, CCS briefing to EPA June 3, 200912
11 A miscible C02 flood injects C02 as a liquid at high pressure to completely mix with oil and make it
flow more easily. An immiscible C02 flood uses lower pressures of C02 to swell the oil and provide
additional gas pressure to move the oil.
12 Note that the C02 being produced by Dakota Gasification Plant in North Dakota is being injected into an
ER project in Canada.
17
-------
A total of 105 active ER projects operated by 23 different companies were reported in the
O&GJ survey. In some cases multiple projects were reported by the same company in an
oil field, reservoir, or operator unit. For the purposes of this analysis, EPA grouped these
reported projects by field and by owner or operator to align with the definition of facility
at 40 CFR 98.6. This computation results in 80 facilities that conduct ER. EPA further
grouped the fields by their source of C02 as determined from geographic location, U.S.
DOE pipeline maps, and operator-published information. As of 2008, four natural sources
(Jackson Dome, MS; Bravo Dome, NM; Sheep Mountain, CO; and McElmo Dome, CO)
and five anthropogenic CO2 sources (Antrim Gas Plant, MI; LaBarge/Shute Creek Gas
Plant, WY; Enid Fertilizer Plant, OK; U.S. Energy Partners Russell, KS Plant; and Val
Verde Gas Plants, TX) supply CO2 for ER that is conducted in the United States.13
The O&GJ survey does not provide the specific volume of CO2 used in each of the active
ER projects, but does provide the total oil production attributed to the ER process. To
calculate the estimated volume of CO2 received by each ER project, EPA determined the
total amount of C02 used daily for ER based on data from the U.S. EPA 1990-2007
Inventory of U.S. Greenhouse Gas Emissions and Sinks14 (Inventory). According to the
Inventory, approximately 2.08 billion cubic feet per day of C02 from natural and
anthropogenic sources is received (purchased) for ER. The threshold analysis was
performed two ways, one based on total daily C02 received combined for all fields in the
U.S., and one based on the total from each source of CO2 received for ER (natural or
anthropogenic). The daily average C02 production from each source was apportioned
among the projects supplied by the source based on an average value for the fractional
production of oil attributed to ER as presented in the O&GJ survey and normalized on an
annual basis (see Appendix B). The analysis was further complicated because
anthropogenic gas from the Val Verde Gas Plants is mixed with natural C02 in the
pipeline carrying the CO2 to the Permian Basin; therefore the specific projects supplied
by the Val Verde Gas Plants could not be determined. To accommodate the mixture of
sources in the Permian Basin, the CO2 from the Val Verde Gas Plants was apportioned
based on the total crude oil production in the Permian Basin and the known quantities of
CO2 supplied from natural sources. The CO2 from the Val Verde Gas Plants represents
approximately 5.39 percent of the CO2 used in Permian Basin Enhanced Recovery
projects.
EPA recognizes that this is likely an oversimplification of the actual volume of CO2
received by each facility, but notes that it follows the principle that higher production is a
function of higher CO2 injection volumes. The volume of CO2 received by a particular
ER project is a function of many factors, including:
13 DOE/NETL, 2009. Carbon Dioxide Enhanced Oil Recovery, Untapped Domestic Energy Supply and
Long Term Carbon Storage Solution. U.S. Department of Energy, National Energy Technology Laboratory,
September 2009. page 11. Available at:
http://www.netl.doe.gov/technologies/oil-gas/publications/EP/small CQ2 eor primer.pdf
14 U.S. EPA. 2009. Inventory of U.S. Greenhouse Gas Emissions and Sinks, 1990-2007, EPA 430-R-09-
004. Available at: http://epa.gov/climatechange/emissions/downloads09/GHG2007entire report-508.pdf.
18
-------
• Reservoir characteristics: Heterogeneity is a significant design consideration,
along with porosity, permeability, oil gravity, production history, depth, and
reservoir pressure.
¦ Flood design: The injection design (e.g., continuous, simultaneous water and
gas, water alternating gas), and number of injector wells are major factors in
overall C02 use. Injection well pattern, miscible or immiscible processes, C02
saturation target as a percent of hydrocarbon pore volume (HCPV), and use of
surfactants and additives also influence C02 use.
¦ Project age: The stage of the project is a significant factor in determining C02
use. New C02 floods use more purchased C02 and produce less oil because
the C02 has not fully penetrated the reservoir. It may take 6-12 months to see
an increase in oil production upon initiation of injection. As the reservoir
becomes saturated with C02, the amount of new C02 added to the project is
reduced, and the majority of the injected C02 is recycled from producing
wells.
The calculation of C02 utilization in the threshold analysis presented in this chapter is
based on the volume of new (purchased) C02 for ER obtained from the Inventory. The
value represents the "Net C02 Utilization" in thousand cubic feet (mcf)/barrel (bbl).
"Gross C02 Utilization" is the total amount injected, including purchased and recycled
C02. Gross C02 Utilization was not analyzed in this threshold analysis because the total
injection data were not available.
To evaluate the reasonableness of the approach, EPA compared the calculated Net C02
Utilization from the threshold analysis to field-specific values for 12 field-scale projects
published by DOE.15 Values for individual projects showed variance; however, overall
the average values were consistent (7.53 mcf/bbl in the Threshold Analysis, versus 6.48
mcf/bbl in the DOE report).
As a secondary check on the reasonableness of the estimates in the threshold analysis,
EPA compared the average age of the projects to the Net C02 Utilization. EPA
performed this comparison because the age of the project is one of the most significant
factors in determining Net C02 Utilization. The threshold analysis evaluated ER projects
by C02 source (essentially by producing basin) and resulted in Net C02 Utilization
estimates by basin. The results provided support that the estimates of C02 use were
reasonable. For example:
• The Permian Basin has the oldest EOR projects (70 percent started before
2000) and the Threshold Analysis shows the lowest Net C02 Utilization (7.2
mcf/bbl) for this basin.
15 DOE. 2006. Evaluating the Potential for "Game Changer" Improvements in Oil Recovery Efficiency
from C02-Enhanced Oil Recovery, Prepared by Advanced Resources International for DOE NETL,
February 2006. Available at:
http://fossil.energy.gov/programs/oilgas/eor/Game Changer Oil Recovery Efficiencv.html.
19
-------
• The Mississippi Interior Salt Basin has the youngest CO2 EOR projects (15
percent started before 2000) and the Threshold Analysis shows the highest
Net CO2 Utilization (19.2 mcf/bbl) of any basin.
Based on the analysis described above, the simplified estimation using production-
apportioned C02 volumes shows a good correlation with independent data and is an
appropriate estimation of the amount of CO2 received by the facility.
Analysis of the ER injection data showed five fields with no enhanced production, which
were correspondingly allocated to no C02 use. One additional field showed near zero
production. Two of the six fields reported the projects as having "just started," and four
reported projects at the "nearing completion" stage. This demonstrated to EPA that
injection activity generally follows a curve that can slowly increase in the beginning, as
the project moves from a pilot test to a larger scale, or slowly decrease toward the end of
operations. Development of a full-scale CO2 ER project requires extensive testing and
ramping up over many years. Additionally, EPA determined that some ER production
operations inject CO2 and water in alternating pulses rather than continuously over years
to maximize production, and the C02 injection periods and the corresponding
(alternating) water injection periods may last from several months to two years or more.
Therefore, the annual snapshot of data compiled for ER does not necessarily represent the
typical operating conditions of full-scale projects expected in the future, as the number of
ER operations increase.
The Interstate Oil and Gas Compact Commission (IOGCC) noted that acid gas is injected
at approximately 20 sites in Michigan, New Mexico, North Dakota, Oklahoma, Texas,
and Wyoming.16 Many of the acid gas injection well locations are within oil fields that
are listed in the O&GJ EOR survey, and inject into oil producing zones, suggesting the
acid gas is being used for ER and not being injected solely for the purposes of disposal.
Other acid gas injection locations identified by the IOGCC that are not associated with
ER include one in Michigan, two in North Dakota, two in the Palo Duro basin of Texas,
one in the Permian Basin in New Mexico, and one in Utah. These sites generally inject
the acid gas for disposal into permeable strata below the deepest currently producing oil
zone. EPA could not find any data to estimate quantity of CO2 received for injection, so
these facilities are not included in the threshold analysis.
Table 2-1 shows that nearly all facilities (92.5 percent) received greater than 1,000 metric
tons of CO2 per year. At a reporting threshold of up to 25,000 metric tons of CO2
received per year, essentially all CO2 (100 percent) received would be included, but not
all facilities would report. At a reporting threshold of 25,000 tons of CO2 received per
year, 65 of the 80 facilities (81 percent) would report. At a 100,000 metric tons of CO2
16 Interstate Oil and Gas Compact Commission, 2005, Interstate Oil and Gas Compact Commission. 2005.
Carbon Capture and Storage: A Regulatory Framework for States - Final Report January 2005. Available
at:
http://groundwork.iogcc.org/topics-index/carbon-seauestration/executive-white-papers/ccgs-task-force-
phase-i-final-report-2005
20
-------
received per year threshold, 97.4 percent of all CO2 used in ER would be reported from
60 percent of the facilities. The majority of the 40 percent of facilities not covered are
characterized in the O&GJ Survey as pilot projects, projects that have just started, or
projects nearing completion.
Table 2-1:
hreshold Analysis Based on
Amount of CQ2 Received by a Facility
Threshold
Level
(metric
tons/yr of
co2
received)
Total
National
(metric
tons/yr of
co2
received)
Total
Number of
U.S.
Facilities
Amount of C02 Received
Number of Facilities
Metric tons/yr
of C02
Received
Percent
Covered
Number of
Facilities
Covered
Percent
Covered
All In
40,111,639
80
40,111,639
100.0%
80
100.0%
1,000
40,111,639
80
40,111,115
100.0%
74
92.5%
10,000
40,111,639
80
40,099,065
100.0%
71
88.8%
25,000
40,111,639
80
40,005,238
100.0%
65
81.3%
100,000
40,111,639
80
39,065,039
97.4%
48
60.0%
Since the proposed subpart RR rule and March 2010 General TSD17 were published, a
new O&GJ survey of ER projects was published. The 2010 O&GJ Worldwide Survey of
Enhanced Oil Recovery projects was conducted by O&GJ and published on April 19,
2010 (Volume 108 Issue 14). The data were collected and presented in the same manner
as the 2008 survey that EPA used in the Threshold Analysis and described in the March
2010 General TSD.
The 2010 survey reflects the status of CO2 ER projects at the end of 2009, and shows that
the popularity of C02 ER is growing. There was a 14 percent increase in the number of
fields employing CO2 flooding, a 4 percent increase in the acreage of projects, and a 2
percent increase in the acreage of the fields. The projects included in the 2010 O&GJ
survey are generally being operated by the same companies and in the same states as was
the case in the 2008 survey. The projects appear to be more successful, producing 14
percent more oil per project with only a 3-5 percent increase in the number of producing
and injection wells. This is likely due to the increased volumes of CO2 being used for
flooding.18
In the threshold analysis described in the March 2010 General TSD, EPA estimated the
CO2 use for each field identified in the 2008 OG&J survey of ER projects operating as of
the end of 2007 using the apportioned amount of CO2 used daily for ER based on the
2007 data from the 1990-2007 U.S. GHG Inventory. The 2009 data from the 1990-2009
U.S. GHG Inventory is not available at this time; therefore the Threshold Analysis cannot
be updated to reflect the new 2010 O&GJ Survey data which identifies ER projects as of
the end of 2009. However, producers of natural sources of CO2 are increasing
17 U.S. Environmental Protection Agency, General Technical Support Document for Injection and Geologic
Sequestration of Carbon Dioxide, Proposed Rule for Mandatory Reporting of Greenhouse Gases, March
2010 (see docket ID No. EPA-HQ-OAR-2009-0926).
18 DOE. 2009. Electricity Use of Enhanced Recovery with Carbon Dioxide (C02 -EOR), U.S. Department
of Energy, DOE/NETL 2009-1354, January 26, 2009.
21
-------
production19,20,21 in response to the growing demand for CO2 in new and existing ER
projects. The 2008 Inventory shows an increase from 2007 of approximately 3 million
metric tons of CO2 production from natural sources. Additionally, as noted above, the
volume of C02 used in each field is increasing, as is the Net C02 Utilization.
19 Denbury Resources 2009 Annual Report page 20 shows a 7.2 percent increase in production of C02 over
the previous year, and page 48 reports a production increase of 17 percent over 2009 in February 2010.
20 Kinder-Morgan Analysts Conference, January 28, 2010, Slide 14 showing Permian Basin C02 Deliveries
increasing 23 percent from 2007 to 2009.
21 The 1990-2008 U.S. GHG Inventory (containing 2008 data) is available at:
http://epa.gov/climatechange/emissions/usinventorvreport.html.
22
-------
3. Monitoring Methods for C02 Received and C02 Injected
This chapter presents monitoring methods for determining the amount of CO2 received
(subparts RR and UU) and the amount of C02 injected (subpart RR).
CO2 is a gas whose volume is typically reported at standard temperature and pressure
(STP). The behavior of CO2 at low pressure and high temperatures can be satisfactorily
approximated by the ideal gas law. The ideal gas law is simple to apply in practice, but
the behavior of CO2 deviates from ideal gas with an increase in pressure or decrease in
temperature as the assumptions made in deriving the ideal gas equation begin to deviate
from actual gas behavior. The GHG Reporting Program addresses CO2 equipment leaks,
vented CO2 emissions, and combustion CO2 emissions as well as CO2 streams, which can
be at a wide range of temperatures and pressures. In addition, the program also covers
facilities that conduct GS and all other facilities that inject CO2 underground, which may
use mixtures of several fluids that predominantly contain CO2. The ideal gas law is only
applicable to fluids in the gas phase; therefore significant error will occur if it is used to
calculate densities for liquids and supercritical fluids.
Measuring CO2 in gas phase should be performed differently from CO2 in the
supercritical phase. In supercritical phase, C02 behaves much like a liquid, and other
gases can dissolve into supercritical-phase CO2. This means the volumetric reading of
supercritical phase C02 stream can be misleading if there is any significant amount of
impurities dissolved in the CO2 stream. The discussion below describes different
measurement and calculation methods that may be required to accurately measure the
different phases of CO2.
3.1 Determining the Amount of C02 Received Using Data from Sales, Contracts,
Invoices, or Manifests Associated with Commercial Transactions
In order to determine the amount of C02 received, a facility would measure the flow rate
of the CO2 at the custody transfer meter at the facility boundary prior to any subsequent
processing operations at the facility, use flow rate data from the sales contract associated
with CO2 received from a one-time commercial transaction, use data from invoices or
manifests for C02 received from an ongoing commercial transaction, or measure the flow
rate at the equivalent of a custody transfer meter, following the provisions of subpart PP,
for C02 supplied from a production process that is part of a facility. Also, for commercial
transactions for which the sales contract specifies a range of CO2 concentration, and if the
supplier of the C02 sampled the C02 stream and measured its concentration per the sales
contract terms, the reporter can use the average CO2 concentration data from the sales
contract or the seller's measured value.
Information concerning the use of mass or volumetric flow meters for measurement of
the amount of CO2 received is included below under Chapter 3.2 concerning the use of
data from sales contracts, invoices, or manifests associated with commercial transactions
is discussed in this chapter.
23
-------
For a one-time commercial transaction between a C02 supplier and a C02 recipient (i.e.,
the reporter) there would generally be a sales receipt indicating the amount of CO2 that
was supplied by the supplier and that was received by the recipient in the commercial
transaction.
For an ongoing commercial transaction between a CO2 supplier and a CO2 recipient (i.e.,
the reporter) involving a series of discrete shipments, there would generally be invoices
or manifests indicating the amount of CO2 that was supplied by the supplier and that was
received by the recipient. Each individual invoice or manifest would cover a specific
period of time (e.g., one month; one quarter) over which CO2 was received, or may cover
one individual transaction in the series of discrete shipments.
To minimize errors in reporting data from sales receipts or invoices and manifests
associated with one-time and ongoing commercial transactions, the sales receipts,
invoices, and manifests provided by the C02 supplier should report the amount of C02
supplied (and the amount of CO2 received) in the same physical units (mass units) as are
required to be reported under the subpart, and using the same standard conditions (STP)
and conversion factors as are required to be used under the subpart. This will minimize
errors by the reporter in converting the amount of C02 received from one set of units
(e.g., volumetric units) to the units (i.e., mass units) that are required to be used under the
subpart. If the commercial transaction is actually tied to a different set of measurement
units (e.g., volumetric units) the sales receipt, invoice, or manifest should report the
amount of C02 supplied in both the units to which the transaction is tied and the units
that are required to be reported under the subpart. The units of CO2 concentration
measurements provided by the C02 supplier in sales contracts used by the reporter to
determine the CO2 content should also be in the same units as required to be used under
the subpart, and the methodology used by the CO2 supplier to measure the CO2
concentration should be consistent with the methodology specified under subpart PP for
CO2 suppliers.
For ongoing transactions, the reporter should compare invoices and manifests received
for CO2 supply transactions from one period to the next (e.g., monthly, quarterly) or
should compare the invoices or manifests from one shipment to the next, to assess the
variability of the amount of CO2 supplied. Unanticipated variation in the amount of CO2
supplied (and received) between one invoice/manifest period and the next could indicate
a reporting error on the part of the CO2 supplier, for example.
3.2 Determining the Amount of CO2 Received, Injected, or Produced Using Mass
Flow Meters or Volumetric Flow Meters
In order to determine the amount of CO2 received, injected, or produced (at ER or other
fluid production operations), a facility would measure the flow rate using either a mass
flow meter or a volumetric flow meter. To determine the mass, the facility would either
multiply the mass flow rate by the concentration of CO2 in that flow or multiply the
volumetric flow rate at STP by the concentration of CO2 in the flow and by the density of
24
-------
CO2 at STP. Either a continuous mass flow meter or a continuous volumetric flow meter
would most accurately account for the fluctuations over time of flow rate.
To minimize the error in calculating the mass from CO2 volumetric readings, the
application of an equation of state22 (EOS) more accurate than the ideal gas law is
recommended. The EOS proposed by Span and Wagner23 represents the state of science
in theoretical prediction of gas phase CO2 properties. This EOS predicts density of CO2
within ± 0.03 percent to ± 0.05 percent for pressures up to 4,350 pounds per square inch
(psi) and temperatures up to 482°F. At the same time, the Span and Wagner EOS may be
too complex to be directly applied by reporters. Therefore, EPA recommends using the
database of thermodynamic properties developed by the National Institute of Standards
and Technology (NIST). This online database, available at
http://webb00k.nist.gov/chemistry/fluidA provides density of CO2 using the Span and
Wagner EOS at a wide range of temperature and pressures.
The following generic protocol is recommended to calculate CO2 mass from volumetric
measurements, with the assumption that the gas stream is predominantly CO2:
• Obtain volumetric measurements using consensus standards at operating
conditions;
• Determine the density of CO2 at operating conditions using the Span and
Wagner EOS from NIST tables;
• Calculate the mass of CO2 in the required units by multiplying the density by
the volumetric measurements of CO2.
• Obtain species and their mass fraction information by sampling the stream of
CO2. The sample can be analyzed using standards from consensus standards
organizations for methods such as chromatography or mass spectroscopy to
identify all chemical constituents and their mass fractions. With knowledge of
the mass of CO2 calculated, the masses of the other individual species present
may be evaluated knowing their individual mass fractions in the CO2 stream.
The density of liquid or supercritical phase24 CO2 depends significantly on process
conditions as well as mixture composition. Therefore, available theoretical models to
calculate density of CO2 may not be robust enough to span all potential process
conditions. Furthermore, theoretical methods, where available, may be too cumbersome
for reporters to implement in practice. EPA has not identified standards from consensus-
based organizations that are suited for the purpose of measuring the density of CO2 in
liquid or supercritical state. Therefore, EPA recommends the following generic protocol
22 An equation of state is a mathematical expression that describes the relationship between thermodynamic
properties of chemical species.
23 Span, R. and W. Wagner. 1996. A New Equation of State for Carbon Dioxide covering the Fluid Region
from the Tripe-point Temperature to 1100 K at Pressures up to 800 MPa, J. Phys. Chem. Ref. Data, Vol 25.
24 Supercritical C02 refers to C02 with both pressure and temperature greater than their critical values.
Critical temperature of a gas is the temperature above which the gas cannot be liquefied at any pressure.
This temperature for C02 is 88°F. Critical pressure is the pressure required to liquefy the gas at its critical
temperature. Critical pressure for C02 is 1072 psi.
25
-------
to calculate density of CO2 in liquid or supercritical state at process conditions or
conditions at the point of transfer/measurement:
• Collect a representative sample of liquid or supercritical C02 mixture from
process equipment in a high-pressure sampling container.
• Determine the volume of the sampling container from manufacturer
specifications.
• Determine the mass of the sample by weighing the sampling container before
and after taking the sample.
• Send the sampling container to an analytical laboratory for compositional
analysis. The laboratory may expand the contents in sampling container to
testing conditions. A sub-sample is analyzed using standards from consensus
standards organizations for methods such as chromatography and mass
spectroscopy to identify all chemical constituents and their mass fractions.
• Determine the mass of each individual chemical species in the sampling
container by the known mass of the sample and mass fractions. Mass of C02
in the process volumes from volume readings is determined by using volume
of the sampling container and mass of individual species in each of the
container; i.e., CO2 Mass flow rate = CO2 Volume flow rate *(Mass of
Representative Sample / Volume of Sample Container) * Mass fraction of
CO2 in the sample.
It is important to note that the sampling of liquid or supercritical CO2 should be
performed so the sample represents the process fluid. Some American Society for Testing
and Materials (ASTM) standards that detail procedures for sampling liquids, especially
hydrocarbons such as liquefied natural gas, may be applicable to conduct this sampling.
In addition, weighing scales may be sensitive to temperature difference between the
empty container and the container with the sample. Care must be taken to minimize such
error.
Facilities subject to the UIC program for permitting of injection wells would already have
flow meters installed to measure the flow rate of the CO2 stream injected for purposes of
compliance with their UIC permits. It is common industry practice to use volumetric
rather than mass flow meters.
In order to determine the portion of the CO2 stream that is CO2, the facility would sample
the stream and analyze it for its CO2 concentration. EPA has identified three industry
standards that provide methods to quantify CO2 that may be applicable to CO2 streams.
Facilities may use any standard method published by a consensus-based standards
organization if such a method exists, or an industry standard practice.
Gas Processors Association (GPA) Standard 2261-00. Analysis for Natural Gas and
Similar Gaseous Mixtures by Gas Chromatography
This standard describes a method for the compositional analysis of natural gas and
similar gaseous mixtures that also contain CO2, in addition to other gaseous constituents,
and various hydrocarbons. The method consists of physically separating the gaseous
26
-------
constituents by gas chromatography with a thermal conductivity type detector.
Procedures are outlined for component concentrations by mathematical procedures and
using calibration data from a reference standard. The method is applicable to mixtures
that confirm to component concentrations as outlined in the standard.
GPA Standard 2177-03. Analysis of Natural Gas Liquid Mixtures Containing Nitrogen
and Carbon Dioxide by Gas Chromatography
This standard describes a method for the compositional analysis of demethanized liquid
hydrocarbon streams that also contain CO2, in addition to nitrogen and various
hydrocarbons. The method consists of physically separating the liquid mixture
constituents by gas chromatography having a thermal conductivity type detector.
Procedures are outlined for component concentrations by mathematical procedures and
using calibration data from a reference standard. The method is applicable to mixtures
that confirm to component concentrations as outlined in the standard.
ASTM E1747-95 (re-approved 2005). Purity of Carbon Dioxide Used in Supercritical
Fluid Applications
This ASTM standard defines purity standards for C02 for use in chemical extraction and
chromatography applications. The guide defines standards for impurities in CO2 used in
extraction and chromatography and it suggests methods of analysis for quantifying these
impurities using gas chromatography with an electron capture detector and flame
ionization detector.
EPA recognizes that these standards are not specifically designed for the analysis of C02
for use in ER or GS applications. The GPA standards note that the methods use gas
chromatography to separate and identify compositional characteristics of the constituents
of the gas/liquid mixtures; however, they are recommended and applied for mixtures with
a relatively low concentration of CO2 (less than 20.0 mol percent in the case of GPA
Standard 2261-00, and less than 5.0 mol percent in the case of GPA Standard 2177-03).
Typical applications of CO2 for ER or sequestration may have CO2 concentrations greater
than 95 percent. There does not seem to be any indication in the standards themselves
that the standards would be inaccurate or even biased for high CO2 streams, but the
reporter should evaluate the suitability of these methods if they are used.
Using the operating temperature and pressure, the volume of CO2 can be converted into
STP conditions and, using a STP density value for CO2, from the NIST online database
(0.0018682 grams per milliliter or 0.11663 lb/ft3) one can calculate the CO2 quantity in
metric tons. A facility would apply the same method in order to calculate the quantity in
metric tons of CO2 received.
Measurements of CO2 volume may be taken at actual pressure and temperature
conditions and may therefore need to be converted to STP conditions for the purposes of
reporting under subpart RR. If the CO2 is not in a liquid or supercritical state the reporter
may apply the Ideal Gas Law to convert measurements at actual temperature and pressure
conditions to STP conditions. Standard Temperature is defined in the Rule as 60 degrees
Fahrenheit. Standard Pressure is defined in the Rule as 1.0 atmosphere.
27
-------
For example, if a CO2 volume of 1,000 actual cubic feet was measured at an actual
temperature of 300 degrees Fahrenheit and an actual pressure of 1.1 atmospheres, the
calculation to convert the measurement to the volume of CO2 at STP would be as
follows:
Volume Conversion using the Ideal Gas Law
The Ideal Gas Law is: P x V = n x R x T in which:
P = Pressure
V = Volume
n = mass (number of lb-moles)
R = Gas Constant
T = Temperature
For pressure in units of atmospheres, volume in units of cubic feet, and temperature in
degrees Rankine, the Gas Constant R = 0.730 atm ft3 / lb-mol deg. R.
Degrees Rankine is defined as Degrees Fahrenheit + 459.67.
From the Ideal Gas Law:
Pi x Vi / Ti = P2 x V2 / T2
Where:
Pi = Actual pressure
T1 = Actual temperature
Vi = Actual volume
P2 = Standard pressure
T2= Standard Temperature
V2 = Volume at Standard Conditions
For the example of a measurement of 1,000 cubic feet of CO2 at actual temperature and
pressure conditions of 1.1 atmospheres and 300 degrees Fahrenheit, the values for the
Ideal Gas Law equation would be as follows:
Ti = Actual Temperature = 300 + 459.67 = 759.67 Deg R
Pi = Actual Pressure =1.1 atmosphere
T2= Standard Temperature = 60 + 459.67 = 519.67 Deg R
P2 = Standard Pressure =1.0 atmosphere
Vi = Actual Volume = 1,000 cubic feet
The equation Pi x Vi / Ti = P2 x V2 / T2 would then be solved for the variable V2.
V2 = (Pi x Vi / Ti) / (P2 / T2)
28
-------
V2 = [(1.1 atm) X (1,000 ft3) / (759.67 Deg R)] / [(1.0 atm) / (519.67 Deg R)]
Therefore:
V2 = 752.48 cubic feet at standard temperature (60 degrees F) and pressure (1.0 atm)
For CO2 that is in a supercritical state, the Ideal Gas Law is not applicable and therefore a
different equation of state is needed to calculate the volume of the C02 at standard
conditions.
In such cases, the reporter may apply the NIST Thermophysical Properties of Fluid
Systems (accessible at http://webb00k.nist. gov/chemistry/fluid/) to look up the density of
supercritical CO2 at actual temperature and pressure conditions.
For example, for supercritical CO2 at an actual temperature of 140 degrees Fahrenheit (60
degrees Celsius or 333.15 degrees Kelvin) and an actual pressure of 100 atmospheres
(10.13 MPa or 1470 lb/in2) the NIST Thermophysical Properties database shows the
density of the supercritical C02 to be 18.614 lb/ft3.
Temperature
(F)
Pressure
(atm)
Density
(lbm/ft3)
State
140
100
18.614
supercritical
At the standard temperature of 60 degrees Fahrenheit and standard pressure of 1.0
atmosphere, the density of CO2 is 0.11663 lb/ft3.
So if a volume of 1000 ft3 of supercritical C02 was measured at actual pressure and
temperature conditions of 100 atmospheres and 140 degrees Fahrenheit, that volume of
C02 would be equivalent to: 1000 ft3 x 18.614 lb/ft3 = 18,614 lb of C02.
The density of C02 is 0.11663 lb/ft3 at STP. Therefore, 18,614 lb of C02 at a density of
0.11663 lb/ft3 = 18,614 lb / 0.11663 lb/ft3 = 159,598.7 ft3 C02 at STP.
Therefore, the volume of the 1,000 ft3 of C02 at actual conditions would be reported as
159,598.7 ft3 C02 under standard conditions.
29
-------
4. Overview of Monitoring Technologies for C02 Leakage Detection and
Quantification
This chapter describes technologies for monitoring of the injection well, subsurface CO2
plume, vadose zone2 , soil zone and vegetation, and atmosphere and how they may be
applied to detect and quantify movement of CO2 to the surface at GS sites. Table 4-1
provides a list of monitoring methods that are being used at CCS projects.26
Table 4-1: Monitoring Technologies as Deployed at 10 Existing CCS Projects
Technology
i
Desc r>iptk)o u
e
3
^ J
||
1
>:>
i
j
ill
U
§urfece-ba**d hwk l2D > 10 >'
Suif«^w4 vsvm umkJ to «TV4r1 e«e«gy *es**wer» w*esl en*^ waves X
bvh rur#fi fi«n mtesM una FrtroaM weugn tiw Aieriuflnw kvs mtivt rnivei s» mw nrw
to rescue slructures y. depf- ZD sewnc produce* a "tKe~ mage of the subsurface. #We JO
pfMhjtwt * niwft gg-caled *40* bmatapee Iwiii. d**a Pt iw&tired ttMMe
peneds r !i-«^ Id etetnae*) tn r* fcjbfejrtB-H! €ttr 1 If a
X )t
X
x 1 f i
. >¦
wtapy paffftmya w im ffjrjwcjiflw am 1 kwiwii awujfl twsa ^ *•* tfueau fft i.
r« of the formation a! 'Jepih rr poncuijf core sa«ptes can be anafyzed w\ T uprod* nJ c4tW 2 arret ol nicest irnbat n cfeyw in an uncased wefl afPttugh tome bji
0 9 amont bond k>g( org ¦~ifirrov SMrtormod rrwgh casing Tntw >091 hid used kr mttal ^uracMruaoon and
mnJ ntf rr m-w It- trK-ok"**-! *• Wc#» Wi«1 HWil •-* fr**
t-Jl rjection cease n orier ha rlroduce Ihp k^mg tool
K X
*
X
sT 1
IT ~5T
GeochwrwEai samptng & anaysia X
Tvw^rwinig Mb Frt vmn Iv rhvvj«% ul rhumnfry tfuf i»u*1 nVMnWt tjh byiruhon AihH ftttm fh«
X X
T
IT1-
F* 1
Z T
& mpKtxm rain msi^tmg
Staniins ftrd rnanagpmcnt wols used IB mentor foimaton v reevJaroe
bm kw tnoMwi (hal mw E RT um an atocrato ri tha (V* waHtM* ba vaad X
h r«e atecvodei eo measure fesistAvrry Data can be acquired art en and vta renuu nuking r« a good opaon ta
1ft Nri (WKi Arfc'tnl* wrWrw* tuifhwyt
X X
x
X
7
¦:
GullWrf HilllK.
Umi *0wcu and r*cftvai (Maoad «i f^r n*4i a^ncfi olkmt lc« gitalw latokAon mi f* aut*u1*:a. tut ot*r *
CeNkrm ma yyewa»
X X
V
T"
r i
r
Q«avrtjir1iU4i4 lurvwy*
Uanutaa ciiangaa to dw yavlat«pii«l fWO ri utiia* to «a«of*a thangna m fatmarton fto«j dan»Cy aa CO; B
(mne «i the Kymat*>n (war Iwne
X X
x"
jT
r«*rthv aenu iMJuwji* 4v i*»vw\q i,mwI) I mi rruvLjv At a fJTl,, mAMAwg tfvfiuw>na X
__
Trftmfier*
R*ao*v«* vniri gh»*v" m a*avabc»» lha giwnd vinfaca tav^ng l»pm CO, al dnA> uvng Uhi^r » pn B
fyt g'ixK*j
IT
T
nMfl (niftKAR
ihM radar in Hi*nQtt% «n ma «nvanr«*. nf ina gmiM fnvn <"jTi? ny*«w*i ai R
*pOi
X *
"3r
SkM 9M ¦' vadwe 2one i* shflBcw aw***
mondaonj
Uon^ofing wqft» m t* t*i9*nr wbwr^ace are used to «rivcl sam^vi. to checfc k* cKanfva «i COj «on5ent»aten». B
« we« as m '»lo» c» C traces and marten r^: rngM «d<*» COi ^ W
77
T
X
¦r t
flMH
mfMfBes man ore nqr/y detKuele at low co^st^raaom a-.i not tsiruiafy prmrt. n me sutHurfan or me iZO.
X X
T
X
stream are r«^*ice4 rta the m.iactaia pnor to rqvdien Deledtrt u«ed at wHace and ri ir^vmvdiata *nrtfr-
Manng mm can toc« lot mesa vaca gases to {ttcsmuva 4 CO, It leaking 1iom the storage nicvat
Alrnevywn; murriaiing
Mvn«uflng of Uni woftmod and vt*i< areas abuvu llw pbrm Tins tan r^udv eddy i»va«lanc* LIDAA and 9SNh B
bwa at CO. MMeuwi
TiY
T-
TT
" T
a Burini.wKlK!bUu«iftaH
(Source: J.J. Dooley, CL Davidson and RT Dahowski. 2009. An Assessment of the Commercial
Availability of Carbon Dioxide Capture and Storage Technologies as of June 2009. Joint Global Change
Research Institute. Pacific Northwest National Laboratory. June 2009. PNNL-18520)
While technologies for quantifying C02 surface leakage from GS sites are continuously
being refined, it is generally recognized that, when properly planned and implemented,
monitoring methods will be effective at detecting surface leakages."' 8 A wide range of
25 The vadose zone is the relatively shallow zone beneath the surface that is not saturated with groundwater.
26 J.J. Dooley, CL Davidson and RT Dahowski. 2009. "An Assessment of the Commercial Availability of
Carbon Dioxide Capture and Storage Technologies as of June 2009." Joint Global Change Research
Institute. Pacific Northwest National Laboratory. June 2009. PNNL-18520.
27 Benson S.M. 2006. Monitoring Carbon Dioxide Sequestration in Deep Geological Formations for
Inventory Verification and Carbon Credits. Society of Petroleum Engineers Paper 102833, 14pp.
http://www.onepetro.org/mslib/servlet/onepetropreview?id=SPE-102833-MS&soc=SPE (at cost).
FutureGen Alliance. 2006. Mattoon Site Environnemental Information Volume. December, 2006.
http://www.futuregenalliance.org/.
30
-------
techniques for monitoring GS have been used at GS sites as well as for a number of years
in other applications, including oil and gas production, ER, and plant and soil science.
These techniques may be used at a GS site to monitor the injected CO2, the surrounding
rocks and fluids, wells and equipment, and the surface conditions. EPA has concluded
that a GS facility would be able to propose a site-specific monitoring, reporting, and
verification (MRV) plan for leak detection and quantification based on the current
availability of monitoring technologies. In addition, it is expected that site
characterization and screening will lead to selection of sites that are suitable for long-
term sequestration, and that incidences of leaks to the surface may be infrequent at well-
selected and well-managed sites.
4.1 Monitoring of the Injection Well
Mechanical Integrity Testing (MIT) is commonly used at UIC permitted injection wells to
demonstrate that the injection wells can resist the high pressures of injection, as well as
the potentially corrosive nature of the injected C02. Please refer to the UIC program
Web site for information on requirements related to monitoring of UIC permitted
injection wells, including MIT.29
Periodic external MIT checks the area between the cement and the formation in the long
string casing to detect gaps or fluid flow.30 Should there be a break in the bond between
the cement and the rock of the wellbore, the injected C02 can make its way vertically to
access shallow formations, and possibly escape to the surface. External MIT is a reliable
and effective method commonly carried out using wireline logs or pressure tests.
Wireline logs are subsurface measurements of the wellbore taken by lowering an
instrument on a wire into the well, and recording the log response continuously as the
tool is pulled upward. This generates a detailed picture of the borehole and nearby rock.
There are numerous types of wireline logs used in the oil industry; the primary focus is
on determination of reservoir properties such as lithology, porosity, and fluid content (oil,
gas or water). Other types of tools are used to evaluate the integrity of the casing and
cement in a cased well. Wireline logs can be run in the open hole of a newly drilled well,
and may be used to evaluate conditions in injection wells or monitoring wells. Common
wireline logs used for external MIT are the cement bond log, the temperature log, and the
noise log. The cement bond log is a sonic (sound-wave) based tool that is used to
evaluate the bond between the casing and cement, and between the cement and the
formation.31'32 The temperature log helps identify hot spots due to flow from deeper
29 http://water.epa.gov/tvpe/groundwater/uic/wells seauestration.cfm.
30 Koplos, Jonathan, B. Kobelski, A. Karimjee, C.H. Sham. 2006, UIC Program Mechanical Integrity
Testing: Lessons for Carbon Capture and Storage. Fifth Annual Conference on Carbon Capture and
Sequestration, May 8-11, DOE/NETL, 2006. http://www.netl.doe.gov/publications/proceedings/06/carbon-
sea/Tech%20Session%20139.pdf.
31 Duguid, Andrew, and John Tombari. 2007. Technologies for Measuring Well Integrity in a C02 Field.
Sixth Annual Conference on Carbon Capture and Sequestration. DOE/NETL, May 2007. Available at:
http://www.netl.doe.gov/publications/proceedings/07/carbon-sea/data/papers/tue 064.pdf.
31
-------
formations, while the noise log detects the sound of CO2 flow behind the pipe. To some
extent, the log response might be used as a qualitative measure of the amount of leakage.
Other external integrity tests include the oxygen activation log and radioactive tracer
surveys, which indicate lack of fluid migration behind the casing.
Internal MIT of the injection well checks the integrity of the well materials: the tubing,
packers, and long string casing. During injection operations, there can be continuous
monitoring of the annulus fluid pressure, which is typically maintained at a higher
pressure than that of the injected CO2. A drop in the annulus pressure may indicate cracks
or holes in the long string casing, or may also indicate a leak in the tubing or packers.
This would not necessarily be an indication of a CO2 leak to the atmosphere. In addition
to continuous pressure monitoring in the annulus, internal well damage can be monitored
with periodic wireline logs, including caliper tools, radioactive tracer tests, and downhole
video. Logs can detect corroded or damaged tubing or casing even before a leak occurs.
All of these methods are commonly used in the oil and gas industry.
In certain circumstances, new technologies allow well logging to be conducted while
drilling using small, sophisticated tools.
4.2 Monitoring of Subsurface CO2 Plume
The location and size (areal extent) of the injected C02 plume can be evaluated using
several approaches. These include the acquisition of active seismic data, passive seismic
data, gravity data, and information from monitoring wells. This information can be used
in concert with reservoir modeling to predict subsurface CO2 plume movement. Please
refer to the UIC program Web site for information on requirements related to tracking
and monitoring the CO2 plume.33
The active seismic geophysical technique involves the generation of sound waves that
propagate downward or laterally through the subsurface, are reflected off of geological
layers, and are subsequently detected and analyzed at the surface by advanced
instrumentation. The method is used to evaluate the structural or spatial configuration of
the subsurface, and can also be used in some cases to determine reservoir properties and
fluid content, including the presence of free-phase (but not dissolved) CO2. In
conventional active seismic, both sound sources and detectors are located on the surface.
In vertical seismic profiling, the source is on the surface and the detector is downhole. In
another arrangement called cross-well seismic, both source and detector are downhole in
different wells. Periodic acquisition of active seismic data can in some cases be used to
detect subsurface CO2 movement within and outside of the injection zone (IZ). This may
include leakage from or around the injection well, from or around older pre-existing
wells, or through geological zones of weakness in the confinement zone (CZ).
32 DOE/NETL. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory. Available at:
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
33 http://water.epa.gov/tvpe/groundwater/uic/wells seauestration.cfm.
32
-------
A variety of active seismic methods are in use today in the oil and gas industry, and can
be used for CO2 plume monitoring. These include two-Dimensional (2-D) seismic, three-
Dimensional (3-D)/four-Dimensional (4-D) seismic, vertical seismic profiling, and cross-
well seismic. Surface seismic data may be either 2-D, 3-D or 4-D. 2-D seismic is obtained
with a linear surface arrangement of receivers. The acoustic energy source is from small
explosive charges or from "vibro-seis" trucks. 3-D seismic may provide better resolution,
accuracy, and tracking of the CO2 if it migrates from the storage reservoir into the
overburden. 3-D seismic evaluates a large volume of the subsurface, has high resolution,
and is more complex and more costly to obtain and analyze. Surface receivers are laid in
a grid pattern, and large amounts of data are simultaneously recorded and subsequently
processed. Sophisticated software allows the conversion of this raw information into a
detailed subsurface representation. 3-D data are being increasingly used in the oil
industry, as they allow the mapping of structural details, small faults, rock properties, and
fluid properties. 4-D seismic is time-sequenced 3-D. The technique is the same as 3-D,
but it is acquired over time. In this way, one can evaluate fluid movement over time.
Vertical seismic profiling uses sensors deployed in a borehole to measure sound
propagation in the immediate vicinity of the well.34 It is a high-resolution method that
evaluates a small volume of the subsurface adjacent to a single well. A variation of
vertical seismic profiling is "walk-away" profiling, in which the source is sequentially
moved away from the well. This creates a mini-2-D seismic line away from the injection
well. Cross-well seismic, as mentioned above, involves the transmission of acoustic
waves between wellbores, and also evaluates a small rock volume. Vertical seismic
profiling and cross-well seismic methods are used to develop a very high-resolution
image of conditions relatively near the injector or monitoring wells, due to the short
distance between seismic sources and receivers. These methods can also be used to
calibrate the acoustic signature of the CO2 plume under very controlled conditions. This
calibration can then be used to better interpret the 2-D and 3-D seismic, which cover a
much larger area.
All these types of active seismic methods can be used to image the free-phase portion of a
CO2 plume under certain subsurface conditions. These methods are known to be effective
in many cases where CO2 is injected into a saline reservoir and has a good density
contrast with the saline fluid. The ability to image and track a CO2 plume depends upon
factors such as lithology (sandstone, limestone, coal, or shale), porosity (pore space),
fluid content, and depth.35 Seismic detection of a CO2 plume in a carbonate (CO3"")
(limestone or dolomite) reservoir may be a challenge due to low porosity and other
factors. Seismic detection of CO2 in a depleted oil reservoir may be more challenging
than monitoring a saline reservoir due to lower average fluid density contrast (both oil
34 DOE/NETL. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory. Available at:
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
35IPCC. 2005. Carbon Dioxide Capture and Storage. Intergovernmental Panel on Climate Change, pg. 237
Available at:
http://www.ipcc.ch/publications and data/publications and data reports carbon dioxide.htm.
33
-------
and free-phase CO2 are less dense than saline water). Studies have been undertaken to
evaluate seismic plume monitoring in coal beds.36
Active seismic methods including 2-D, 3-D/4-D, and cross-well seismic can achieve high
plume detection resolution, and therefore can be used to develop a first-order estimate of
plume dimensions for leakage quantification; a drawback is that seismic methods only
image free-phase CO2 in the reservoir and not the dissolved CO2 component. In addition,
the amplitude of the seismic signal (which is used to detect the C02) is only partially
related to the concentration and therefore cannot fully quantify even the free-phase part
of the C02. However, amplitude anomalies have been observed in C02 sequestration
monitoring that have been related to C02 concentration. Such amplitude effects have
been observed at the Sleipner and Weyburn projects.37 To be detectable, the C02
accumulation must have lateral and vertical dimensions sufficient to produce a
discernible seismic response. One study based on theoretical resolution considerations38
has suggested that C02 buildups as small as 10,000 to 20,000 metric tons may be
detectable at typical injection depths, but amounts would be difficult to quantify, as
saturation would remain a key uncertainty. In practice, results from the Sleipner time-
lapse surveys suggest that repeatability noise (which depends on the accuracy with which
successive surveys can be matched), rather than resolution, may be the key parameter in
limiting the detection of small changes of seismic signature due to leakage. Under
favorable conditions, such as those at Sleipner, Weyburn and Frio, accumulations on the
order of 1,000 to 10,000 metric tons of C02 are detectable at depths less than 1,000
meters (m), and smaller accumulations could be detected at shallower depths.39
Passive seismic methods, which use only receivers (no active acoustic sources), monitor
the sound waves that are emitted from high-pressure injection sites with sophisticated
instrumentation and decipher the results using computational methods. As the pressurized
CO2 moves through the reservoir, it creates microfractures that release acoustic energy
that is detected in 3-D space. Passive seismic methods can be used to track the subsurface
plume and the areas of high pressure. They can also be used to detect where fractures are
occurring that might signal breaks in the CZ or activation of old faults.
Gravity surveys measure the earth's gravitational field at a series of points over a
subsurface target of interest. Gravity surveys have been used since the early 1900's in oil
exploration, and are now being evaluated for use in GS sites. Changes in the earth's
36 Lawton, Don. 2002. A 3C-4D Surface Seismic and VSP Program for a Coalbed Methane and C02
Sequestration Pilot - Red Deer, Alberta. CREWES Research Report 14. Available at:
http://www.crewes.org/ForOurSponsors/ResearchReports/2002/2002-64.pdf.
37 DOE. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory. Available at:
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
38 Myer, L. R., Hoversten, G. M., and Gasperikova, E. 2002. Sensitivity and Cost of Monitoring Geologic
Sequestration Using Geophysics. In: Gale, J. & Kaya, I. (eds.) Greenhouse Gas Control Technologies.
Elsevier, 377 - 382. Available at: http://www.sciencedirect.com/science/book/9780080442761.
39 Benson, S.M.. 2006. Monitoring Carbon Dioxide Sequestration in Deep Geological Formations for
Inventory Verification and Carbon Credits. Society of Petroleum Engineers Paper 102833, 14pp. Available
at: http://www.energy.utah.gov/Utah Actions/documents/dec2006/spel02833.pdf.
34
-------
gravitational field are caused by changes in density in the underlying rock layers.
Because C02 is usually less dense than native fluids and petroleum, reservoirs that fill
with CO2 will appear as lower gravity areas. Gravity surveys are performed in a time
lapse mode like 4-D seismic, to evaluate plume movement. Data processing is complex
and involves removal of several external influences from the data set prior to creating a
model of the gravity field. Studies at Sleipner site indicate the method shows promise to
help constrain the reservoir simulation models40. This technology is still in early stages of
field testing and model development.
Downhole instrumentation is used in monitoring wells to measure temperature, pressure,
conductivity/salinity, and fluid characteristics. The individual measurements are quite
reliable and accurate, but their limitation is that they represent a single point in time and
space. The monitoring wells can be logged periodically with wireline logs that can be
used to detect the presence and depth of C02. Certain wireline tools may be used to
collect a fluid sample under reservoir pressure and temperature conditions, and retrieve
the sample to the surface for laboratory analysis.41 They can also be used for cross-well
seismic data acquisition, in which seismic signals are broadcast from one well and
detected in an adjacent well to characterize the formation between the wells. Monitoring
wells can be designed to allow testing of the CO2 reservoir, or of units above the
reservoir. U-tube devices can be used to retrieve pressurized samples for laboratory
testing. Based on the specific (electric) conductance and the pH of the pressurized
samples, the concentration of dissolved inorganic carbon (DIC) can be determined using
complex algorithms. DIC includes aqueous CO2, carbonic acid (H2CO3), bicarbonate
(HCO3"), and C03"". C02 leakage from the reservoir can be detected long before the C02
can escape to shallower zones or to the surface.
Geochemical sampling of the fluid from monitoring wells can also be used to detect
natural or artificial "tracers" in the injected CO2. Monitoring to gather additional data
may be necessary if the instrumentation indicates a significant anomaly relative to
background levels.
Intermediate monitoring wells can be used to test zones within a depth range of
approximately 200 to 2,000 feet below the surface. As with the deeper wells, downhole
instruments can measure pressure, temperature, and conductivity/salinity. Further
monitoring may be implemented when readings depart from background levels; for
example, a 3-D seismic survey may be undertaken if the anomaly is within an appropriate
depth range.
Surface deformation monitoring is currently being tested as a rapid cost effective method
to identify plume movement and potential leakage pathways in the subsurface. As CO2 is
injected the ground surface may be raised a minute amount in response to the addition of
fluids to the subsurface. Highly sensitive tiltmeters (much like a highly sensitive
40 Arts, R., Chadwick, A., Eiken, O., Thibeau, S., Nooner, S., 2008. Ten years' experience of monitoring
C02 injection in the Utsira Sand at Sleipner, offshore Norway, First Break, Volume 26, January 2008
41 FutureGen Alliance. 2006. Mattoon Site Environnemental Information Volume. Available at:
http://www.futuregenalliance.org/news/fg mattoon eiv vl master revl.pdf.
35
-------
electronic carpenters level) and Differential Global Positioning Systems can be used to
measure ground movements over small areas, and specialized radar (Interferometric
Synthetic Aperture Radar (InSAR)) can be used to survey large areas from the air.
Tiltmeters are capable of detecting changes in slope on the order of one millimeter (mm)
over 1000 kilometers. InSAR satellite images are collected monthly and can cover from
2,500 to 10,000 square kilometers (km) with accuracy in the millimeter range.42 When
combined with tiltmeter and global positioning system (GPS) data, accuracy and
resolution can be improved. Surface deformation at In Salah was monitored using InSAR
from 2003 to 200843 and showed ground swelling on the scale of 3 to 14 mm per year
near the injection wells, and several millimeters of subsidence near the natural gas
production wells. The data can be used to infer underground pressure gradients and fluid
movements from the changes in surface deformation over time. Surface deformation
surveys are influenced by vegetation, frost heaves and other natural conditions that can be
removed from the data set with additional information.
4.3 Vadose Zone Testing
The vadose zone is the relatively shallow zone beneath the surface that is not saturated
with groundwater. Due to absence of water, monitoring in this zone therefore is limited to
testing the chemistry of the gases contained within the pore space. The C02 concentration
of gas samples taken in this zone can be measured using commercially available infrared
gas analyzers (IRGAs), which measure the absorption of specific portions of the infrared
spectrum to determine the concentration of CO2 (or other gases) in the sample.
Background levels must first be measured to determine the statistical variations of CO2
concentration accounting for sample temperature, seasonal variations, diurnal variations,
and the like. Later, a presumptive leak is detected if the monitored CO2 concentration
exceeds a value corresponding to a very high (e.g., exceeds the 99 percentile) prediction
for the natural CO2 concentration accounting for sample temperature, seasonal variations,
diurnal variations, etc. Should significant anomalies be detected, an expanded soil or
atmospheric testing program may provide additional information. In addition to the use of
IRGAs for CO2 concentration, vadose zone sampling can include testing for tracers,
should they be used, and for sampling for carbon and oxygen isotopes, to determine
whether the CO2 is that which is being injected or naturally occurring CO2.44 As with
data collected in monitoring wells, these individual measurements are quite reliable and
accurate, but their limitation is that they represent a single point in time and space.
4.4 Soil Zone and Vegetation Testing
The soil zone is generally present within the first few inches to possibly tens of feet
beneath the surface, the uppermost layer of the vadose zone. The soil zone is the fertile
42 McColpin, L. 2009. Surface Deformation Monitoring as a Cost Effective MMV Method. Energy
Procedia, 1, (2009) pp. 2079-2086
43 Onuma, T, and Ohkawa, S. 2009 Detection of surface deformation related with C02 injection by
D InSAR at In Salah, Algeria Energy Procedia, 1, (2009) pp. 2177-2184
44 FutureGen Alliance. 2006. Mattoon Site Environmental Information Volume. Available at:
http://www.futuregenalliance.org/news/fg mattoon eiv vl master revl.pdf
36
-------
portion that can retain water (i.e., topsoil). A monitoring program in the soil zone can
detect the vertical C02 flux. It is especially important in the soil zone that background
levels (which vary with time of day and season) be determined to provide a well-
characterized baseline against which statistically significant anomalies can be detected
and estimated.
Accumulation chambers can be used to estimate CO2 flux from the ground surface to the
atmosphere, and be used to directly detect and quantify the mass of C02 leaking from the
subsurface.44 The approach may consist of a grid of accumulation chambers in which the
C02 flux is periodically measured.45 Commercially available accumulation chamber
instrument packages are currently being used to track seasonal variations in CO2 flux as
part of climate change studies. The accumulation chamber is a method of measuring soil
CO2 flux that involves the placement of a collection chamber directly on or into the soil
surface, with the rate of C02 accumulation measured periodically with an IRGA. The rate
of change in CO2 concentration defines the rate of flux. The calculated rate of flux can be
used as a standalone quantification method if C02 leakage is detected. This type of
measurement could be used to quantify leakage that is moving up through the soil zone
across a wide area, as opposed to point sources such as a leaking injection well. To
account for local variations in the underlying geologic, soil characteristics, and soil
moisture levels, several accumulation chambers would have to be distributed over the
area to quantify the leakage rate. CO2 flux measurement systems based on accumulation
chambers are commercially available and can provide a measurement of anomalous C02
flux at a point, with a precision better than 5 micrograms per square meter per second.
That point measurement precision corresponds to a leak rate of 25,000 metric tons per
year distributed across an area of 150 square kilometers.
Surface leaks detected in the soil zone may also be quantified through tenting methods, in
which large tents (made of impermeable material) are used to capture and accurately
measure the C02 leakage from a much larger soil surface area. Tenting methods have
been used to detect CH4 leaks from natural gas pipelines and could also be used for CO2
quantification. Variations of this method are commonly used in measuring CH4 flow rates
from equipment leaks.46 A tent or bag is constructed above a known leakage site. An
inert gas such as nitrogen is conveyed through the bag at a known flow rate. Once the
carrier gas attains equilibrium, a gas sample is obtained from the bag and the CH4 content
of the sample is determined. The leak flow rate is calculated from the purge flow rate
through the enclosure and the concentration of CH4 in the outlet stream. Table 4-2
illustrates the general approach to quantify the leak.
45 Oldenburg, C.M., Jennifer L. Lewicki, and Robert P. Hepple. 2003. Near-Surface Monitoring Strategies
for Carbon Dioxide Storage Verification. Earth Sciences Division, Lawrence Berkeley National
Laboratory, publication LBNL-54089. Available at: http://www.osti.gov/bridge/servlets/purl/840984-
dTw752/native/840984.pdf.
46 Envirotech Engineering. 2007. Review and Update of Methods Used for Air Emissions Leak Detection
and Quantification. February 5th, 2007. Available at:
http://eipa.alberta.ca/media/31357/1666%20review%20and%20update%20of%20methods%20used%20for
%20air%20emissions%201eak%20detection%20and%20auantification%20-%20final%20report.pdf
37
-------
Table 4-2: Example Leak Quantification Using Tenting and Known Test Gas Flow47
Inlet Flow of Conveyed
Test Gas
(cu. ft./minute)
Measured C02
Concentration at Outlet
(%)
Calculated Outlet
Flow Rate
(cu. ft./minute)
Calculated C02
Leak Rate
(cu. ft./minute)
10
50%
20
10
10
10%
11.1
1.1
This approach is limited to sampling a point source or small area of leakage. However, if
the site of the leakage can be mapped by airborne methods such as are described below, it
should be possible to set up a number of tent measurement locations such that the leaks
over a large area can be quantified using statistical sampling. However, the cost and time
to set up a large number of tents and/or accumulation chambers might favor the
development and use of innovative mobile quantification methods (e.g., radial plume
mapping with path-integrated optical remote sensing).
CO2 in the shallow subsurface can also be detected through shallow wells drilled to
sample the groundwater. Such wells may be on the order of only several hundred feet
deep. Water samples collected from the wells are analyzed for a wide range of parameters
that may reflect changes due to the presence of CO2. Some indicator parameters such as
pH, alkalinity, electrical conductance and dissolved oxygen can be measured in the field,
and other parameters (trace metals, dissolved organic carbon, organic compounds and
isotopes) are submitted to a laboratory for analysis. Pilot tests and controlled release
experiments demonstrate that the rapid and significant changes in chemical parameters
were observed in response to the presence of CO2.48 As with soil testing, background
levels must first be established in order to detect changes that may indicate CO2 leakage.
Vegetative stress can also be a leakage indicator. Vegetative stress can be detected
through tower-mounted or airborne imaging instruments, such as digital color infrared
orthoimagery or hyperspectral and multispectral imaging, to detect changes to vegetation
coloration in the visible, infrared, and ultraviolet spectrum. These are relatively new
technologies with uncertainty in the interpretation of the results. Recent work has shown
that vegetation in the vicinity of a leak may be negatively affected, but vegetation farther
from a leak may be positively affected.49 A controlled release experiment at the Zero
Emissions Research and Technology (ZERT) facility in Montana reported that the
surface area of CO2 leakage hot spots could be delineated to within 2.5 meters using
portable hyperspectral imagers. The plant response was dependent on plant species but
was discernable when soil concentrations reached 4 to 8 percent CO2 concentration.50
47 The Calculated Outlet Flow Rate (cu. ft./minute) is inlet flow volume divided by concentration. The
Calculated C02 Leak Rate (cu. ft./minute) is calculated as outlet flow minus inlet flow.
48 Kharaka, Y., J. Thordsen, E. Kakouros, G. Ambats, W. Herkelrath, S. Beers, J. Birkholzer, J. Apps, N.
Spycher, L. Zheng, R. Trautz, H. Rauch, and K. Gullickson 2009. Changes in the Chemistry of Shallow
Groundwater Related to the 2008 Injection of C02 at the ZERT Field Site, Bozeman Montana. Environ.
Earth Sci. (2010) 60:273-284
49 Rouse, J.H. 2008. Measurements of Plant Stress in Response to C02 Using a Three-CCD Imager.
Montana State University. Available at: http://etd.lib.montana.edu/etd/2008/rouse/RouseJ1208.pdf
50 Male, E.J., W.L. Pickles, E.I. Silver, G.D. Hoffmann, J. Lewicki, M. Apple, K. Repasky, and E.A.
Burton, Using hyperspectral plant signatures for leak detection during the 2008 ZERT C02 sequestration
38
-------
4.5 Atmospheric Monitoring
The atmosphere above a GS site may be monitored with an array of C02 detectors
(closed-path point measurements and open-path line measurements), tracer gas and CO2
isotopic measurements, eddy covariance methods, Raman LIDAR (Light Detection and
Ranging), or Differential Absorption LIDAR (DIAL). In all methods, CO2 concentration
data are integrated with data on wind speed and direction to determine the vertical C02
flux and locate the CO2 leakage footprint. As noted previously, establishing baselines are
important to distinguish between natural C02 fluxes and C02 flux due to leakage from
geologic storage.51
CO2 detectors are commercially available for short closed-path and short open-path
(point) measurements, as well as long open-path (radial line) measurements. Similar
detectors have been integrated into stationary, mobile, and airborne monitoring packages
that are commonly used in combination with high-resolution GPS to detect and quantify
CH4 leaks in areas with road access. While these packages have not been widely tested
for C02, various types of C02 monitors are commercially available, generally relatively
low cost, reliable, and could be used in these applications. The technologies include
IRGAs (including Fourier transform infrared (FTIR) and non-dispersive infrared
analyzers), tunable diode lasers (TDLs), cavity ring-down techniques, and others. The
sample path can range from 10 cm to 1 km, by reflecting a laser beam off retro-reflecting
mirrors. These devices measure the gas concentration, and, when packaged with
measurements of wind speed and wind direction, they measure the total gas flow. The
method is described in detail in EPA Other Test Method 10 —Optical Remote Sensing
for Emission Characterization from Non-Point Sources.52 The protocol describes three
methodologies, each for a specific use:
¦ Horizontal radial plume mapping (for locating the source of emissions or hot
spots);
¦ Vertical radial plume mapping (for estimating the rate of gaseous emissions
from an area fugitive emission source); and
¦ One-dimensional radial plume mapping (for profiling pollutant concentrations
along a line-of-sight which is downwind of an emission source).
These methodologies use an open-path path-integrated optical remote sensing system in
multiple beam configurations to measure path-integrated concentration data. The protocol
suggests four scanning systems:
field experiment in Bozeman, MT. Environmental Earth Sciences, 60, 251-261, doi: 10.1007/s 12665-009-
0372-2, 2010.
51 David Etheridge, Ray Leuning, Donald de Vries, Kevin Dodds, Cathy Trudinger, Colin Allison, Paul
Fraser, Fred Prata, Cirilo Bernardo, Mick Meyer, Bronwyn Dunse and Ashok Luhar. 2005. Atmospheric
monitoring and verification technologies for C02 storage at geosequestration sites in Australia. C02CRC
Report No: RPT05-0134, Prepared by CSIRO Marine and Atmospheric Research for CANSYD Australia
and the CRC for Greenhouse Gas Technologies (C02CRC). Available at:
http://www.co2crc.com.au/publications/index.php.
52 http ://www. epa. gov/ttn/emc/prelim. html.
39
-------
¦ Open-Path FTIR Spectroscopy;
¦ Ultra-Violet Differential Optical Absorption Spectroscopy;
¦ Open-Path TDL Absorption Spectroscopy; and
¦ Path-Integrated Differential Absorption LIDAR.
EPA is also testing mobile versions of these systems for methane leak detection, in which
the detectors are integrated onto ground or airborne vehicles with high-resolution GPS to
aim the lasers at the retro-reflectors and correctly locate the methane leak.
Other atmospheric monitoring for leakage detection and quantification approaches
include tracer gases and CO2 isotopic measurements. The incorporation of trace
amounts of chemical compounds into the injected CO2 may be useful in circumstances in
which a suitable baseline for a natural parameter (e.g., soil C02 flux or surface-to-
atmosphere CO2 flux) cannot be established due to the parameter's wide and
unpredictable natural variation. In such cases, tracers could be added to the injected C02
and continuous or periodic monitoring could be conducted to determine if the tracer
could be identified outside the IZ in water, soil or air. The UIC Class VI rule does not
require the use of tracers.
Tracers can be manmade molecules such as perfluorocarbons (PFCs) and sulfur
hexafluoride, or noble gases such as helium (He), argon (Ar), and xenon (Xe), including
specific isotope concentrations. Tracers can be detected at levels of a few parts per billion
or even parts per trillion.
PFC tracers have been applied in several pilot studies and experimental tests and appear
to have good detection capabilities for leakage detection and potentially for quantification
in both soil gas53 and the atmosphere.54 The CO2 injection pilot test at West Pearl Queen
ER site in New Mexico used PFC tracers to detect and quantify leakage from injection
wells at soil gas monitoring points within 300 meters of the injection well.55 PFC tracer
analysis has also been used in monitoring wells in the IZ to map the location of the
advancing CO2 plume56 in the Frio Brine Test, and to monitor for potential surface
leakage at the Zama Acid gas injection project in Alberta.57
53 Strazisar B.R., Wells A.W., Diehl J.R. (2009) Near surface monitoring for the ZERT shallow C02
injection project. Int J Greenh Gas Control 3(6):736-744. doi:10.1016/j.ijggc.2009.07.005
54
Wells, Arthur, Strazisar, B., Diehl, J., and Veloski, G. 2010. Atmospheric tracer monitoring and surface
plume development at the ZERT pilot test in Bozeman, Montana, USA, Environ. Earth Sci. Volume 60,
Number 2, 299-305, DOI: 10.1007/sl2665-009-0371-3
55 Arthur W. Wells, J. Rodney Diehl, Grant Bromhal, Brian R. Strazisar, Thomas H. Wilson, Curt M.
White. 2007. The use of tracers to assess leakage from the sequestration of C02 in a depleted oil reservoir,
New Mexico, USA. Applied Geochemistry, Volume 22, Issue 5, May 2007, Pages 996-1016, ISSN 0883-
2927. Available at: http://www.sciencedirect.com/science/article/B6VDG-4N3GFGS-
3/2/fl6al33aleda34353elcbcc8dd05c917.
56 McCallum, S.D., Riestenberg, D.E., Cole, D.R., Freifeld, B.M.,Trautz, R.C., Hovorka, S.D., and Phelps,
T.J. 2005. Monitoring Geologically Sequestered C02 during the Frio Brine Pilot Tests Using
Perfluorocarbon Tracers. Fourth Annual Conference on Carbon Capture and Sequestration, May 2005.
57 Smith, Steven A., J.A. Sorensen, A. A. Dobroskok, B. Jackson, D. Nimchuck, E. N. Steadman, and J. A.
Haiju. 2009. Injection of Acid Gas (C02/H2S) into a Devonian Pinnacle Reef at Zama, Alberta, for
40
-------
The amount of tracer added to the injected C02 will depend on the minimum detectable
amount within the medium to be monitored and the probable leakage volumes and
dispersion patterns for the relevant leakage pathways. It also depends on how the tracers
will move along the leakage pathway relative to the leaking CO2. Therefore, if tracers are
proposed to be used, the monitoring strategy should consider the proposed tracers
characteristics in terms of:
• dissolution in water more/less than CO2;
• sorption into coals/shales more/less than C02;
• chemical reactions with fluids and rocks; and
• physical separation due to differences in molecular size and weight.
In some cases, tracers can move more quickly than the C02 and will reach sensors or
sample collection points well before any CO2 leak. In other instances a tracer might be
trapped along the leakage pathway, reducing its effectiveness. Certain tracers are GHGs
and if leaked, would contribute to total GHG emissions.
Another consideration for tracers is the possibility of contamination of sensor locations or
sample collection point through:
• spills during transport and loading of the tracer;
• equipment leaks and vented emissions from surface equipment;
• equipment (tubing) brought to surface for repair and replacement;
• fluid retrieved from IZ for analysis; and
• old leaks that have been repaired.
Therefore a monitoring strategy that includes the use of tracers should include ways of
preventing contamination by misplaced tracers through careful design of where and how
tracers are injected and how equipment and fluids that contact the tracers are handled.
Considerations should be made on what could be done in the event of contamination
including options for:
• moving sampling point away from contaminated locations (e.g., from surface
to subsurface);
• changing type of tracer used; and
• measuring a tracer "baseline" and looking for deviations from the baseline.
Alternatively, CO2 isotopic measurements are able to distinguish between CO2 derived
from fossil fuels and naturally occurring atmospheric and underground CO2. Researchers
in Australia note that even relatively large leaks at the well could dissipate rapidly away
from the source and be indistinguishable from background concentrations of CO2. They
suggest that the use of radioactive tracers or natural carbon isotope ratios in the
Enhanced Oil Recovery and Carbon Sequestration, Search and Discovery Article #40355. Adapted from
extended abstract prepared for oral presentation at AAPG Convention, San Antonio, TX, April 20-23,
2008.
41
-------
sequestered fluids will improve detection sensitivity.58 Field studies of carbon ratios in
injected and recovered C02 have been conducted at Weyburn-Midale C02 Monitoring
and Storage Project (Saskatchewan), the Pembina Cardium CO2 Monitoring Project
(Alberta), and the Lost Hills oil field (California). Recently a commercially available
carbon isotope instrument was modified for portable use and deployed during the
controlled release field tests at the ZERT site in Bozeman Montana.59 The advantage of
the technique is that it provides essentially real time results of the carbon isotope
concentrations across a potential leak location. The results showed that the portable
instrument was capable of real time spatially distributed measurements of carbon isotopes
to detect leakage for a subsurface source. The authors note that other techniques would
need to be used to quantify the leakage rate and background fluxes may complicate the
interpretation of results.
Eddy Covariance systems can measure the vertical CO2 flux in the atmosphere. They
combine an open-path IRGA on a tower alongside a sensitive, high-speed, 3-D
anemometer, a device used for measuring instantaneous wind speed and direction. Time
series data are recorded and evaluated using computer methods. The size and shape of the
sampling footprint (the surface area that is sampled by the instrument) is derived
mathematically from the anemometer data. A typical station consists of sensors mounted
on a tower from several meters to 30 meters or more high. The stations can be operated
with solar power and can be set up for data telemetry for transmission to a central facility.
Deployment of a grid of such detectors over an area provides information regardless of
wind direction. CO2 concentration data are integrated with meteorological data including
wind speed and direction, relative humidity, and temperature.
Such methods have the ability to measure C02 concentration in the atmosphere over
relatively large areas. A limitation of the method is that it assumes a horizontal and
homogeneous surface to interpret the data. Variations in plant cover, land use, and
topography may create challenges. As with soil sampling, the rate of leakage of CO2 from
the site must be a statistically significant anomaly above the background variability.
Therefore, to provide the required precision, eddy covariance towers would have to
collect data over an appropriate time period to provide a baseline which includes diurnal
and seasonal variations. With careful installation and the collection of baseline data, the
eddy covariance method can potentially provide a precise average of CO2 flux over a
large surface area. Although eddy covariance systems are currently research tools
assembled from a variety of commercially available components, the eddy covariance
method may eventually provide the anomalous CO2 flux averaged over a large surface
area, with an precision less than 50 micrograms per square meter per second, which
58 Paul Steele, Zoe Loh, David Etheridge, Ray Leuning, Paul Krummel, Aaron Van Pelt. 2008. Continuous
Greenhouse Gas and Isotopic C02 Measurements via WS-CRDS-based Analyzers: Investigations in Real
Time Monitoring at C02 Geological Storage Sites. Presented at American Geophysical Union, Fall
Meeting 2008, abstract #U41C-0024, poster available at:
http://www.picarro.com/assets/docs/AGU poster sequestration.pdf
59
Krevor, S., Perrin, J.-C., Esposito, A., Rella, C., Benson, S., 2010. Rapid detection and characterization
of surface C02 leakage through the real-time measurement of [delta] 13 C signatures in C02 flux from the
ground. International Journal of Greenhouse Gas Control 4, 811-815.
42
-------
corresponds to a leak rate of 25,000 metric tons per year distributed across an area of 17
square kilometers.
Another potential technology for measurement of CO2 in air is Raman LIDAR. LIDAR is
the optical analog of radar, and it is based upon the use of laser radiation to measure
various compounds in the air, including C02. The Raman LIDAR method involves
transmitting laser light into the atmosphere and then detecting the scattered laser radiation
that has been shifted in wavelength due to interaction with the target scattering molecules
(e.g., CO2) along the resolved path length. By comparing the Raman signal of the CO2 to
the Raman signal of nitrogen or oxygen, a direct measurement of C02 concentration can
be obtained.
A similar method, DIAL, would use two wavelengths of laser light to measure the CO2
concentration in the atmosphere. The wavelengths used are specific to C02. One
wavelength is selected to correspond to a CO2 spectral absorption line, while the other is
a non-absorbing wavelength. The average C02 concentration over the path length can be
determined from the ratio of the backscatter signals for the two laser wavelengths. The
instrumentation for both methods (Raman LIDAR and DIAL) can be ground, truck,
helicopter, or airplane mounted and can provide similar precision at similar cost. Truck
methods can cover up to tens of square kilometers per day. Helicopter and airplane
mounted platforms can cover a much larger area.
Airborne DIAL methods are currently in use for methane leakage detection along natural
gas pipelines.60 Instrumentation includes the LIDAR, a digital mapping camera, a color
video system, and an optical guidance system. The airplane flies a survey over the
pipeline at an altitude of about 1,000 feet, approximately perpendicular to the wind
direction. The LIDAR instrumentation measures the concentration of methane by
measuring how much of the reflected laser pulse has been absorbed. These data could be
integrated with real time wind direction and velocity to define the leakage footprint and
estimate a leakage rate.
While DIAL is commercially available for methane, Raman LIDAR and DIAL are
currently under development for commercial CO2 applications. Etheridge61 reports that
ground-based and airborne LIDAR and DIAL technologies appear to be feasible
technologies when used in conjunction with other sources of information, however
additional research is needed to improve precision for measuring and monitoring leaks at
geosequestration sites. Researchers at Montana State University are developing and
testing a portable DIAL with the ability to monitor several square kilometers at a
60 Brake, D. 2005. Detection and Measurement of Fugitive Emissions Using Airborne Differential
Absorption Lidar (DIAL). EPA Gas Star Program Annual Implementation Workshop. Available at:
http://www.methanetomarkets.org/expo china07/docs/postexpo/oag brake.pdf
61 Etheridge, D, Leuning, R, de Vries, D, Dodds, K, Trudinger, C, Allison, C, Fraser, P, Prata, F, Bernardo,
C, Meyer, M, Dunse, B and Luhar, A, 2005. Atmospheric monitoring and verification technologies for
C02 storage at geosequestration sites in Australia. Cooperative Research Centre for Greenhouse Gas
Technologies, Canberra, Australia, C02CRC Publication Number RPT05-0134. 83pp. Available at:
http://www.co2crc.com.au/publications/index.php.
43
-------
resolution on the order of 100 m.62 Airborne methods may also be integrated with
ground methods to provide the best approach for quantification. Airborne methods could
be used to map out the surface leakage area, and ground methods (such as accumulation
chambers and tents) could be installed on the surface leakage area to measure the
combined background and anomaly CO2 flux and outside the surface leakage area to
measure just the background C02 flux.
4.6 Leakage Detection and Quantification in the Offshore Environment
Reservoir modeling and monitoring techniques, such as seismic and monitoring wells,
can also be applied to the detect leakage in subsurface geologic formations and shallow
ocean bottom sediments located offshore. However, the fate of the C02 entering the
ocean from an offshore geologic formation is dependent on the temperature, pressure, and
C02 flux, and may result in the formation of C02 hydrates, may dissolve into the
seawater, or may be released into the atmosphere. Shallow monitoring systems aim to
detect and quantify C02 that has migrated into the shallow sediments and related water
bearing strata or seabed, and, ultimately, into the seawater or atmosphere. Shallow
monitoring includes those methods that detect and measure C02 in the subsurface (e.g.,
potable aquifers, soil, sub-seabed) and those that actually measure C02 in the water
column or atmosphere. 63
Sparker surveys detect reflections of acoustic signals produced from electro-capacitive
sources that penetrate several hundred meters beneath the seafloor. Boomer surveys
detect reflections of acoustic signals produced from electromagnetic (EM) sources that
penetrate about 100 meters beneath the seafloor. Both methods can potentially resolve
bed thickness of a meter or less, and would likely have considerable potential for
resolving small amounts of gas. However sparker and boomer data are acquired along 2-
D profiles, which render them less effective for systematic areal detection of
undiscovered leaks. Their main use may lie in high-resolution imaging of shallow
features previously detected on 3-D conventional data.
Shallow sedimentary deposits can also be examined using high-resolution acoustic
imaging. The technique should reveal anomalous features in the top 100 meters beneath
the seafloor as well as direct imaging of rising gas bubbles in the water column.
Transducers such as Compressed High Intensity Radar Pulse Sonars, operating at
frequencies between 2 and 8 kilohertz, have been widely employed on the continental
margins, and have also been used to map glacial and other sedimentary deposits in
shallow water. These techniques are potentially very useful for shallow storage
monitoring. They offer high-resolution profiling of seabed morphology but also have
sufficient depth penetration to identify anomalous features in the subsurface; for example,
62 DOE/NETL, 2010, Development and Deployment of a Compact Eye-safe Scanning Differential
Absorption Lidar (DIAL) for Spatial Mapping of Carbon Dioxide for MVA at Geologic Carbon
Sequestration Sites, Project 642, March 2010
http://www.netl.doe.gov/publications/factsheets/proiect/Proiect642.pdf
63 Cleaner Fossil Fuels Programme, 2005. Monitoring Technologies for the Geological Storage of C02.
Technology Status Report, DTI/Pub URN 05/1032, Didcot, Oxfordshire, UK, March 2005. Available at:
http://www.co2net.eu/public/reports/tsr025abc.pdf.
44
-------
zones of acoustic blanking interpreted as signifying high gas saturations. The techniques
can also permit direct imaging of gas plumes in seawater.
Several acoustic sonar bathymetric systems have been developed to provide area
mapping of sea bed bathymetry, resolving features as small as 1 cm, such as pockmarks
on the seafloor and free gas in the water column. Naturally occurring pockmarks and
shallow gas chimneys may act as potential pathways for CO2 leakage. Two principal
types are available: sidescan sonar using piezoelectric transducers mounted on towfish,64
and multibeam echo sounding using magneto-restrictive transducers from ships.
Multibeam echo sounders integrate bathymetric and backscatter information and can
provide a highly efficient way of obtaining very useful views of the sea floor. A multi-
beam echo sounder may be considered as a series of around 100 single-beam echo
sounders, mounted in a fan-shaped array on either side of a ship collecting echoes from
the entire swath width as the ship advances. The fan is narrow in the fore and aft
direction, typically 1.5 degrees, and wide in the port and starboard direction—around 120
degrees. For each beam, the system reports depth and the echo maximum amplitude over
time. This allows the detailed mapping of the seafloor bathymetry and inferences about
the nature of the sediment.
Electrical and EM methods offer the theoretical potential for low-resolution, low-cost,
site monitoring. The techniques utilize the propagation of electrical or EM fields within
the earth to map subsurface variations in electrical conductivity. A distinction can be
made between electric methods that use zero frequency (DC) or very-low-frequency
methods, where no EM induction occurs, and EM techniques that use frequencies where
time-variant source fields induce secondary electrical and magnetic fields that carry
information about subsurface electrical structure. CO2 is resistive, so electrical/EM
methods are likely to be suitable for monitoring storage in saline formations where CO2 is
displacing more conductive formation waters. Placing the electrical sources or receivers
(or both) in the subsurface, within or around the storage reservoir, radically increases
spatial resolution while reducing power requirements. Downhole configurations seem to
offer the greatest potential for useful offshore application.
Recent developments of sea floor EM systems, previously used (typically) for deep-water
tectonic studies, have led to "direct-detection-of-hydrocarbons" systems. Two of the most
advanced systems are referred to as SeaBed Logging65 and Offshore Hydrocarbon
Mapping66. The systems consist of towed Controlled Source transmitters and static sea-
bed receiver arrays. The source is a horizontal electric dipole, towed behind an
64 A towfish is an underwater equipment package towed a few hundred feet above the seafloor using an
umbilical cable connected to a support craft.
65 Ridyard, D. 2007. Seabed Logging - A Proven Tool for Offshore Exploration and Development.
Summarized from a Presentation given to AAPG European Region. Search and Discovery Article #40277
(2008) Posted January 31, 2008. Available at:
http://www.searchanddiscoverv.net/documents/2008/08008ridvard/images/ridvard.pdf.
66 Greer, A. MacGregor, L. Weaver, R. 2003, Remote mapping of hydrocarbon extent using marine Active
Source EM Sounding EAGE 65th Conference & Exhibition — Stavanger, Norway, 2-5 June 2003.
45
-------
instrumented towfish on a neutral buoyant streamer. During a survey the source is towed
around and through the array of receivers that each measure 2 or 3 orthogonal
components of the electric field. The technique is sensitive to thin resistive anomalies at
depths between several tens of meters and several kilometers. Recent surveys have
successfully determined the presence and absence of hydrocarbons within reservoirs.
Direct detection of resistive C02 zones within more conductive lithologies would,
theoretically, be possible.
Technologies to monitor surface seawater geochemistry, coupled to hydrographic and
meteorological conditions advanced significantly in the 1990s in an effort to understand
the role oceans play in global carbon budgets and the relative size and directions of
carbon (mainly C02) fluxes between the sea, the atmosphere and the sediment. These
technologies are relatively well tested and are deployed extensively across the world,
often routinely on ships regularly crossing oceans. A sensitive pH sensor for marine
studies has recently been developed in the United Kingdom and would be useful to
monitor changes in seawater pH resulting from C02 leaks. Four parameters are typically
measured that, together with ancillary information such as temperature, pH, and salinity
etc., can be used to describe the C02 system in a given sample. These parameters are:
total DIC (a measure of the concentrations of C02 (aq), H2CC>3 (aq), HCO3" and CO32"),
total alkalinity (a form of mass-conservation relationship for the hydrogen ion), fugacity
of C02 in equilibrium with seawater (a measure of the partial pressure of C02), and total
hydrogen ion, primarily controlled by total sulfate concentration.
Seawater samples obtained at depth, which are more appropriate to the needs of
monitoring for C02 escape, require the pressure to be maintained in order to ensure that
degassing does not occur. One such gas-tight sampling device has been developed to
obtain fluids samples (both waters and gases) at elevated temperatures and pressures. A
key feature is that fluid within the chamber is maintained at seafloor pressures during
sampling, allowing fluid aliquots to be withdrawn for an almost unlimited number of
analytical techniques without degassing the remaining fluid.
4.7 Detection Range, Accuracy and Precision of Various Monitoring Technologies
Each of the monitoring methods discussed above involves the application of technologies
to detect and quantify C02 leakage. Monitoring is effective in detecting the presence of
both subsurface and atmospheric leakage. Methods of surface leak quantification
continue to be an active research area.
The ability to detect or quantify a plume or leak is dependent upon factors such as the
rate of leakage, its depth and location, the atmospheric background C02 variability, the
number and distribution of sampling stations, and the limits of the technology itself. Each
technology typically has strengths and weaknesses, and only a combination of approaches
is expected to ensure reliable monitoring and leak detection.
To evaluate the capabilities of monitoring technologies, one must look at the theoretical
limits of C02 detection, the size of anomaly that can be detected, and the precision and
46
-------
accuracy of the method. Table 4-3 presents this information for a number of technologies
that can potentially be used for leak quantification. 67'68' 69'70 The terminology in the
table is as follows:
• Sampling rate. The frequency of sampling.
• Detection range. The instrument detection limit or a range of instrument detection
limits in appropriate units for the instrument.
• Anomaly detection threshold. The size of the C02 anomaly that can be detected
under typical background conditions.
• Precision. The degree of mutual agreement among a series of individual
measurements.
• Accuracy. The degree of conformity of a measured quantity to its actual value.
Table 4-3: Detection Range, Accuracy, and Precision of Various Monitoring Technologies
Technology
Reference
Sampling
rate
Detection
Range
Anomaly
Detection
Threshold
Precision
Accuracy
Comments
Pressure
gauges in
monitoring
wells
1,4
Continuous
NA
0.1 psi
change
NA
NA
Established
and reliable.
3-D seismic
1,4
Periodic
surveys
NA
1,000 to
10,000
metric ton
accumulatio
ns
NA
+/-20% of
plume vol.
Established
technology;
expensive,
resolution
decreases
with depth.
Cannot
image
dissolved
C02.
Eddy
Covariance
1, 2, 4
Continuous
>10
microgram
/ m2/ sec
>45
microgram/s
q. m/sec
5 - 30% of
the C02 flux
NA
Commerciall
y available
technology
used in other
fields; signal
to noise
challenges
co2
Detectors
(closed path
and open
path)
2
1 -10 Hz
during
sampling
0 - 3,000
ppmv
NA
0.2 ppm at
350 ppm
C02
concentratio
n
NA
Established
technology;
Portable and
inexpensive.
Accumulatio
n Chamber
2, 3, 4
Continuous
>0.46
microgram
si m2/ sec
NA
10% of the
C02 flux
12.5% of
the C02
flux
Commerciall
y available
technology;
67 Benson, S.M. 2006. Monitoring Carbon Dioxide Sequestration in Deep Geological Formations for
Inventory Verification and Carbon Credits. Society of Petroleum Engineers Paper 102833, 14pp. Available
at: http://www.energy.utah.gov/Utah Actions/documents/dec2006/spel02833.pdf.
68 Oldenburg, C.M., J.L. Lewicki, and R.P. Hepple. 2003. Near-Surface Monitoring Strategies for Geologic
Carbon Dioxide Storage Verification. Report #LBNL 54089, Lawrence Berkeley National Laboratory,
Berkeley, CA, October 30, 2003. Available at: http://www.osti.gov/bridge/servlets/purl/840984-
dTw752/native/840984.pdf.
69IPCC. 2006. 2006 IPCC Guidelines for Greenhouse Gas Inventories. Intergovernmental Panel on Climate
Change, Simon Eggleston, ed. Available at http://www.ipcc-nggip.iges.or.ip/public/2006gl/index.html.
70 DOE. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory. Available at:
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
47
-------
Many
sampling
stations
needed;
average data
from multiple
chambers;
samples
small area;
signal to
noise
challenge.
Raman
LIDAR and
DIAL
2, 4
Continuous
or periodic
<1 ppmv to
several %
C02
NA
1 -5% of the
C02
concentratio
n
NA
Technology
available for
methane;
C02
technology
under
development
; Very large
sampling
area; long
range;
Precision of
3-27 ppm at
1 km path
length
References:
1. Benson, S.M. 2006. Monitoring Carbon Dioxide Sequestration in Deep Geological Formations for
Inventory Verification and Carbon Credits. Society of Petroleum Engineers Paper 102833, 14pp. Available
at: http://www.energv.utah.gov/Utah Actions/documents/dec2006/spel02833.pdf.
2. Oldenburg, C.M., J.L. Lewicki, and R.P. Hepple. 2003. Near-Surface Monitoring Strategies for Geologic
Carbon Dioxide Storage Verification. Report #LBNL 54089, Lawrence Berkeley National Laboratory,
Berkeley, CA, October 30, 2003. Available at: http://www.osti.gov/bridge/servlets/purl/840984-
dTw752/native/840984.pdf.
3. IPCC. 2006. 2006 IPCC Guidelines for Greenhouse Gas Inventories. Intergovernmental Panel on
Climate Change, Simon Eggleston, ed. Available at: http://www.ipcc-
nggip.iges.or.ip/public/2006gl/index.html.
4. DOE. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory. Available at:
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
48
-------
5. Monitoring, Reporting, and Verification Plans
Facilities conducting GS are required to develop and implement an EPA-approved MRV
plan that includes the following five major components (which are identified at 40 CFR
98.448):
• Delineation of the maximum monitoring area (MMA), and active monitoring
areas (AMAs);
• Identification of the potential surface leakage pathways and an assessment of
the likelihood, magnitude, and timing of surface leakage of CO2 through these
pathways;
• Strategy for detection and quantification of surface leakage;
• Approach for establishing the expected baselines; and
• Considerations made to calculate site-specific variables for the mass balance
equation.
Chapters 5.1 through 5.5 describe these MRV plan requirements and present
considerations for developing site-specific MRV plans. For each requirement, the chapter
lays out key information that may address the requirement, and provides illustrative
examples and metrics to help the reporter determine the level of performance necessary
to design and implement a system that meets the regulatory requirements. Examples are
provided throughout the chapter to illustrate how these requirements may be addressed
under different scenarios.
In the MRV plan, the reporter should show that the approach taken to address each MRV
plan component provides an adequate level of assurance that the regulatory requirement
will be met. This assurance could include a demonstration that the statistical basis of the
monitoring strategy or the quantification method meets a certain level of performance. In
some cases the MRV plan may not include specific details of implementation because
there are many site-specific or event-specific factors that will influence the selection of a
particular method or technology. For example, prior to writing the plan the reporter will
not know the exact location or cause of a future surface leak, and therefore should
provide a description of the process and tools used to detect and quantify the potential
leaks. The reporter should provide the process and decision rationale for how such event-
specific decisions will be made.
EPA is taking this site-specific flexible approach for three reasons. First, each facility
will have a unique set of geologic, environmental, and operational conditions that are best
addressed with site-specific solutions to satisfy each MRV requirement. Second, as
projects mature, reporters will collect new information and may choose to improve their
conceptual site models and modify their monitoring, modeling, and evaluation
techniques. Third, EPA recognizes that the uncertainties and inherent variability in the
natural systems will necessitate modifications to the selected methods and approaches
over time and in response to unexpected events.
49
-------
The site-specific approach also allows the reporter to leverage the site characterization,
modeling, and monitoring approaches (e.g. monitoring of injection pressures, injection
well integrity, groundwater quality and geochemistry, and CO2 plume location, etc.)
developed for their UIC permit. The UIC Class VI permit (including the Testing and
Monitoring plan) and subpart RR MRV plan have separate monitoring objectives.
Requirements under the UIC program are focused on demonstrating that underground
sources of drinking water (USDWs) are not endangered as a result of CO2 injection into
the subsurface, while requirements under the GHG Reporting Program through subpart
RR will enable reporters to quantify the amount of CO2 that is geologically sequestered.
However, monitoring approaches and technologies employed for these two objectives
may overlap. The MRV plan submitted for subpart RR may describe (or provide by
reference to the UIC permit) the relevant elements of the UIC permit (e.g. assessment of
leakage pathways in the monitoring area) and how those elements satisfy the subpart RR
requirements.
Beyond describing how the existing or planned UIC monitoring procedures form the
basis for fulfilling the subpart RR leakage detection requirements, the subpart RR MRV
plan should show how any detected anomalies would be further studied to verify and then
quantify leakage. Leakage verification and quantification could include more frequent
monitoring, denser spatial coverage, or the deployment of additional monitoring
technologies.
For more information on UIC regulations, including requirements and guidance
documents, please go to http://water.epa. gov/type/groundwater/uic/index.cfm.
5.1 Delineation of the MMA and AMA
The MRV plan must include a delineation of the area that will be monitored. The
maximum areal extent of the plume area of the life of the project should be determined
using a reservoir simulator model informed by site characterization data and monitoring
results. Reservoir modeling or simulation is a powerful mathematical tool that is used to
evaluate the movement of injected CO2 in the reservoir, to predict the size and location of
the plume, and is an integral aspect of project design, planning, site characterization, and
monitoring program design.71 An initial model can be used to forecast how the plume is
expected to move and change. After the beginning of injection, the data from the
injection well and data from the other types of monitoring should be used to calibrate and
history-match the model.
Reservoir modeling has been in use for decades in the oil and gas industry, and is used to
design the development of fields, and is applied at GS sites. The model is a 3-D construct
of the subsurface reservoir, and uses a numerical approach in which the reservoir is
divided into a large number of discrete 3-D volumes for analysis. Input data include
reservoir thickness, depth, pressure, porosity, permeability, fluid saturations, and other
71 DOE. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory.
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
50
-------
parameters. Each 3-D volume has its own specific characteristics. In the case of GS, the
model is set up using all available data, and the model is run to estimate future movement
of CO2 through the reservoir. The model is then calibrated and updated through time as
the injection proceeds. This is accomplished through hi story-matching of actual
subsurface data, such as pressure or the detection of CO2.
There are three general types of simulation models used for GS: multiphase fluid flow
models, reactive transport models, and geomechanical models.72 The simulators allow the
prediction of the following:
• Temporal and spatial migration of the plume;
• Geochemical reactions;
• CZ and wellbore integrity;
• Potential leakage pathways and estimates of travel times for these routes;
• Effects of unplanned hydraulic fracturing;
• Potential leakage near the injector well; and
• Consequences of wellbore failure.
Like all numerical models, reservoir simulation models for GS are not an exact
representation of the subsurface reality, but provide an approximation of the conditions
throughout the reservoir based on the given input values and the computational codes
used in the simulation. Some uncertainty is inherent in the model output as a result of
uncertainties related to the construction of the underlying governing equations, and
uncertainties in the values used to represent the actual site conditions73. The accuracy of
the model will improve as model input parameters more accurately reflect the actual
subsurface conditions within the modeled area. It is important, therefore, for reporters to
evaluate the accuracy of the model, and identify the input parameters which are most
sensitive to the model output. Developing accurate input values for the most sensitive
parameters will improve the models predictive accuracy. The model input parameters
may need to be updated often to reflect new information gained from the operation of the
GS project.
A recent evaluation of advances in numerical modeling by Michael et al.74'75 notes that
significant improvements have been made in the numerical models particularly in the
linking of geochemical, geomechanical, and flow models to predict reservoir
performance. The authors also note, however, that relative permeability and residual CO2
saturation are sensitive input parameters for accurate reservoir models and need to be
better constrained with data from current projects. The evaluation also notes that there is
72 DOE. 2009. Best Practices for Monitoring, Verification, and Accounting for C02 Stored in Deep
Geologic Formations. U.S. Department of Energy, National Energy Technology Laboratory.
http://www.netl.doe.gov/technologies/carbon sea/refshelf/MVA Document.pdf.
73 US EPA, 2003. Draft Guidance on the Development, Evaluation, and Application of Regulatory
Environmental Models. Prepared by The Council for Regulatory Environmental Modeling.
74
Michael, K., Arnot, M., Cook, P., Emus-King, J., Funnell, R., Kaldi, J.G., Kirste, D., Paterson, L., 2009.
C02 storage in saline aquifers I—current state of scientific knowledge. Energy Procedia 1, 3197-3204.
75 Michael, K., Golab, A. Shulakova, V., Ennis-King, J., Allinson, G., Sharma, S., and Aiken, T.2010.
Geological storage of C02 in saline aquifers - A review of the experience from existing storage operations
51
-------
limited data from post injection monitoring to calibrate models for the long term
assessment of reservoir conditions.
As described below, subpart RR requires that two monitoring areas be delineated: the
MM A and the AM As.
5.1.1 Maximum Monitoring Area (MMA)
Under 40 CFR 98.448(a)(1) the reporter will need to establish the maximum areal extent
of the injected C02 over the life of the project, including the post-injection period, to
define boundaries for monitoring leakage. This boundary is used to help define the
MMA. The MMA is defined as equal to or greater than the area expected to contain the
free-phase CO2 plume until the CO2 plume has stabilized, plus an all-around buffer zone
of at least one-half mile. The buffer is intended to encompass leaks that might migrate
laterally as they move towards the surface. EPA has determined that a buffer zone of at
least one-half mile will have an acceptable probability of encountering leaks in many
circumstances. In some cases the use of a one-half mile buffer zone may not account for
uncertainties in the subsurface conditions, or may not incorporate potential leakage risks
from faults and fractures that extend into the edge of the modeled MMA, and the buffer
zone should be expanded.
For example, if there is a specific leakage pathway or a pronounced regional dip that
might carry a leak further, the buffer zone should be expanded beyond one-half mile. The
reporter should derive the MMA from site characterization, monitoring, and models. The
MRV plan should describe the rationale for how the MMA was determined, including a
discussion of the uncertainties of the models and buffer zones.
In order to determine the MMA, the reporter should estimate, by modeling, the future
area of the free-phase plume. The geometry of the free-phase plume will be a function of
the amount and rate of CO2 injected, as well as the geologic characteristics of the IZ
including the CZ geometry, IZ thickness, permeability, and porosity, and the amount of
anisotropy within the IZ. The resolution of the reservoir model used to predict plume
behavior will also influence the delineation of the area of free-phase CO2. A model that
can predict characteristics of thin layers at the upper boundary of the IZ, or predict the
presence of lower CO2 saturations, may be able to resolve a larger area of free-phase
CO2. The reporter should describe the rationale for defining the free-phase plume
boundary by presenting the results of the reservoir simulation including the minimum
CO2 saturation that defines free-phase, and the thickness of the zone over which the
saturation is estimated. The determination of both minimum saturation and saturated
thickness help define the plume edge.
Plume stabilization is the basis of the free phase plume boundary and depends on the rate
of movement of the free-phase CO2 and the moderation of pressures within the free-phase
plume. The reporter should define what criteria will be used to determine when the free-
phase plume is to be considered stable. For example, this could be stated in terms of
when the rate of movement of free phase CO2 is less than a certain value (X foot per
52
-------
year), in any direction, greater than the natural (or not influenced by the site)
hydrodynamic movement of the IZ, and the pressure change within the reservoir is less
than a certain value (Y psi per year). The values of fluid movement and pressure would
be generated from runs of the reservoir model. The values that define plume stability
should be consistent with the proposed monitoring and modeling methods.
The reporter should redefine the MMA to reflect changes to the plume area if
observations of the plume extent are significantly different from the modeled or expected
extent and should also resubmit the MRV plan (see Chapter 6.7).
5.1.2 Active Monitoring Area (AMA)
It may not be practical or cost-effective to implement a full-scale monitoring program for
the entire MMA when in the early years of operation the free-phase plume covers only a
small portion of the MMA. The deployment of monitoring locations may be phased in
over time to include those areas where the potential leakage may occur. The AMA is the
part of the project area that will employ the monitoring methods and systems described in
the MRV plan for a period of time less than the life of the project but greater than a year.
Over the life of the project there will likely be several monitoring phases, each with an
AMA that will increase in size as the plume expands. The MRV plan should describe the
area and duration of each AMA as defined by the reporter, and should be determined
using the most recent monitoring and characterization data available. The MRV plan
should include a description of how the AMA was determined.
To ensure that the proposed leak detection monitoring systems provide adequate
coverage, the AMA must extend beyond the modeled plume position at the end of the
monitoring period. The AMA is established by superimposing two areas: the first is a
one-half mile buffer zone around the outline of the anticipated plume location at the end
of the AMA period, and the second is the area projected to contain the free-phase CO2
plume five years after the end of the AMA. The area encompassed by either or both of
the two areas will represent the AMA.
Subpart RR requires that the AMA be larger than the projected free-phase plume, for two
reasons. First, there may be uncertainty in the projected plume location, given
uncertainties in the characteristics of the IZ. The biggest element of uncertainty may be
the "storage efficiency" of the IZ, or what portion of the pore space will be filled with
CO2 versus the portion that still will contain naturally occurring saline fluids. This
uncertainty will be greatest at the start of injection and will be reduced over time as the
reservoir simulation models are calibrated through history-matching of observed plume
migration. The initial AMA must encompass the area that the free-phase plume is
expected to occupy five years beyond the end of the initial monitoring period, so as to
account for this IZ-related uncertainty in reservoir characteristics.
The second reason why the AMA needs be larger than the expected free-phase plume
area is that any CO2 leaks from the IZ will not necessarily follow a straight path upward
to the surface. The flow path of leaking CO2 will be influenced by buoyancy forces that
53
-------
move the CO2 laterally in an up-dip direction, via hydrogeologic pressure differences.
The leaked C02 will tend to follow formation fluid movements, or geologic features such
as reservoir heterogeneities or faults that may produce circuitous upward leakage
pathways. Therefore, a monitoring buffer around the free-phase plume needs to be
applied. To account for sites where lateral movement of the plume will be controlled by
structural features (and where the free-phase plume five years past the end of the active
monitoring period will be nearly the same location as at the end of the long-term
monitoring period), EPA has determined that a monitoring buffer of at least one-half mile
will have an acceptable probability of encountering leaks in many circumstances. At sites
where there are known geologic features that may carry leaks laterally more than one-half
mile beyond the free phase plume, then that monitoring buffer zone should be extended
to encompass those specific leakage pathways.
C02. PJume
.(Free Gas)
Free Gas Butler
Zor»
inil&nrm flfoa (I
unitary defined
RR Maximum
(MMA)
try
presence of Free Phaae
CO, unl.ll the plume has
Ma bill ted, plus a one-hair
mile all around buffer
j * Acbve
Monitanrw} A/ea
= CQ7 miction
Q = AtwKJon-wfWtf
Subpart KR Active
Moretorinq Arw (AM A)
boundary defined by the
area that will be
monitored over a specific
tim# Interval
Figure 5-1: Monitoring Areas as They Relate to UIC Class VI and subpart RR
Requirements
5.2 Assessment of the Risk of Potential Leakage of CO2 to the Surface
Reporters are required to evaluate the risk for leakage of CO2 through all potential
pathways within the monitoring area described in Chapter 5.1. As discussed in Chapter
5.1, leakage pathways outside the monitoring area should be evaluated if the pathway has
the potential to carry a leak outside the monitoring area. Once a potential leakage risk
pathway is identified, it should be evaluated to determine the likelihood, magnitude, and
timing of potential leakage. Each leakage pathway should be numerically identified and
referenced in the MRV plan. The evaluation of leakage pathways will likely be
UIC Class VI Area
of Review
fautl/Praclijra
UIC Class VI Area at
Review boundary defined
by tlio region surrounding
Hie geologic ssgufretialion
project where USDWs
may be endangered by
\the Injection activity
54
-------
qualitative in nature. If a reporter has sufficient information to develop a quantitative
analysis of the likelihood, magnitude, or timing of leakage risk, it may be presented in the
MRV plan. One example of a quantitative analysis of leakage risk is presented in the Risk
Assessment appendix to the Final Environmental Impact Statement for the FutureGen
project.76 Another method of quantifying leakage risk was developed by Oldenburg,
Bryant and Nicot who applied their methodology to a GS site in the Gulf Coast. 77
One or a combination of physical phases of C02 that have varying chemical and physical
properties could be present in leakage from a GS system. These phases of CO2 include:
• Dry supercritical CO2;
• The wetted portion of the supercritical C02 plume;
• Native fluids in the IZ in which high concentrations of CO2 are dissolved; and
• Fluids above the confining zone (ACZ) that mix with C02-containing fluids
migrating from the GS system.
Fluids in these phases may be present in the IZ or adjacent geologic layers, and could
come into contact with the potential subsurface leakage pathways. Escape of any of these
C02-containing substances to the surface could constitute a reportable release requiring
quantification by the reporter. While the nature of individual GS systems can be expected
to differ with respect to site-specific geologic attributes, potential release pathways at GS
sites could be:
. Wells;
• Fractures, faults, and bedding plane partings;
• Competency, extent, and dip78 of the confining system.79
Please refer to relevant UIC regulations80 for information on suitable sites for GS, and the
Vulnerability Evaluation Framework81 for further background information on potential
leakage pathways.
5.2.1 Assessment of the Risk of Potential Leakage of CO2 through Wells
Of the potential conduits for CO2 leakage, abandoned wells are the most likely conduits
for leakage of CO2 from the IZ.82 The presence of oil, gas, and water wells completed
76 U.S. Department of Energy (DOE). 2007. FutureGen Environmental Impact Statement.
77 Oldenburg, C. Bryant, S., and Nicot, J., 2009. Certification Framework Based on Effective Trapping for
Geologic Carbon Sequestration. International Journal of Greenhouse Gas Control
Volume 3. Issue 4. July 2009, Pages 444-457
78 Dip is the angle at which the rock unit is included from horizontal.
79 Leakage related to the CZ can occur if the CZ allows C02 to pass though it via permeable zones or
fractures, or if the CZ is not continuous across the plume area and C02 escapes around the CZ. Leakage
may also occur if the CZs is at an angle and C02 moves up the incline to the surface or to other permeable
zone.
80 http://water.epa.gov/type/groundwater/uic/wells_sequestration.cfm
81
US EPA. 2008. Vulnerability Evaluation Framework for Geologic Sequestration of Carbon Dioxide.
U.S. Environmental Protection Agency Technical Support Document EPA430-R-08-009, July 10, 2008.
55
-------
and/or abandoned over the last century poses a risk for CO2 leakage due to the potential
for casing and cement failure, as well as the potential for incompetent bonding between
cements and either well casing and/or the confining formation.83 Abandoned oil and gas
wellbores and exploratory boreholes may be present at GS sites located in depleted or
active oil and gas fields, or in formations near current or historic petroleum exploration or
production.
Well plugging and abandonment methods have evolved over time from makeshift
plugging with available materials to a regulated, documented process with quality control
measures. Gasda et al.83 describe the possibility of improperly cemented and abandoned
wellbores to function as conduits for CO2 due to a variety of mechanisms, including:
• Degradation of cement and piping by corrosive fluids from the GS system;
• C02 flow through degraded cements;
• CO2 flow along the borehole wall;
• C02 flow along the interface of the casing and cement; and
• CO2 flow through holes in the casing or annular seals.
As a depleted oil and gas reservoir is repressurized by GS activities, potentially to
pressures exceeding those present at the start of the field's development, risks posed by
weak well seals can be expected to increase significantly. Newer wells constructed
specifically for service in CO2 environments are less likely to show cement and casing
deterioration than older oil and gas production wells. Recent studies of cement cores from
a 52 year old well exposed to CO2 ER in the Permian Basin for 30 years84, and a 30 year
old C02 production well in Colorado85 showed effective sealing ability of the cement.
These studies concluded that although the cement reacted with the CO2 it was not
compromised possibly due to re-mineralization of the cement into impermeable
materials.
Although relatively rare, blowouts from operating oil and gas wells can occur as a result
of injection. A study of the rate of oil well blowouts in California's southern San Joaquin
Valley between 1991 and 2005, indicated that approximately one-third of the blowouts
occurred at active producing wells, most of which were in fields undergoing thermally
82
Gasda, S.E., S. Bachu, and M.A. Celia. 2004. The potential for C02 leakage from storage sites in
geological media: Analysis of well distribution in mature sedimentary basins. Environmental Geology
46(6-7) :707-720.
83 Gasda, S.E., S. Bachu, and M.A. Celia. 2004. Spatial Characterization of the Location of Potentially
Leaky Wells Penetrating a Deep Saline Aquifer in a Mature Sedimentary Basin. Environmental Geology
46, 707-720.
84 Carey, J.W., M. Wigand, S.J. Chipera, G. WoldeGabriel, R. Pawar, P.C. Lichtner, S.C. WehnerM.A.
Raines and G.D. Gutherie Jr., "Analysis and Performance of Oil Well Cement with 30 Years of C02
Exposure from the SACROC Unit, West Texas, USA", preprint submitted to International Journal of
Greenhouse Gas Control, October 18, 2006
85
W. Crow, B. Williams, D. Carey, J.W. Celia, S. Gasda, "Wellbore integrity analysis of a natural C02.
producer, Energy Procedia, Vol 1 (1), Proceedings GHGT- 9, Washington, D.C., (2008).
56
-------
enhanced oil recovery86. However, the study also showed that the number of blowouts
per year declined by 80 percent, which the authors attribute to improved industry
operational and safety practices.
As part of an MRV plan, the reporter should identify the risk of leakage to the surface
through abandoned wells by considering the location, depth, type and age of well, and
reviewing the available documentation to evaluate the quality of construction and
effectiveness of the closure methods. The associated risks should be described in terms of
the relative likelihood, magnitude, and timing for surface leakage. The rationale for
estimating likelihood, magnitude and timing of each pathway should be presented.
Injection and monitoring wells of the GS project are potential leakage pathways but are
also the conduits over which the facility will have the highest level of control. The UIC
program regulates injection well construction and integrity testing. Please refer to the
UIC program Web site for information on requirements related to monitoring of
permitted UIC wells, including MIT.87 Injection and monitoring well construction,
evaluation, and leakage risk is also an active area of research.88'89'90
Injection wells pose a potential CO2 leakage risk because they will be subjected to high
pressures and flows of C02. Monitoring wells in the IZ may also pose a risk for potential
CO2 leakage because they will be within the area of increased pressure from the injection
wells, and could be in contact with native fluids containing C02 which may affect
cement integrity. Additionally, nearby monitoring wells within the IZ will experience
extended periods of exposure to corrosive mixtures of resident fluids and C02 as the C02
plume migrates outward from the injection wells. The combination of potential corrosion
and high pressures will increase risks associated with monitoring wells in the IZ.
Injection and monitoring wells constructed of materials that can withstand the pressures
and the corrosive environments can minimize risk of an accidental release. Mechanical
integrity testing of cement, cement bonds, casing, tubing, packers, valves, and piping can
ensure that injection and monitoring wells are capable of withstanding the pressures and
geochemical conditions anticipated to occur during GS injection and post-injection
period.
86 P. Jordan and S. M. Benson, 2009. Well blowout rates in California Oil and Gas District 4: update and
trends. Exploration and Production: Oil and Gas Review.
87 http://water.epa.gov/type/groundwater/uic/wells_sequestration.cfm.
88 Chabora, Ethan and Benson, S. 2008. Monitoring Pressure Transients in Zones Overlying C02 Storage
Reservoirs as a Means of Leak Detection and Diagnosis. Poster presented at 7th Annual Conference on
Carbon Capture and Sequestration - Pittsburgh, PA, May 7, 2008.
89Nordbotten, J.M., D. Kavetski, M.A. Celia, and S. Bachu. 2009. Model for C02 leakage including
multiple geological layers and multiple leaky wells. Environmental Science and Technology 43: 743-749.
90 Celia, Michael A., Bachu, S., Nordbotten, J., Gasda, S., Dahle, H. 2004. Quantitative Estimation of C02
Leakage from Geological Storage: Analytical Models, Numerical Models and Data Needs. Proceedings of
the 7th International Conference on Greenhouse Gas Control Technologies, September 5-9 2004,
Vancouver, B.C.
57
-------
The MRV plan should describe and reference construction and MIT methods to be used
for GS system injection and monitoring wells,91 and include an analysis of the likelihood
and magnitude of releases from wells.
5.2.2 Assessment of the Risk of Potential Leakage of CO2 through Fractures,
Faults, and Bedding Plane Partings
Fractures, faults, and partings along bedding planes are potential leakage pathways in a
GS system. Fractures are breaks in the rock caused by compression or stress, which result
in areas of weakness in the rock that can also serve as conduits for fluid movement.
Individual fractures are often part of a larger group of similarly oriented fractures
resulting from regional and local stresses. Faults are features in the earth at which the
rock has broken and along which movement has occurred. Depending on the properties of
the strata in which a fault occurs, nature of the movement, and subsequent mineralization
or dissolution, faults may function as conduits for fluids moving through the subsurface
or as barriers to fluid movement. Bedding plane partings are naturally occurring areas of
weakness, often found between poorly bonded sedimentary layers, which may be
laterally extensive and intersect vertical migration pathways. Fractures, faults, and
bedding plane partings can be forced open by fluids whose pressures exceed the
fracture/fault reactivation pressure.
These features may not be observed at the surface or be apparent during geologic site
characterization, due to overburden or insufficient offset to be observed by standard
geophysical imaging. Additionally, due to the significantly lower viscosity of
supercritical CO2 relative to water, these features may not be observed during site
characterization hydraulic testing of the CZ and the ACZ. Risk presented by these
features exists due to the ability of CO2 to move along open fractures and faults, and the
ability of pressurized fluids to open faults,92 fractures, and bedding plane partings.93
These potential conduits can be identified from core samples and various geophysical and
hydraulic testing methods. Once identified, these features should be monitored for early
detection of leaks, in accordance with the approved MRV plan. Please note that for GS
wells permitted by the UIC Class VI program, there are extensive site characterization
and suitability requirements, including that the CZ must be free of transmissive faults and
fractures.
91 Please see http://water.epa. gov/tvpe/groundwater/uic/wells sequestration.cfm for UIC program MIT
requirements and guidance documents. Additional guidance is provided at the EPA Region V UIC Web
site: http://www.epa.gov/r5water/uic/techdocs.htm.
92 Evans, D.J. 2007. An appraisal of Underground Gas Storage Technologies and Incidents, for the
Development of Risk Assessment Methodology. British Geological Survey. (OR/07/023), Nottingham, UK,
Available via http://nora.nerc.ac.uk/7877/.
93 Watney, Willard L., Byrnes, Alan P. Bhattacharya, Saibal, Nissen, Susan E., and Anderson, Allyson K.
2003. Natural Gas Explosions in Hutchison, Kansas: Geologic Factors. Geological Society of America,
North-Central Section - 37th Annual Meeting, Kansas City, Missouri, March 24-25. Updated presentation
available from: http://www.kgs.ku.edu/Hvdro/Hutch/GSA Watnev updated082004.pdf
58
-------
The possibility of induced seismicity from fluid injection94 compounds the risks
associated with the above leakage pathways. Seismic events (e.g. small earthquakes)
resulting from overpressure in the IZ, CZ, or well materials can cause fracturing and
create preferential pathways for C02 movement. While induced seismicity may occur
along pre-existing faults identified during site characterization, it may also occur along
features that were not detected prior to injection, and be activated as a result of increased
pore pressure from injection.
The MRV plan should describe the leakage risk from known and inferred faults and
fracture systems based on their extent, depth and historical data. The analysis of risk
should consider the possibility of lateral transmission of CO2 and splays in the fault and
fracture systems as well as offsets within the subsurface. The evaluation of faults and
fractures should encompass a large enough area to account for plume movement that is
different from the predicted reservoir simulation model.
Existing data sets may need to be supplemented and updated with measurements specific
to the needs of understanding CO2 leakage. For example, existing or new geophysical
survey data may need to be re-processed and re-interpreted to evaluate the extent of faults
and fractures, the competency of the CZ or the current location of the CO2 plume. In
areas where there is limited geological information from exploration wells, reporters may
need to increase the density of data sampling.
5.2.3 Assessment of the Risk of Potential Leakage of CO2 based on the
Competency, Extent, and Dip of the Confining Zone
The CZ overlying a GS system performs a critical function of preventing the upward
migration of highly buoyant supercritical CO2 into the ACZ. The risk of a release of CO2
into the ACZ and subsequently into the atmosphere is greatly dependent on the
competency, extent, and slope of this rock unit.
The presence of fractures, faults, interconnected bedding plane partings, susceptibility to
degradation by CO2 and impacted native fluids, or low capillary entry pressures95 could
render the CZ capable of transmitting CO2 from the IZ into the ACZ. These properties
should be evaluated during site characterization, and any potential conduits should be
monitored during baseline data collection and facility operation. In addition, anticipated
injection pressures should be compatible with the geomechanical properties of the CZ to
prevent fracturing and creation of conduits to the ACZ. Geochemical, hydraulic, and
geomechanical testing of the CZ materials should be performed as part of GS system
characterization and facility construction.96
94 Ingebritsen, S. E , Sanford, W. E, Neuzil, C. E. 2006. Groundwater in Geologic Processes. Cambridge
University Press, p. 29.
95 Capillary entry pressure is defined in the Vulnerability Evaluation Framework as the added pressure that
is needed across the interface of two immiscible fluid phases (e.g., supercritical C02 and water or brine) for
C02 to enter the confining system. If the IZ pressure exceeds the capillary entry pressure, C02 could be
forced out of the IZ into the CZ.
96
US EPA. 2008. Vulnerability Evaluation Framework for Geologic Sequestration of Carbon Dioxide.
U.S. Environmental Protection Agency Technical Support Document EPA430-R-08-009, July 10, 2008.
59
-------
The lateral extent of the CZ should be sufficient to contain the C02 plume as it spreads
and migrates laterally, during and after injection, until physical and geochemical trapping
mechanisms can sequester the C02. General CZ extent and thickness should be available
from previous oil and gas or ER studies or regional geologic studies used in the GS
system siting. These data can be supplemented by well logging performed during
injection and monitoring well drilling, records from oil and gas exploration and
production drilling, and geophysical studies. At sites where the CZ is not laterally
extensive, an increased risk of leakage around the edge of the CZ may be present.
The dip of the confining system is an important component of the GS system. Due to the
buoyancy of supercritical C02, until geochemical trapping mechanisms sequester the
C02, the injected fluid can be expected to move in an up-dip direction until a structural or
stratigraphic trap97 is encountered or the lateral extent of the confining unit is reached
and the plume moves upward into the ACZ. A relatively steep dip of the overlying CZ
may result in the need for a more laterally extensive confining system to prevent the
escape of CO2 to the ACZ and subsequently to the atmosphere. The risk of leakage at the
up-dip locations should be considered a potential leakage pathway for sites with steeply
dipping CZ without structural closure. GS system modeling studies including simulation
of multiphase fluid flow may be performed to determine expected travel times and
evaluate the potential for the confining system to retain supercritical, and subsequently
dissolved-phase, C02 until geochemical and structural trapping mechanisms can
effectively sequester the CO2. The likelihood, magnitude, and timing of the risk of
releases through the CZ should be assessed and presented in the MRV Plan.98 The MRV
plan should present an assessment of the risk of leakage from the CZ with supporting
information for how the risk was determined. The description should evaluate the
uncertainties in the risk evaluation and consider movements of the plume throughout the
lifecycle of the sequestration.
5.3 Strategy for Detecting and Quantifying any CO2 Leakage to the Surface
5.3.1 Detecting Leakage
After the potential leakage pathways in the MMA have been identified, the reporter will
design a strategy to assess the pathways for conditions that could indicate leakage from
the AMA within specified time intervals. The strategy should be designed so that
potential leakage pathways are monitored in a comprehensive manner that allows for
timely and accurate identification of leaks. Subpart RR requires quantification and
reporting of leakage to the surface; however, the reporter's leakage detection program
may rely on subsurface measurements to detect leaks (such as may be implemented for
compliance with a UIC permit). A leak may be initially detected in the subsurface in
97 A stratigraphic trap is caused by changes in the permeability or porosity of the reservoir rock that restrict
flow. An example of a stratigraphic trap would be a sand lens that is surrounded by silt or clay.
98 K. P. Saripalli, N. M. Mahasenan, and E. M. Cook. 2002. Risk and hazard assessment for projects
involving the geological sequestration of CO2. In Kaya et al. GHGT-6: Sixth International Conference on
Greenhouse gas control technologies, Kyoto, Japan, September 30-October 4 2002. Elsevier.
60
-------
which case subsequent verification and quantification analysis will be needed to
determine if the leak has reached or is likely to reach the surface (described in Chapter
5.3.2 and 5.3.3).
The strategy for detecting CO2 leakage to the surface could include taking measurements
on a continuous basis, such as pressure readings in injection and monitoring wells, or
continuously reading eddy covariance monitoring. The leakage detection strategy could
also include regularly scheduled periodic monitoring events and surveys designed to
evaluate conditions at a snapshot in time. Regularly scheduled monitoring events could
include periodic sampling of water chemistry, MIT of injection and monitoring wells, or
whole-area airborne surveys conducted at regular intervals. The area covered by the
leakage detection measurements may change over time as the plume expands, and the
AMA is expanded, or new wells are added, but the practices would likely remain the
same.
The detection capability of leakage detection monitoring systems should be described in
the MRV plan. Reporters should develop a robust monitoring system that ensures
adequate detection capability. In determining the detection capability, the reporter should
consider:
• The accuracy and precision of the instruments used to measure data that will be
quantified.
• The statistical variability of the expected baselines against which the
measurements will be compared.
• The time required for a leak to be transmitted to a monitoring station or
measurement point.
• The ability to use the monitoring program to reliably recognize leaks of the
targeted magnitude.
The identification and influence of site characteristics such as topography, land use,
weather, and climate on the detection methods should be presented in the MRV plan as
part of the assessment of adequacy for detection methods. The considerations for the
approach and techniques are described below.
Part of the suite of leakage detection monitoring should include MIT of injection wells,
using a suite of well logs that are capable of determining leakage under the site-specific
conditions. Such logs may include but are not limited to noise logs; cement bond/variable
density logs; mechanical, electrical, or acoustical caliper tools; tracer logs; and other
specialized well inspection tools. MIT utilized for satisfying UIC requirements could be
described as part of the leakage detection strategy in the MRV plan. Although there are
no UIC requirements for performing MIT on monitoring wells, it may be prudent for
reporters to include in their MRV plans a periodic evaluation of the integrity of
monitoring wells that penetrate the IZ. The reporter's process for selecting and
interpreting MIT should be provided in the MRV plan. Additional information on the use
of these and other monitoring techniques is provided in Chapter 4 of this TSD.
61
-------
The process and technologies for detecting leakage through CZ or spill points should be
presented in the MRV plan. In some cases, changes to the ACZ will likely be a result of
leakage through the CZ or structural spill points, which are detectable in the monitoring
wells and with other subsurface imaging technologies. For example, pressure changes
observed in multiple monitoring wells can be used to locate the general area of leakage
through triangulation. The process and technologies for detecting leakage through CZ or
spill points should be presented in the MRV plan. The reporter should determine the
targeted leak size and estimate the reliability of its detection for the proposed monitoring
technologies and procedures.
Faults, natural fracture systems, and abandoned boreholes are potential leakage pathways,
where a range of different monitoring technologies could detect leakage, depending on
site-specific conditions. Near-surface monitoring devices may include eddy covariance,
LIDAR, accumulation chambers, soil gas surveys, isotopic analysis of C02 at the surface
or in shallow wells, and monitoring for tracers injected with the CO2. Leakage from the
ground into surface water, such as a leaking fault that intersects a lake, may require
different approach to monitoring and flux calculation."
The reporter should consider the characteristics of the GS site and the capabilities of the
monitoring systems to analyze what leak sizes or surface C02 fluxes are detectable using
the proposed methodology. Subpart RR does not have a specific leak detection minimum
volume or reliability standard. Instead, it is requested that reporters compute the
statistical reliability of the monitoring element given one or more targeted leak rate(s).
Note that a GS project operating under a UIC permit will have separate UIC requirements
if a leak is detected that may include reporting of the leak, ceasing injection, and taking
corrective action.
The targeted leak rates for the probability calculations in the MRV plan should be
appropriate to the site, and for example could be on the order of thousands or tens of
thousands of metric tons per year as measured in the subsurface. There is little published
research on measured subsurface leakage rates for GS sites, so the reporter should
provide a justification for selection of the target rate. The reporter may wish to review the
risk assessment performed for the proposed FutureGen project100 which tabulated surface
emissions rates from 28 locations including natural CO2 sources (domes, volcanic, and
hydrothermal sites) and ER/GS sites. The surface emissions ranged over eight orders of
magnitude, but all of the ER and GS sites had surface flux of less than one micromole per
square meter per second. The reporter may also wish to review the design of studies
conducted at the Zero Emissions Research and Technology Center in Bozeman Montana
where controlled release experiments were performed to simulate low-level seepage in
order to test the detection capability of several monitoring methods101.
99 Oldenburg, C.M., and J.L. Lewicki. 2006. On leakage and seepage of C02 from geologic storage sites
into surface water. Environmental Geology, 50(5), 691-705 LBNL-59225
100 U.S. Department of Energy (DOE). 2007. FutureGen Environmental Impact Statement.
101 Spangler, L.H., Dobeck, L.M., Repasky, K.S., Nehrir, A.R., Humphries, S.D., Barr, J.L., Keith, C.J.,
Shaw, J.A., Rouse, J.H., Cunningham, A.B., Benson, S.M., Oldenburg, C.M., Lewicki, J.L., Wells, A.W.,
Diehl, J.R., Strazisar, B.R., Fessenden, J.E., Rahn, T.A., Amonette, J.E., Barr, J.L., Pickles, W.L.,
Jacobson, J.D., Silver, E.A., Male, E.J., Rauch, H.W., Gullickson, K.S., Trautz, R., Kharaka, Y.,
62
-------
For the evaluation of near-surface monitoring techniques, these leak rates should be
translated to surface leak areas and fluxes based on dispersion modeling or calculated
factors to be discussed below. It should be understood that these leak rate assumptions
are not intended as predictions of actual leaks or limits on what leaks can or should be
detectable. Rather, they are benchmarks against which the performance of the leak
detection strategy can be measured using statistically valid techniques. The performance
would generally be stated in terms of a leak rate target, a time period, and percent
probability of detection.
For example a plan to detect leaks with pressure monitoring using wells in the ACZ
might have a performance evaluation concluding that it can detect "leaks of X metric tons
of CO2 per year within Y days with a probability of Z percent." The performance
estimate does not mean that leaks smaller or larger than this cannot be detected. Smaller
leaks can be detected, but they may take more time or will be detectable with less
reliability; larger leaks may be detectable in less time with greater reliability.
A consideration in designing a leakage detection strategy (and in establishing expected
baselines) for the MRV plan is ensuring adequate data resolution to cover a range of
points in space and time. A non-representative data set that contains too few samples at
too few times of day/year from too few locations will generate a biased picture and result
in erroneous decision making. A sampling program should be designed to ensure that it
considers the appropriate frequency of sampling, aerial extent of sampling, and sample
size. The MRV plan is likely to include approaches that involve collecting data at one or
more locations within the MMA.102 For example, CO2 isotope ratios in soil gases might
be measured throughout the entire AMA as one leak detection method. The MRV plan
could present the statistical coverage adequacy of a given sampling program (i.e., the
probability a given size of leak will be detected). Please refer to Appendix E for further
discussion of this topic.
5.3.2 Verifying Leakage
The MRV plan should describe the strategy to verify and confirm the location and source
of leakage that has been detected. The leak verification process is intended to be a quick
and cost effective step to help the reporter focus efforts on deviations from expected
measurements that would need to be quantified. The MRV plan should describe the
methods and criteria for determining how an anomalous reading or condition will be
evaluated to determine if it represents a leak, for verifying the location and source of the
leak, and what level of accuracy is anticipated. This description should include details of
the approach for determining how readings will be distinguished from background or
Birkholzer, J., and Wielopolski, L.„ A shallow subsurface controlled release facility inBozeman, MT,
USA, for testing near surface C02 detection techniques and transport models. Environmental Earth
Sciences, 60, 227-239, doi: 10.1007/s 12665-009-0400-2, 2010
102 Jones, David. 2009. Ground surface monitoring of C02 and associated gases. Presented at C02 GeoNet
Open Forum, March 2009, Venice, Italy.
63
-------
operational variability (see Chapter 5.4 for a detailed presentation of this topic). If
possible, the MRV plan should provide criteria that will be used to identify the presence
of a leak. Monitoring systems for each pathway are likely to employ different
technologies at different measurement frequencies. For example, ACZ pressure
monitoring may be conducted continuously at several locations over a broad part of the
C02 plume however, leak detection at active or abandoned wells may be performed using
a truck-mounted laser scanning survey that is conducted once per year at specific well
locations within the AMA. Likewise, leak detection and verification methods will vary
with the monitoring technology. Since each monitoring system will be different, and the
monitoring systems may change location over time as the plume changes, the reporter
may simplify the presentation of the leak detection and verification process in the MRV
by describing the leak detection and verification in a flowchart or process diagram.
The verification steps may include:
1. Verification that the reading is accurate and truly representative of the condition.
This may require repeating the measurement, recalibrating the measurement
device and repeating the measurement, taking a second measurement with a
redundant device or a different device of greater accuracy or higher precision, or
evaluating other data and measurements that may be affected by the condition.
Verification of the reading may also necessitate collecting measurements for a
longer period of time to determine if the anomalous reading is of a short duration
or part of a longer or continuous trend of anomalous readings. For very large
leaks that are manifested at the surface, direct visual observation of ice crystals,
vapor plumes, or gas bubbles in well cellars may provide adequate verification of
a large leak.
2. Identification of the location or the source of a suspected leak. Once the accuracy
of the measurement is confirmed, other methods may be used to further refine the
location of a leak, including triggered or episodic monitoring that takes place after
a leak is suspected. For example, if a release of CO2 from a fault zone is suspected
based on fixed location eddy covariance data, a soil gas survey of the fault area
could be performed to more accurately define the location of the leak. Resolution
of the source can also be made with surface measurements by using a denser
sampling grid or more frequent time scale to improve the resolution of the source
area. For subsurface leakage in the injection or monitoring wells, mechanical
integrity well logs can be used to help define the area and pathway of leakage.
5.3.3 Quantifying Leakage
The MRV plan must include a discussion of how leaks will be quantified once they are
detected and verified.103 Given the uncertainty concerning the nature and characteristics
of leaks that will be encountered, EPA expects that this section of the MRV plan will
103
Any leakage detected would constitute a failure of mechanical integrity and would necessitate the
cessation of injection in accordance with the UIC permits.
64
-------
provide a list of possible quantification methods and a discussion of when and how those
methods might be employed for each risk pathway identified in the leakage pathway
assessment. To the extent possible, the expected accuracy of the techniques should also
be discussed. EPA understands that it may be difficult to measure C02 leakage quantities
and that modeling and estimation processes with possibly wide margins of error may be
utilized.
In some cases the leak volume or rate may be estimated using near-surface parameters,
and quantification will reflect a mass of CO2 leaked to the surface that is then reported to
EPA. In other cases, the estimation process will employ subsurface parameters, and the
initial estimates will be of the subsurface leak rate and amount. These subsurface leakage
rate and amount estimates then need to be translated to an estimate of the mass of C02
leaked to the surface, and reported to EPA. The choice of quantification methods in an
MRV plan will be site specific. They could include but are not limited to:
1. Estimation of leakage in the subsurface by material balance, using known injected
quantities and monitored pressures in the IZ. Since leaks will cause monitored
pressure in the IZ to be lower than expected, leaked quantities might be estimated
as those amounts that bring the modeled IZ pressure down to measured values.
Given the uncertainty and variability in IZ reservoir conditions, this technique
would be expected to have a very wide margin of error, particularly in the early
phases of injection when pressure differences above background levels are small
and reservoir characteristics have not been confirmed/adjusted by history-
matching. The estimated leaked amounts will be reported in volume of fluids
(e.g., cubic feet or cubic meters) and should be broken out into the amounts of
water versus amounts of C02. Then the amount of subsurface C02 leakage must
be converted to the amount of C02 that has leaked or will leak to the surface. (See
discussion below on these two estimation steps.)
2. Estimation of subsurface leakage by material balance using monitored pressures
in the ACZ. Leaks through the CZ likely will cause monitored pressure in ACZ to
be higher than expected. This second technique should be more accurate than the
first technique using IZ pressures, because there should be less uncertainty
determining the value of "normal" pressure in the ACZ. However, a wide margin
of error will exist because of the given uncertainty in ACZ reservoir
characteristics. The volume of fluids leaked will have to be converted to CO2
leaked to the surface.
3. Estimation of leakage by reservoir simulation that triangulates leak location and
calculates leak rate by interpreting pressures in ACZ at three or more monitoring
wells. When several wells in the ACZ record pressure increases, it will be
possible to locate and estimate volumes of a leak that is concentrated at one
location (i.e., not made up of several scattered leaks). The volume of fluids leaked
will have to be converted to CO2 leaked to the surface.
4. Estimation of leakage by measuring CO2 flux at several locations on the surface
with eddy covariance, accumulation chamber, tenting, etc., can be used in
extrapolating the entire area of leak. This process would involve identifying that
65
-------
the surface area through which the leak is occurring and directly measuring the
leak rate at a sample of locations at or near the surface. The accuracy of this
technique will depend on the shape and size of the leak at the surface, the sample
footprint size, (i.e., point sampling for accumulation chambers versus a
measurement area for an eddy covariance tower), density of sampling, and the
ability to distinguish leak volumes from natural flux rates.
5. Estimation of a leak by remote sensing or ground surveys of the effects of leaked
C02 on vegetation104. It may be possible to approximate the surface extent and
rate of leakage by looking at vegetative effects. This option is likely to have a
very uncertain accuracy. It may best be used to estimate the areal extent of the
leak or to determine whether subsurface leak has reached the surface.
6. Estimation of the plume size and location using seismic survey by mapping the
primary plume size and density in the IZ and the subsurface leakage plume in the
ACZ. Under favorable geologic circumstances, it will be possible to use periodic
seismic surveys to track movements of the CO2 plume in the IZ through time.
Under those same circumstances, it might be possible to use the seismic surveys
to estimate how much of the CO2 plume is still located in the IZ and how much
has leaked into the ACZ.
Though the state of the science does not currently support using existing logging
techniques to quantify leakage in wells, estimation of well leakage through interpretation
of mechanical integrity well logs of injection wells or monitoring wells drilled into the IZ
can help to bound leakage estimates. When logs indicate the leak of fluids along the
wellbores, it might be possible to estimate the extent of leakage by analyzing those logs
or supplemental logs. For example, the rate of fluid movement may be indicated by the
analysis of the frequency of sound signals from noise logs. The analysis of the velocity of
radioactive tracers up a borehole could also indicate rate of fluid leaks. In addition, an
analysis of dimensions of damaged well materials (e.g., gaps in cement or corroded
casing) determined from sidewall cores and pressure drawdown tests might be used to
constrain of the size of leaks. The volume of fluids leaked will have to be converted to
CO2 leaked to the surface.
When leaks are quantified in the subsurface as volumes of CO2 or fluids, it will be
necessary to convert this value to an estimate of the amount of CO2 that has leaked to the
surface. One method to estimate CO2 volume leaked at the surface would be to apply one
or more of the near-surface options listed in items #4, #5, and #6 above to measure CO2
leak flux and leakage area, and then estimate the leak volume by multiplying leakage area
by flux and duration. Alternatively, the reporter can choose to use techniques such as
modeling to more directly estimate what portion of the subsurface leak is CO2 that has or
will reach the surface. When the leak has been measured as total volume of fluids, the
amount of the leak that is CO2 versus water may be inferred by analysis of where the leak
occurs relative to the position of the plume. The reporter may determine that leaks far
104
Keith C.J., Repasky K.S., Lawrence R.L., Jay S.C., Carlsten J.L. (2009) Monitoring effects of a
controlled subsurface carbon dioxide release on vegetation using a hyperspectral imager. International
Journal of Greenhouse Gas Control 3(5):626-632. doi:10.1016/j.ijggc.2009.03.003
66
-------
outside of the plume area are assumed to be 100 percent water and no CO2. Leaks
through the CZ within the plume area might be estimated based on the relative modeled
CCVwater saturation levels at the point of the leak and modeled flow characteristics
through the CZ.
The quantification of leakage should include consideration of the time at which the
leakage began and when the monitoring system first detected the leak. Subsurface leaks
may be identified relatively quickly if they occur near a monitoring well, but may take
months or years to reach the surface. Leak quantification methods should also consider
the effect of time on leakage rates because the leakage rate at the surface is likely to
increase over time.
The portion of CO2 that leaks through the CZ, out of structural spill points, or along
boreholes that will ultimately reach the surface may also be estimated using modeling.
This will include careful characterization of the rocks and fluids to be encountered by the
C02 along the leakage pathways to the surface. If any part of the leak is modeled as
trapped in a USDW, the assumption should be made that the water with the trapped CO2
may eventually be produced from the USDW, and therefore counted as a leak to the
surface, unless the reporter demonstrates that the CO2 will not be emitted to the
atmosphere. Alternatively, the reporter may report to EPA that 100 percent of the
subsurface leak has reached the surface. Note that if the injectate reaches a USDW, the
reporter will be in violation of their UIC permit.
5.4 Strategy for Establishing the Expected Baselines
Baseline monitoring is essentially the first step in implementing the leakage detection and
quantification monitoring strategy. The primary goal of establishing expected baselines
is so that the reporter can discern whether or not the results of monitoring are attributable
to leakage of injected CO2. This chapter describes considerations for developing
expected baselines that will ensure adequate leakage detection capability within the
monitoring system.
Leakage of sequestered CO2 may result in detectable deviations from the expected
baseline values in one or more of a number of environmental conditions, such as
subsurface pressure, groundwater chemical composition, the concentration of CO2 in air,
soil, surface or near-surface CO2 flux rates, surface CO2 isotope ratios, and other
geophysical and geochemical parameters, or deviations from expected operational
conditions, such as the injection pressure and the annular pressure in the well. The MRV
plan leakage detection and quantification strategy may include monitoring a selection of
these indicator parameters to detect potential CO2 leakages. To judge whether a set of
measured parameter values obtained during GS operations presents a cause for concern,
reporters should know what those parameter values would be expected to be in the
absence leaks. In the MRV plan, reporters should credibly explain how an expected
baseline that represents the most probable range of the indicator parameters will be
obtained. Once the expected baseline has been established, reporters would then have a
67
-------
basis for statistically evaluating if the measured parameter values obtained during GS
operations are of a level that suggests a leakage of injected C02.
In most cases seepage fluxes will far exceed background;105 however, some natural
systems, especially the concentration of CO2 in near-surface soil and air, are quite
variable and are affected by many factors, making prediction of baseline difficult. There
may be situations where conditions limit the ability to collect or model representative
conditions, either for baseline or monitoring applications. For example, in the subsurface,
heterogeneity may create complex pressure pathways between the injection well and the
monitoring well that can result in anomalous or unexpected readings that cannot be
correlated to the injection well pressure. In ER areas with multiple stages of primary,
secondary, and tertiary production and reservoir manipulation, pressure transients and
well stimulation practices may make pressure monitoring unreliable.
5.4.1 Approaches for Establishing an Expected Baseline
The baseline approach described in the MRV plan should be reliable, representative of
the site conditions, and provide enough resolution such that an anomalous reading can be
identified without an unacceptable level of false negative or false positive results.
Inaccuracies in establishing the expected baseline could lead to erroneous conclusions. If
the baseline is inaccurately determined to be too low, then "false positives" would result;
in other words, CO2 leakages would be detected even when they were not occurring. On
the other hand, if the baseline is inaccurately determined to be too high, then "false
negatives" would result; that is, CO2 leakages would not be detected even when they
were in fact occurring.
In the MRV plan submittal, the reporter should provide a description of the baseline
monitoring that demonstrates the statistical validity of the selected approach for each
monitoring method proposed. This discussion should include documentation of the
environmental variability and instrument capability, and a determination of the
probability of detecting a leak of the targeted size(s). The identification and influence of
site characteristics such as topography, land use, weather, and climate on the monitoring
methods for baseline should be addressed in the MRV plan as part of the assessment of
adequacy and potential for variability in conditions.
Pre-injection Monitoring of Environmental Parameters
Optimally the indicator parameters would be measured at the proposed locations prior to
injection. Given the seasonal variability of some parameters, such as CO2 flux, this
historical baseline may need to encompass a range of seasonal and climatic conditions.
An advantage of this approach is that the operational stage and pre-injection values are
derived from the same location, thus minimizing variance derived from geographical
differences. Using the same geographic location for baseline and measurements will
105 Benson, S. 2004. Monitoring Protocols and Life-Cycle Costs for Geologic Storage of Carbon Dioxide.
Presentation at Carbon Sequestration Leadership Forum meeting in Melbourne Australia. September 13-15,
2004.
68
-------
eliminate one potential source of error and variability in the determination of anomalous
readings and potential leakage. However, the disadvantage is that operational stage and
pre-injection values are measured at different points in time, resulting in possible
variance from temporally varying factors such as weather.
Contemporaneous Monitoring of Environmental Parameters
There are two potential approaches involving contemporaneous monitoring of
environmental parameters. One approach involves contemporaneous monitoring of the
indicator environmental parameters at a site similar to the sequestration facility, but at an
offsite location that is unaffected by the CO2 injection. This reference site should be
established to be similar to the MMA in terms of all the relevant environmental
parameters so that it serves as a proxy for what conditions would have been like in the
MMA had the C02 facility not been built and the C02 not been injected. The advantage
of this method is that the reference site readings are contemporaneous to the monitoring
zone readings and would therefore be matched in terms of weather conditions and season
(provided the reference site is fairly close to the MMA). The disadvantage of this
approach is that it is usually difficult to find a reference site that is a perfect proxy for the
facility site in every respect. For this reason it may be necessary to use data from the
reference site indirectly in predictive models of the expected baseline values.
A second approach involves contemporaneous monitoring of the indicator environmental
parameters at reference points within the maximum monitoring area. These reference
points are expected to be currently unaffected by the C02 injection, although they lie
within the maximum monitoring area of the sequestration facility. The advantage of this
approach is that it is likely to be cost-effective to implement. However, it suffers from the
drawback that the reference sites may unexpectedly be affected by the leakage, which
would result in false negative determinations.
Use of Predictive Models
A predictive model for the baseline is a mathematical model, usually developed through
regression analysis, that estimates the expected baseline as a function of other variables
(e.g., predicting expected C02 flux in soils as a function of season, time of day, recent
rainfall, and temperatures). Predictive models may be useful when there may be bias in
the historical data, or when the variation of the monitored parameters is very wide and
when it is known that certain measurable influences, such as weather, account for much
of that variation. The creation of predictive models adds to the complexity and cost of the
MRV plan development and implementation, but may improve the statistical reliability of
a monitoring strategy. If the MRV plan relies on predictive models, then the plan should
lay out how the predictive model will be developed and how the model will change the
bias and variance in the expected baseline. The MRV plan should describe how the
predictive model is based on available monitoring data. Examples of predictive models
69
-------
used to calculate flux can be found in recent studies of natural CO2 releases, ER and
sequestration sites.106'107
5.4.2 Considerations in Establishing Expected Baselines
Adjustments for a Shifting Baseline
Changing natural and land-use factors may necessitate revisiting the expected baseline. In
the case of a baseline defined by pre-injection historical environmental data, changes in
land use in the surrounding region could affect current measured parameter values,
biasing them in a certain direction and increasing the possibility of a false positive or
negative. To prevent this, either the pre-determined baseline would have to be adjusted to
factor in the effect of the current land use change, or the measurement of the current
environmental parameter values would have to be adjusted to factor out the effect of the
land use change.
For a baseline defined by contemporaneous reference data, natural or land use changes in
the reference area that do not influence the monitoring area would be a cause for concern.
In such a situation, reporters would need to define a new reference site more reflective of
original conditions at the monitoring area and create a new baseline there, or they would
need to factor out the influence of the natural/land use changes at the existing reference
site. Depending on the particular influence of the change, failure to make these
adjustments could lead to either false negative or false positive determinations.
If adjustments are made to the strategy for establishing the expected baseline after the
MRV plan is approved by EPA, the reporter must describe the changes in the annual
report. Changes to baseline may result in a material change and would require the MRV
plan to be re-submitted.
Uncertainty and Variability in Baseline Data: The Choice of a Monitoring Method
In the choice of monitoring methods for leakage detection and quantification, the reporter
should consider the issues posed by uncertainty and variability in both the baseline and
the measured operational data. Certain environmental parameters such as CO2 flux are
likely to be harder to measure accurately than others. For example, surface-to-atmosphere
CO2 flux varies considerably by time of day, by season, and by location. A number of
106 Etheridge D., R. Leuning, A. Luhar, D. Spenser, S. Coram, P. Steele, M. van der Schoot, S. Zegelin, C.
Allison, P. Fraser, C.P. Meyer and P. Krummel. 2007. Atmospheric monitoring and verification of
geosequestration. At the C02CRC Otway Project 2007 Annual Report, Report # RPT07-0735.
107 Stephanie Trottier, William D. Gunter, Bernice Kadatz, Mark Olson, Ernie H. Perkins. 2009.
Atmospheric monitoring for the Pembina Cardium C02 Monitoring Project using open path laser
technology, Energy Procedia, Volume 1, Issue 1, Greenhouse Gas Control Technologies 9, Proceedings of
the 9th International Conference on Greenhouse Gas Control Technologies (GHGT-9), 16-20 November
2008, Washington DC, USA, February 2009, Pages 2307-2314. Available at:
http://www.sciencedirect.com/science/article/B984K-4W0SFYG-
BM/2/e8c3fb 1745fea401053 If65c3 5 lbd23 3.
70
-------
variables influence surface-to-atmosphere CO2 flux, many of which are associated with
measurement error. The measurement of surface-to-atmosphere flux using eddy
covariance is reported to have a precision of up to 30 percent, with short-term error of 7
to 12 percent.108 There is also uncertainty about the statistical model that converts wind
and CO2 concentration readings to flux and the values of the coefficients in the model.
This suggests that C02 flux estimates may have sizeable confidence intervals or error
bars.109 In a study of the CO2 leakage potential from the ER operations at the Rangely oil
field in Colorado, Klusman found that C02 flux and carbon isotope ratios varied
significantly at reference locations and over the suspected source areas.110 A predictive
model based on regression analysis of the historical data could predict expected baseline
flux as a function of weather and other conditions.
A sampling program that measures an environmental parameter with a considerable error
bar or confidence interval is likely to be difficult to discriminate from baseline
conditions. High error bars in either or both the operational stage monitoring estimates or
the baseline estimates will increase the probability of false negative readings, thus
potentially leading to erroneous conclusions. An environmental parameter with a low
error bar is preferable to alternative indicators that are associated with high error bars.
Predictive models have been applied at ER and sequestration pilot tests to compare
background CC^flux and concentrations to measured values.111'112'113'114 These studies
demonstrate the use of predictive models and the challenges115 of discriminating natural
conditions from leakage especially over small areas or with low leakage rates.
108 Oldenburg, C.M., and J.L. Lewicki. 2003. Near-surface monitoring strategies for geologic carbon
dioxide storage verification, Lawrence Berkeley National Laboratory Report LBNL-54089, October 2003.
109 Confidence intervals or error bars are used to describe the most probable range of values with which a
measured reading should be associated. Readings that closely match the actual values will have narrow
confidence intervals or error bars; readings with high variability will have wide confidence intervals or
error bars.
110 Klusman, R., 2003. A geochemical perspective and assessment of leakage potential for a mature carbon
dioxide-enhanced oil recovery project as a prototype for carbon dioxide sequestration; Rangely field,
Colorado. AAPG Bulletin V. 87, No. 9 pp 1485-1507
111 Lewicki, J.L., G.E. Hilley, and C.M. Oldenburg. 2005. An improved strategy to detect C02leakage for
verification of geologic carbon sequestration. Geophysical Research
Letters, 32 (19), L19403, doi:10.1029/2005GL024281.
112 Loh, Z., Leuning, R., Zegelin, S., Etheridge, D., Bai, M., Naylor, T., Griffith, D. 2009. Testing
Lagrangian atmospheric dispersion modelling to monitor C02 and CH4 leakage from geosequestration.
Atmospheric Environment doi: 10.1016/j.atmosenv.2009.01.053
113 Benson, S. 2004. Monitoring Protocols and Life-Cycle Costs for Geologic Storage of Carbon Dioxide,
http://www.cslforum.org/documents/benson_sally_wed_Pal_AB_0915 .pdf.
114 Lewicki, J.L., G.E. Hilley, M.L. Fischer, L. Pan, C.M. Oldenburg, L. Dobeck, and L. Spangler. 2009.
Detection of C02 leakage by eddy covariance during the ZERT project's C02 release experiments. Energy
Procedia, 1,2301-2306.
115 Etheridge, D, Leuning, R, Luhar, A, Spencer, D, Coram, S, Steele, P, Van der Schoot, M, Zegelin, S,
Allison, C, Fraser, P, Porter, L, Meyer, C and Krummel, P. 2007. Atmospheric monitoring and verification
of geosequestration at the C02CRC Otway Project 2007 Annual Report to the C02CRC. Cooperative
Research Center for Greenhouse Gas Technologies, Canberra, Australia, C02CRC Publication Number
RPT07-0735. 14pp.
71
-------
Statistical Methods for Comparing Operational Stage Measurements and Expected
Baseline Data
In the MRV plan, the reporter may propose to use a statistical approach to conclude if the
operational stage measurements differ from the expected baseline. Using statistics, the
reporter can infer about a population using data from a sample. In many cases a statistical
approach is advisable because environmental parameters tend to be inherently variable-
even the best sampling program would only generate an estimate of the true value of a
parameter. This sampling estimate could be mathematically expressed as a statistical
probability distribution, which represents the most probable range of values for the
parameter based on the sample data. A range of statistical tests are available to determine
if the measured operational stage data differs significantly from the baseline.
Heterogeneous populations should be split into homogeneous subdivisions to ensure that
like quantities are being compared and the statistical comparisons are meaningful. For
more information, see EPA's Data Quality Assessment statistical guidance.116
Using the Regional Hydrologic Evaluation to Determine Expected Conditions in the
Monitoring Zones
An understanding of the regional hydrogeologic characteristics in which the GS facility is
situated will provide the context for conceptual site model development, and may be used
to help quantify potential leaks from a GS reservoir. As an example, the presence of
small-scale structural, stratigraphic, or geochemical features, such as faults, trapping
structures, or mineral deposits, are often indicated by the broader geologic setting of an
area. However, these features, which could impact geochemical monitoring parameters,
may not be observed during site-specific investigations. IZ hydrogeologic and
geochemical conditions should be evaluated during development of an expected baseline
to anticipate the movement of injected CO2 after injection has begun. As part of baseline
data collection, data could be gathered to establish the ambient pressures, direction,
velocity, and geochemical nature of native fluids in the reservoir and ACZ prior to GS
facility operations. These data can be used to predict CO2 plume migration, assist
calibration of reservoir models, and during periodic reviews to quantify differences
between observed changes in the hydrologic regime and predicted effects. The regional
model can also be combined with the site characterization model to assist siting of
monitoring infrastructure.
In the strata comprising the ACZ, determination of baseline hydraulic and geochemical
conditions may allow determination of the presence and magnitude of any leakage that
may be occurring. If strata in the ACZ are in hydraulic communication with a USDW,
pumping regime changes in the USDW may cause pressure fluctuations that could be
misinterpreted as being indicative of a release from the reservoir. With a sufficient period
of baseline monitoring, seasonal fluctuations in ACZ pore pressures can account for
seasonal trends in groundwater pumping, resulting in regular increases and decreases in
ACZ fluid pressures.
116 US EPA. 2006. Data Quality Assessment: Statistical Methods for Practitioners (EPA QA/G-9S, 2006)
72
-------
Geochemical data are also an important part of baseline data collection, as naturally
occurring localized mineral deposits or the presence of organic constituents such as
natural gas or naturally occurring CO2 may mimic or obscure conditions indicative of a
C02 release. Depending on the site and on the proposed monitoring strategy, the GS
reservoir and ACZ geochemistry could be measured as part of a baseline assessment, due
to the possible presence of naturally occurring C02, metals that may mobilize in the event
of a GS reservoir leak such as iron and manganese, or naturally occurring constituents
that could also be introduced by the C02 stream such as argon, nitrogen oxide, hydrogen
sulfide, and methane.
5.5 Site-Specific Variables for the Mass Balance Equation
The mass balance equations RR-11 or RR-12 are used to calculate the amount of CO2
that is reported as sequestered and include the terms C02fi and C02Fp. C02fi is defined as
the total annual C02 mass emitted (metric tons) as equipment leakage or vented
emissions from equipment located on the surface between the flow meter used to measure
injection quantity and the injection wellhead. C02fp is defined as total annual C02 mass
emitted (metric tons) as equipment leakage or vented emissions from equipment located
on the surface between the production wellhead and the flow meter used to measure
production quantity. The MRV plan should include a summary of considerations made to
calculate equipment leaks and vented emissions from surface equipment between the
flow meters and either injection or production wellheads, and the quantity of C02 that is
produced with oil and water.
The mass balance equation assumes that all C02 measured at the flow meter is injected at
the wellhead. In most cases this will provide an accurate estimate of injection volume.
However, if additional surface equipment is located on the injection line after the point of
flow/mass measurement, and before the injection well, there is a potential for equipment
leaks and vented emissions that will not be included in the mass balance. Similarly, the
measurement point for produced C02 (at the flow meter immediately downstream of the
gas-liquid separator) is assumed to represent the amount of C02 exiting the wellhead.
However if additional surface equipment other than the separator is located between the
production wellhead and point of C02 measurement, an estimate of equipment leaks and
vented emissions from that equipment must be made in accordance with the monitoring
methods in §98.233 of the Petroleum and Natural Gas Systems Reporting Rule (subpart
W of the GHG Reporting Program).117
The following table lists the surface C02 emission sources at an injection facility that
may be located between the point of transfer onsite and the injection wellhead (selected
from the list of all emissions sources under onshore petroleum and natural gas production
at §98.232) and describes the methodologies provided in §98.233 of subpart W. Note that
the compressors may or may not be a part of this operation depending on how the
individual site configures its recycle stream. Compressors are used to compress the
recycle stream, which is eventually converted to critical phase C02that is mixed with the
critical phase stream coming from outside facility boundary. In the event that there are
117 http://www.epa.g0v/climatechange/emissi0ns/subpart/w.html
73
-------
other vented or fugitive CO2 emissions sources that are between the flow meter and the
wellhead, the Reporter will need to use the respective monitoring methods for such
sources as provided in §98.233.
Table 5-1: Surface Components as Potential CQ2 Emissions Sources at Injection Facilities
Emissions Source
Engineering
Estimates
Direct
Measurement2
Equipment Count
and
Population Factor3
Natural gas pneumatic high bleed
device venting4
X
Natural gas pneumatic low bleed
device venting4
X
Natural gas pneumatic intermittent
bleed device venting4
X
Natural gas driven pneumatic pump
venting4
X
Reciprocating compressor rod and
packing venting
X
EOR injection pump
X
EOR injection pump blowdown
X
Centrifugal compressor wet seal oil
degassing venting
X
Other equipment leaks (valve,
connector, open-ended line,
pressure relief valve)
X
(1) The engineering calculation methods use monitored process operating parameters and either
engineering calculations (in the case of EOR injection pumps) or emission factors provided by the
equipment manufacturer (in the case of pneumatic device or pneumatic pump).
(2) Direct measurement involves use of rotameters, turbine meters, or other meters, as appropriate, for
emissions measurement directly from the vent stack. This would be repeated once each year to establish a
site-specific, equipment-specific emission factor for use between repeat measurements.
(3) For the use of population factors, the relevant emission factors would be applied to all components
using a count of the sources.
(4) ER operations that power pneumatic equipment by instrument air rather than natural gas would not have
natural gas pneumatic emissions sources.
Equipment leaks and vented CO2 emissions at an ER operation can also occur from
surface components located between the production well and the point of CO2
measurement. Between these two points, an injection facility will typically have valves,
connectors, meters, and headers (which are large pipes that mix the oil stream from
74
-------
multiple wellheads). The procedures provided in subpart W for determining CO2
emissions from all of these sources are equipment count and population factor.
The MRV plan must provide the basis for calculations of equipment leaks and vented
emissions estimates from these points and include the emissions in the mass balance
equations RR-11 or RR-12.
Fluids, including oil and water, produced at GS facilities may contain C02 that is not
captured in the gas phase CO2 measurement downstream of the gas-liquid separator. The
C02 is present as dissolved gas in the oil and water phase of the produced fluids. To
account for the amount of produced CO2 that is not sequestered, the reporter must tailor
the mass balance equation to account for the mass of C02 in all produced fluids. Equation
RR-9 describes the mass balance of CO2 at the gas-liquid separator where CC>2,p
represents the total mass of C02 in the incoming production stream (oil, water and C02),
and C02,w represents the amount of CO2 in the separated gas stream. The value of "X" is
the percent of C02 expected to remain with the produced oil and gas when referenced to
the CO2 separated for recycle or reuse.
The MRV plan must describe how the volume of CO2 in produced fluids will be
determined, and how the value of "X" will be calculated. The reporter should describe
how the analysis of CO2 in produced fluids will be performed. Commercial laboratories
perform crude oil analyses and separator liquid analyses that can include quantification of
CO2 as a component. Because separators operate under pressure, the CO2 saturation in
water and oil inside the separator is higher than it would be at atmospheric conditions.
Care must be taken to ensure the measured CO2 concentration is representative of the
pressure conditions at the measurement location. For analysis, the water and oil may need
to be separated and analyzed using different methods. Several steps would be needed in
order to estimate the quantity of CO2 produced with oil, water or gas. The oil or gas
stream of a production well at an ER operation is first separated into liquid phase and gas
phase; the liquid phase is further separated into a water fraction and an oil fraction. CO2
may be found in the following three streams following the gas-water-oil separation phase:
1) CO2 in oil, 2) CO2 in water, and 3) CO2 in gas phase (along with other produced
hydrocarbons). The amount of CO2 in the oil (or water) can be estimated by multiplying
the gas-to-oil (or gas-to-water) ratio by the volume of oil (or water) produced. The gas-
to-oil (or gas-to-water) ratio can be determined using any standard test method (the ratio
determination as well as chromatography are well understood techniques). A method to
determine the amount of CO2 in gas phase is to use a mass flow meter or volumetric flow
meter to measure mass or volumetric flow rate of the produced gas stream that contains
CO2, take a sample of the stream, and analyze the stream for CO2 composition, which can
be done using chromatography.
A total fluids analysis from the separator liquid stream can also be performed to
determine the amount of dissolved CO2 in the oil and produced water. The total fluid
analyses are often a combination of several ASTM or other methods. The selected
sampling and analytical method(s) should consider the pressure at which the fluid is
separated from the gas and differences in CO2 solubility in oil and water. EPA has not
75
-------
identified consensus-based or industry standard analytical methods specifically for
determining the amount of C02 in mixed produced fluids (oil and water); however, two
potentially applicable methods for CO2 in hydrocarbons were identified:
ASTM D2597 - 94(2004) Standard Test Method for Analysis of Demethanized
Hydrocarbon Liquid Mixtures Containing Nitrogen and Carbon Dioxide by Gas
Chromatography
GPA Standard 2177 Analysis of Demethanized Hydrocarbon Liquid Mixtures Containing
Nitrogen and Carbon Dioxide by Gas Chromatography
To measure the C02 content in water co-produced with the oil, the reporter may use the
following potentially applicable method:
ASTM D2513 - 06 Standard Test Method for Total Dissolved Carbon Dioxide in Water
The reporter must describe the analytical methods used and demonstrate the applicability
of the methods in the MRV plan.
It is common in centralized production collection and processing sites receiving
production from dozens of surrounding wells to have a "test trap." This is a small
gas/oil/water separator, outfitted with accurate flow measuring devices on all streams
in/out, and manifolded to divert any one well through the test trap to gauge gas/oil/water
ratio of each well. This is important in balancing the extraction of production between
wells. If a well is drawing too much water, it can be slowed down and adjacent wells
increased in production. This is where gas/oil/water production would be monitored for
each well individually and regularly. Therefore, it is not necessarily the case that the
sample point would involve a single CO2 stream from a single well. There may be an
aggregate point of monitoring from a collection of wells.
At some geologic sequestration sites where oil is not produced, saline water may be
removed from the IZ to manage the pressure and capacity of the CO2 storage system. The
amount of dissolved CO2 in the saline water removed from the IZ may be measured using
ASTM D2513 - 06 Standard Test Method for Total Dissolved Carbon Dioxide in Water,
or other consensus-based or industry standard methods. The MRV plan should describe
the fate of the produced water and account for all CO2 in the produced water.
5.6 Quality Assurance and Quality Control
To ensure accuracy in reporting the data from natural and engineered systems, reporters
should document their quality assurance program for data collection and analysis. The
MRV plan should define what quality assurance and quality control procedures will be
implemented for each technology applied in the leak detection and quantification process.
For example, the plan should describe how the reporter will ensure that sensor readings
are correct, and describe (or reference) the quality control measures that are in place to
ensure precision, accuracy, representativeness, completeness, and comparability of data.
76
-------
Examples of quality control measures that may be described in the MRV plan include
standard operating procedures, calibration requirements, redundancy of the sensor or
measurement, criteria for equipment selection, maintenance schedules and procedures,
and surveillance of subcontracted surveys or vendor materials.
5.7 Missing Data
Missing data procedures for the quarterly values of mass or volume, density, and
concentration for both CO2 injected and CO2 received are provided at 40 CFR 98.445.
Procedures are also provided for missing data on C02 production and for equipment leaks
and vented emissions.
The rule does not require the reporter to develop missing data procedures for monitoring
data described in the MRV plan. However, if leakage is detected for which a
quantification approach is not outlined in the plan, information on the quantification
approach should be included in the annual report. Similarly, if additional leakage
pathways are identified, they should be assessed in accordance with the procedures in the
"Assessment of potential surface leakage pathways" section of the MRV plan, and
included in the annual report. Procedures for developing missing leak quantification data
should be included in the MRV plan.
77
-------
6. MRV Plan Approval Process Overview
If a facility is subject to subpart RR, the MRV plan must be submitted:118
• By June 30, 2011 if the facility was issued a UIC permit authorizing CO2
injection on or before December 31, 2010.
• Within 180 days of receiving a final UIC permit (or offshore authorization)
authorizing CO2 injection.
• Any time if injecting CO2 to enhance oil and gas recovery and not permitted
under UIC Class VI.
Facilities will be allowed one extension of up to an additional 180 days. After submission
of the MRV plan, the following review and approval process steps apply:
• Notification of receipt of MRV plan.
• Completeness check.
• Technical review.
• Issuance of EPA decision.
This chapter describes each step of the MRV plan approval process.
6.1 Suggested Outline for the MRV Plan
A suggested outline for the MRV plan may be found in Appendix D. The reporter may
choose to modify this outline to meet site-specific requirements. The MRV plan outline
addresses characteristics within the MMA; however, the plan may include information on
geologic, topographic, cultural, and potential CO2 source conditions outside the MMA,
where appropriate, to assist EPA in developing a full understanding of potential
influences on the proposed monitoring strategy.
6.2 Notification of Receipt of MRV Plan
EPA will accept MRV plan submittals from reporting entities that have established a
"Certificate of Representation" through the GHG Reporting Program's Electronic
Greenhouse Gas Reporting Tool (e-GGRT). All MRV plans must be submitted by the
reporter electronically through e-GGRT. EPA has provided information and directions on
how to electronically submit MRV plans on the subpart RR Web site:
http://www.epa.gov/climatechange/emissions/subpart/rr.html. Upon MRV plan
submittal, EPA will send a notice of receipt to the reporter within 15 days to
acknowledge that EPA has received the MRV plan submission.
6.3 Completeness Check
118 Please refer to http://www.epa.gov/climatechange/emissions/data-reporting-svstem.html for information
on electronic data reporting.
-------
EPA will first conduct a completeness check of the MRV plan submittal based on the
regulatory requirements identified in Table 6-1. In addition, if a facility uses their UIC
permit to fulfill or provide the basis for certain MRV plan requirements, the reporter
should clearly identify relevant sections of the UIC permit as well as reference this
information clearly in the MRV plan. EPA will determine if the MRV plan is complete
within 45 days of the notice of receipt and will notify the reporter whether the plan is
complete or incomplete.
EPA will issue a written notice that either requests additional information from the
reporter or provides notice to the reporter that the MRV plan is complete and that the
technical review will commence. In the case that the MRV plan is incomplete, the notice
would specify the MRV plan requirements that were deemed to be incomplete, what
additional information is needed, and the rationale for why such additional information is
needed. If incomplete, the reporter must submit an updated MRV plan within 45 days of
EPA notification unless otherwise specified by EPA.
Note that the completeness check is not a substitute for the technical review process. EPA
will initiate the technical review process only upon determination of the completeness of
the MRV plan.
Table 6-1: Completeness Check Criteria for MRV Plan Submittals and Resubmittals
Requirement
Subpart RR Section
The facility holds a certificate of representation
40 CFR 98.448(g)
Delineation of the maximum monitoring area and the active
monitoring areas as defined in §98.449
40 CFR 98.448(a)(1)
Identification of potential surface leakage pathways for C02 in
the maximum monitoring area and the likelihood, magnitude,
and timing, of surface leakage of C02 through these pathways
40 CFR 98.448(a)(2)
A strategy for detecting and quantifying any surface leakage of
co2
40 CFR 98.448(a)(3)
A strategy for establishing the expected baselines for
monitoring C02 surface leakage
40 CFR 98.448(a)(4)
Summary of considerations to calculate site-specific variables
for mass balance equation
40 CFR 98.448(a)(5)
If a well is permitted or is to be permitted under the UIC
Program, the well identification number in the permit or permit
application and the UIC permit class
40 CFR 98.448(a)(6)
If an offshore well is not subject to SDWA, any well
identification number and any identification number used for
the legal instrument authorizing GS
40 CFR 98.448(a)(6)
Proposed date to begin collecting data for calculating total
amount sequestered according to equation RR-11 and RR-12 of
subpart RR
40 CFR 98.448(a)(7)
For MRV plan revisions, the reason for resubmission
40 CFR 98.448(d)
6.4 Technical Review
79
-------
After the determination of completeness, EPA will initiate technical review of the MRV
plan. After 60 days of technical review, EPA will send the reporter a written request for
additional information, including clarifying technical questions, if necessary. The reporter
will be encouraged to provide a response to this request within 15 days; however, EPA
recognizes that there may be circumstances where additional time is needed for the
reporter to collect the information requested.
6.5 Issuance of EPA Decision
Within a reasonable period of time, EPA will issue a final MRV plan as submitted, or
with revisions. EPA will post the approved MRV plan on a public Web site, subject to
any limitations or requirements in its CBI determination.119
6.6 Appeals Period
If the reporter, or any interested person, objects to EPA's final decision, it may be
appealed to EPA's Environmental Appeals Board. See Section II.B of the preamble for
more information on the appeals process and see
http://www.epa.gov/climatechange/emissions/subpart/rr.html for information on joining
the interested person list.
6.7 Resubmittal of MRV Plans
An MRV plan may require revision and resubmittal if a material change was made to
monitoring and/or operational parameters that was not anticipated in the original plan, if
there has been a change in the permit class of the reporter's UIC permit, if the reporter is
notified by EPA of errors in their MRV plan or annual monitoring report, or if the
reporter chooses to revise their MRV plan for any other reason.
Examples of material changes include but are not limited to: large changes in the volume
of CO2 injected; the construction of new injection wells not identified in the MRV plan;
failures of the monitoring system including monitoring system sensitivity, performance,
location, or baseline; changes to surface land use that affects baseline or operational
conditions; observed plume location that differs significantly from the predicted plume
area used for developing the MRV plan; a change in the maximum monitoring area or
active monitoring area; or a change in monitoring technology that would result in
coverage or detection capability different from the MRV plan.
The reporter must submit a revised MRV plan if notified by EPA of errors in the MRV
plan or annual monitoring report (AMR). A suggested format is provided in Appendix D.
119 At the time of publication of this TSD, the CBI determination rulemaking for the GHG Reporting
Program is not finalized (see Section II.B of the final preamble for a brief description of the CBI
determination rulemaking). Please refer to http://www.epa.gov/climatechange/emissions/CBI.html for
more information.
80
-------
7. Annual Monitoring Report and Records Retention
Under the GHG Reporting Program, all reporters are required to electronically submit an
annual GHG report to EPA that includes general reporting elements that are common
across all sectors covered under the GHG Reporting Program,120 as well as reporting
elements that are specific to each subpart. This chapter outlines information to be
included in the subpart RR annual monitoring report (AMR), per 40 CFR 98.446(f)(12).
AMRs will be submitted each year by reporting entities through the e-GGRT system with
other GHG data, and will summarize monitoring activities conducted in the past year.
Reporters must include four components in AMRs:
• A narrative history of the monitoring effort conducted;
• A description of any non-material changes made to the MRV plan;
• A narrative history of any monitoring anomalies that were detected in the
previous calendar year and how they were investigated and resolved; and
• A description of surface leakages, if any.
These four components are described in this chapter. Appendix D provides a suggested
outline of an AMR. This outline includes all the key elements required to be included in
an AMR, and is intended to be a suggestion for reporters to follow. The reporter may
choose to modify this outline to meet site-specific requirements. Other information that
reporters may deem as necessary to include in an AMR can and should be included as
well.
7.1 Narrative History of Annual Monitoring Efforts
The first required component of an AMR is a narrative history of the monitoring effort
conducted over the reporting year, including a listing of all monitoring equipment that
was operated and its period of operation and any relevant tests or surveys that were
conducted. This narrative history may include, for example, a general discussion cross-
walking the monitoring activities implemented during the reporting year with the
activities outlined in a facility MRV plan. Reporters may include a table that presents all
the monitoring equipment used during a reporting period and the related information.
7.2 Report of Non-Material Changes to the MRV Plan
The second component required to be included in the AMR is a detailed report of any
non-material changes made to the MRV plan over the course of the reporting year. Non-
material changes to the MRV plan are those that do not warrant the submission of a
revised MRV plan to EPA. Non-material changes to an MRV plan may include, for
example, the use of newer versions of a technology that are already included as part of
the EPA-approved MRV plan, or changes in the location of equipment that do not alter
the coverage of detection capability of monitoring. To present this information in the
AMR, a reporter could include a description of the non-material changes made to the
120 Please refer to EPA's GHG Reporting Program Web site at:
http://www.epa.gov/climatechange/emissions/ghgrulemaking.html
-------
MRV plan, including an identification of the part of the MRV plan affected by the non-
material change, supporting information for why the change is non-material, the purpose
of the change, when the non-materials change occurred, and if it affected facility
operations. To articulate non-material changes in the AMR, it may be useful for reporters
to identify, in a list or table, the conditions that were used to determine what constituted a
non-material change to their EPA-approved MRV plan. (See Appendix D)
7.3 Narrative History of Monitoring Anomalies
The AMR must also contain a narrative history of monitoring anomalies that were
detected in the year and how they were investigated and resolved. This discussion could
include a detailed description of all monitoring anomalies detected, and how the
anomalies were investigated in line with the methods in the EPA-approved MRV plan, to
determine if they represent a leak. This part of the AMR may also contain a description
of how the monitoring anomalies were resolved. As an example, if an anomalous
pressure reading at a monitoring well, that is later found out to be caused by a vent line
obstruction, is detected, a reporter would include the following information in the AMR:
a description of the pressure anomaly detected, a description of the steps taken to further
investigate the reason for anomalous pressure reading and evaluate if it represents a leak
(i.e., how the obstruction in the vent line was identified). The reporter may also include a
description of the actions the reporter has taken/is planning to take to rectify the
obstruction in the vent line. Reporters could also present this information along with
other monitoring information presented in a table format. (See Appendix D)
This section of the AMR should also include a qualitative discussion of any unexpected
operational or maintenance issues that caused down-time for the equipment. If the
evaluation and investigation of a monitoring anomaly did result in the detection and
quantification of a leak, reporters should include information discussed in Chapter 7.4.
7.4 Description of Surface Leakage
The fourth required component of the AMR is a description of surface leakages that
happened during the reporting year (if any). Reporters must provide a description of any
surface leakages of CO2, including a discussion of all methodologies and technologies
involved in detecting and quantifying the surface leakages and any assumptions and
uncertainties involved in calculating the amount of CO2 emitted.
In this section of the AMR, as part of the discussion of the methodologies and
technologies involved in detection and quantifying the surface leakages, reporters could
provide EPA with a demonstration showing that the methods and technologies used for
leak detection and quantification followed the methods specified in the facility's MRV
plan. This section might include, for instance, monitoring measurement records that
showed monitoring anomalies that resulted in the reporter having to confirm the detection
of a leak and quantify the leakage in that reporting year. If non-material deviations are
made from the leakage detection and quantification approaches specified in the MRV
plan it must be documented in the AMR. (See Chapter 7.2.)
82
-------
This component of the AMR may also include qualitative descriptions and quantitative
identifications (i.e., if a leak rate were assumed) of the assumptions that went into each
leakage calculation. Reporters may also provide information demonstrating the
assumptions' validity. Information on uncertainties may include a qualitative analysis of
the uncertainty around the amount of C02 emitted. For instance, such an analysis may
include a description of how variables such as the accuracy and precision of measurement
instruments may have affected calculating the quantity of leakage determined, and how
uncertainty in baseline conditions may have affected calculations as well. Reporters may
also, if feasible, include a detailed quantitative analysis of the uncertainty around the
quantity of leakage calculated. If a quantitative analysis is done, reporters should include
examples of all calculations made, and descriptions of the statistical methodologies and
parameters used.
In the AMR, reporters should show the verification of the quantity of leakage in
accordance with the methods in the EPA-approved MRV plan. Reporters should provide
calculations used to determine the quantity of leakage.
7.5 Records Retention Requirements for MRV Plans
Subpart RR requires facilities to retain records specified for retention in the MRV plan.
These may include data that support the development and implementation of the MRV
plan. Such records may be required to provide an auditable record of documentation in
the event of an EPA-requested audit. The records to be retained should be sufficient to
document the rationale of the decision-making process and demonstrate the completeness
and accuracy of the data and calculations used in the detection and quantification of leaks
to the surface. The MRV plan should state which documents are to be retained, and the
format and/or location of the records. The duration for retention is 3 years per 40 CFR
98.3(g). Examples of records to be retained may include, but are not limited to,
operational data, data sets used to construct expected baseline conditions, data relating to
the detection or quantification of surface leakage, logs of mechanical integrity tests and
pressure tests, and field books and photographs that record observations of field
conditions.
83
-------
Appendix A: Glossary
Above Confining Zone (ACZ): The geologic formation that is immediately above the
caprock or CZ.
Active Monitoring Area (AMA): The area that will be monitored over a specific time
interval from the first year of the period (n) to the last year in the period (t). The
boundary of the AMA is established by superimposing two areas: 1) the area projected to
contain the free-phase CO2 plume at the end of year t, plus an all-around around buffer
zone of one-half mile or greater if known leakage pathways extend laterally more than
one-half mile; and 2) the area projected to contain the free phase CO2 plume at the end of
year t+5. (40 CFR 98.449)
Carbon dioxide (C02) plume: The extent underground, in three dimensions, of an
injected CO2 stream. (40 CFR 146.81(d))
Carbon dioxide (CO2) stream: CO2 that has been captured from an emission source
(e.g. a power plant) plus incidental associated substances derived from the source
materials and the capture process, and any substances added to the stream to enable or
improve the injection process. (40 CFR 146.81(d))
Confining Zone (CZ): A geologic formation, group of formations, or part of a formation
strati graphically overlying the IZ(s) that acts as a barrier to fluid movement. For UIC
Class VI wells operating under an injection depth waiver, CZ means a geologic
formation, group of formations, or part of a formation stratigraphically overlying and
underlying the IZ(s). (40 CFR 146.81(d))
Equipment leak: Those emissions that could not reasonably pass through a stack,
chimney, vent, or other functionally-equivalent opening. (40 CFR 98.449)
Expected baseline: The expected baseline is the anticipated value of a monitored
parameter that is compared to the measured monitored parameter. (40 CFR 98.449)
Formation or geological formation: A layer of rock that is made up of a certain type of
rock or a combination of types. (40 CFR part 144 and part 146)
Free-phase CO2 plume: That part of the CO2 plume in which the injected CO2 stream
exists in gaseous, liquid or supercritical free phase. The precise definition will be stated
by the reporter in the MRV plan based on site characteristics and the proposed methods
of monitoring and modeling.
Free-phase CO2 plume area: The 2-D vertical projection onto the surface of the free-
phase CO2 plume. The free phase CO2 plume area is included within the AMA and
MMA.
84
-------
Geophysical surveys: The use of geophysical techniques (e.g., seismic, electrical,
gravity, or EM surveys) to characterize subsurface rock formations. (40 CFR part 144
and part 146)
Injectate: The fluids injected (40 CFR part 144 and part 146). For the purposes of this
rule, this is also known as the C02 stream.
Injection Zone (IZ): A geologic formation, group of formations, or part of a formation
that is of sufficient aerial extent, thickness, porosity, and permeability to receive CO2
through a well or wells associated with a GS project. (40 CFR 146.81(d))
Maximum Monitoring Area (MMA): The area that must be monitored under this
regulation and is defined as equal to or greater than the area expected to contain the free
phase C02 plume until the C02 plume has stabilized plus an all-around buffer zone of at
least one-half mile. (40 CFR 98.449)
Mechanical integrity (MI): The absence of significant leakage within the injection
tubing, casing, or packer (known as internal mechanical integrity), or outside of the
casing (known as external mechanical integrity. (40 CFR part 144 and part 146)
Mechanical Integrity Test (MIT): A test performed on a well to confirm that a well
maintains internal and external mechanical integrity. MITs are a means of measuring the
adequacy of the construction of an injection well and a way to detect problems within the
well system. (40 CFR part 144 and part 146)
Operational stage: The operational stage of a GS system includes the period of time
during injection.
Pore space: Open spaces in rock or soil. These are filled with water or other fluids such
as brine (i.e., salty fluid). CO2 injected into the subsurface can displace pre-existing fluids
to occupy some of the pore spaces of the rocks in the IZ. (40 CFR part 144 and part 146)
Separator: means a vessel in which streams of multiple phases are gravity separated into
individual streams of single phase. (40 CFR 98.449)
Stratigraphic zone (unit): A layer of rock (or stratum) that is recognized as a unit based
on lithology, fossil content, age, or other properties. (40 CFR part 144 and part 146)
Subsurface CO2 Leakage: Subsurface leakage is injected CO2 that is present outside the
IZ but has not reached the surface.
Surface CO2 Leakage: The movement of the injected CO2 stream from the IZ to the
surface, into to the atmosphere, indoor air, oceans, or surface water. (40 CFR 98.449)
Vented emissions: intentional or designed releases of CH4 or CO2 containing natural gas
or hydrocarbon gas (not including stationary combustion flue gas), including process
85
-------
designed flow to the atmosphere through seals or vent pipes, equipment blowdown for
maintenance, and direct venting of gas used to power equipment (such as pneumatic
devices). (40 CFR 98.449)
86
-------
Appendix B: U.S. Oilfields Using C02 Injection for Enhanced Oil Recovery
*Note that this table is a based on data summarized from the Oil & Gas Journal Enhanced Oil Recovery
Survey published in April 2008 and April 2010 with permission of the publisher. Any other use of these
data may be subject to copyright restrictions.
NR indicates that no production data were reported in the Oil & Gas Journal Survey.
Field
Operator
State
Incremental
Production
(bbl per day)
Rangely Weber Sand
Chevron
CO
11,600
Hall-Gurney (removed 2010)
Murftn Drilling
KS
3
Lockhart Crossing
Denbury Resources
LA
NR
Charlton 30-31
Core Energy
MI
75
Charlton 6
Core Energy
MI
10
Chester 2 (added 2010)
Core Energy
MI
NR
Dover 33
Core Energy
MI
75
Dover 3 5
Core Energy
MI
286
Dover 36
Core Energy
MI
195
Brookhaven
Denbury Resources
MS
3,100
Cranfield (added 2010)
Denbury Resources
MS
10,150
East Mallalieu
Denbury Resources
MS
1,800
Eucutta
Denbury
MS
3,000
Lazy Creek
Denbury Resources
MS
250
Little Creek
Denbury Resources
MS
1,650
Martinville
Denbury
MS
830
McComb
Denbury Resources
MS
1,650
Smithdale
Denbury Resources
MS
600
Soso
Denbury Resources
MS
1,700
Tinsley
Denbury Resources
MS
NR
West Heidelberg (added 2010)
Denbury Resources
MS
2,050
West Mallalieu
Denbury Resources
MS
6,200
North Hobbs
Occidental
NM
6,300
Vacuum
Chevron
NM
2,950
Vacuum
ConocoPhillips
NM
5,200
Bradley Unit
Merit Energy
OK
600
Camrick
Chaparral Energy
OK
1,175
Northeast Purdy
Merit Energy
OK
1,800
Postle
Whiting Petroleum
OK
4,500
Postle Expansion
Whiting Petroleum
OK
1,700
Sho-Vel-Tum
Chaparral Energy
OK
1,170
Abell (added 2010)
Fasken
TX
85
Adair San Andres Unit
Hess
TX
900
Adair Wolfcamp Unit (added
2010)
Apache
TX
420
87
-------
Albert Spicer Unit (added 2010)
Chaparral Energy
TX
NR
Anton Irish
Occidental
TX
4,000
Booker Trosper Unit (added
2010)
Chaparral Energy
TX
NR
Cedar Lake
Occidental
TX
2,860
Central Mallet Unit
Occidental
TX
2,100
Cogdell
Occidental
TX
5,900
Cordona Lake
XTO Energy Inc.
TX
400
Dollarhide (Clearfork "AB") Unit
Pure Resources (Chevron 2010)
TX
1,970
Dollarhide Devonian Unit
Pure Resources (Chevron 2010)
TX
124
East Ford
Orla Petco
TX
128
East Penwell (SA) Unit
Energen Resources
TX
450
El Mar
Occidental
TX
270
Frazier Unit
Occidental
TX
925
Garza
George R. Brown
TX
NR
GMK South
Occidental
TX
375
Goldsmith
XTO Energy Inc.
TX
20
Gramstorff Unit (added 2010)
Chaparral Energy
TX
NR
Hanford
Fasken
TX
300
Hanford East
Fasken
TX
45
Hanford San Andres (added 2010)
Fasken
TX
NR
Hansford Marmaton
Stanberry Oil
TX
102
Igoe Smith
Occidental
TX
440
Levelland
Occidental
TX
950
Mabee
Chevron
TX
2,000
Means (San Andres)
ExxonMobil
TX
8,700
Mid Cross-Devonian Unit
Occidental
TX
296
N. Cross-Devonian Unit
Occidental
TX
835
North Cowden Demo.
Occidental
TX
80
North Dollarhide
Occidental
TX
1,000
North Perryton
Chaparral Energy
TX
170
North Ward Estes
Whiting Petroleum
TX
700
Reinecke
Pure Resources (Chevron 2010)
TX
830
River Bend (Devonian) (added
2010)
Fasken
TX
20
S. Cross-Devonian Unit
Occidental
TX
5,790
SACROC
Kinder Morgan
TX
24,227
Salt Creek
Occidental
TX
6,600
Seminole San Andres Unit - Main
Pay and ROZ
Hess
TX
24,350
Sharon Ridge
Occidental
TX
400
Slaughter
Apache
TX
4,580
Slaughter (5 Units)
Occidental
TX
3,800
Slaughter Sundown
Chevron
TX
4,747
South Cowden
ConocoPhillips
TX
250
88
-------
South Welch
Occidental
TX
865
T-Star (Slaughter Abo)
Occidental
TX
2,100
Twofreds
Great Western Drilling
TX
170
Wasson (Cornell and Mahoney
Units)
XTO Energy Inc.
TX
2,250
Wasson (4 Units)
Occidental
TX
44,325
West Welch
Occidental
TX
0
Yates
Kinder Morgan
TX
6,280
Greater Aneth
Resolute Natural Resources
UT
400
Greater Aneth Area
ExxonMobil
UT
3,000
Beaver Creek (added 2010)
Devon
WY
2,425
Lost Soldier
Merit Energy
WY
7,221
Patrick Draw Monell
Anadarko
WY
3,000
Salt Creek
Anadarko
WY
6,000
Sussex
Anadarko
WY
NR
Wertz
Merit Energy
WY
4,019
89
-------
Appendix C: Summary of Other End Uses of Captured or Produced C02
The following is a summary of end uses of captured or produced CO2 (with the exception
of ER and GS).
Cement. While not a common practice in the cement industry, some companies are
currently working to employ a cement manufacturing process that uses captured CO2.121
The process will convert CO2 into carbonic acid, which will be made into CO3"". The
CO3"" will be combined with calcium and magnesium from seawater to make calcium
carbonate and magnesium carbonate for the cement.122 EPA does not have enough
information to conclude whether cement production using this CO2 capture method
permanently sequesters CO2.
Precipitated Calcium Carbonate (PCC). PCC is produced through a chemical reaction
process that utilizes calcium oxide (quicklime), water, and CO2. The CO2 used to
manufacture the PCC is in some cases captured from lime kilns operated at pulp and
paper manufacturers. The PCC is available in numerous crystal morphologies and sizes,
which can be tailored to optimize performance in specific applications. EPA does not
have enough information to conclude whether PCC production using this CO2 capture
method permanently sequesters CO2.
Food and Beverage Manufacturing. The food and beverage industry is one of the largest
consumers of manufactured and captured CO2 in the U.S. CO2 is used widely throughout
the food and beverage industry for a variety of applications including:
• Carbonation of soft drinks;
• Beer and wine production;
• Decaffeination of coffee and tea;
• pH adjustment and shelf life lengthener for dairy products;
• Frozen food production; and
• Canning.
Pulp and Paper Manufacturing. CO2 is used in the pulp and paper industry in production
processes to control pH, decrease calcium levels, increase de-watering, buffer the paper
making system, and recover chemicals in the mill process.123'124
121 Engineering News-Record. 2009. New Green-Concrete Process Combines Seawater, Flue Gas, February
18, 2009, http://64.62.167.123/images/uploads/news/press-coverage/enr.pdf (accessed July 26, 2009).
122 Biello, David. 2009. Cement from C02: A Concrete Cure for Global Warming?, Scientific American,
http://www.scientificamerican.com/article.cfm?id=cement-from-carbon-dioxide (accessed July 26, 2009).
123 Linde Gas, AGA and the Pulp and Paper Industry.
http://www.aga.fi/international/web/lg/aga/like35agacom.nsf/docbyalias/nav_pulp_paper (accessed
October 25, 2010)
124 US Patent 6200416 - Recycled paper production process which incorporates carbon dioxide, March
2001 http://www.patentstorm.us/patents/6200416/description.html (accessed July 26, 2009).
90
-------
Industrial and Municipal Wastewater Treatment. Industrial and municipal wastewater
treatment facilities use C02 to control pH levels in the water and treat soft water by
dissolving lime.125
Metal Fabrication. CO2 is used for metal fabrication to enhance the hardness of casting
molds; C02 is also used during welding and cutting of metal as a shielding gas to prevent
the molten metal from oxidation.
Greenhouse Uses for Plant Growth. CO2 is used to increase plant growth in greenhouses
that produce flowers to full growth, nursery plants in early developing stages, and some
produce (e.g., tomatoes).
Fumigants and Herbicides. CO2 is used in the agricultural sector as an herbicide for
organic produce that successfully kills insects and increases storage life; and as a
fumigant for grain storage to control insects in the storage facility.
Medical Treatment. For medical treatment, CO2 is added to oxygen to stimulate breathing
and balance C02/02 levels in patients.
Construction. In the construction and manufacture industry, C02 is used on a large scale
as a shield gas in metal inert gas (MIG) welding and metal active gas (MAG) welding,
where the gas protects the weld puddle against oxidation by the surrounding air.126 A
mixture of argon and C02 is commonly used to achieve a higher welding rate and reduce
the need for post-weld treatment. Additionally, dry ice pellets are used to replace
sandblasting when removing paint from surfaces. Dry ice aids in reducing the cost of
disposal and cleanup.
Rubber and Plastic Industry. CO2 cleaning, also known as dry ice blasting, is a method of
cryogenic cleaning. Dry ice is used to remove flash from rubber objects by tumbling
them with crushed dry ice in a rotating drum. Almost all major tire manufacturers use
CO2 cleaning in their rubber molds. Shoe companies as well as gasket and other small
rubber companies have successfully been using CO2 cleaning as part of their routine
machine maintenance. Additionally, in the production of polymer foams, liquid CO2 is
used as a blowing agent instead of environmentally hazardous substances. CO2 is a
physical blowing agent with properties desired for ideal foaming, while simultaneously
fulfilling many requirements related to quality, efficiency, and the environment.
Fire Suppression. CO2 fire extinguishers are used for Class B fires (involving flammable
and volatile liquids) and Class C fires (involving electrical equipment).
Cleaning and Solvent Use. Liquid CO2 is used a solvent for cleaning a variety of products
during manufacturing processes, e.g., electronics cleaning and metal cleaning. Liquid
125 Linde Industrial Gases, Products and Supply, Wastewater, http://www.lmde-
gas.com/en/products_and_supply/water_treatment/index.html (accessed October 25, 2010).
126 Metal inert gas (MIG) welding and metal active gas (MAG) welding are two subtypes of gas metal arc
welding (GMAW).
91
-------
C02's solvent potential has been employed in some dry cleaning equipment as a
substitute for conventional solvents.
Refrigeration and Cooling. Liquid and solid C02 are important refrigerants, especially in
the food distribution industry, where they are employed during the transportation and
storage of ice cream and other frozen foods, primarily in small retail food stores. In those
cases, the CO2 may be stored on-site or be provided by refrigerant technicians at the time
of recharging.
Transportation and Storage of Explosives. CO2 is used in the transportation and storage
of explosives as it reduces the explosion risk.
Algae Production. Algae plantations for the production of biofuels have recently been
under the spotlight, primarily with respect to potential capture of C02 from electric power
and chemical production projects. The biofuel production from captured CO2 represents a
potential GHG/carbon sink. Also of strong interest is the use of algae as a source of
feedstock material for the production of biodiesel, and perhaps fermentation. CO2 is an
ingredient used by algae for normal growth, during photosynthesis. The process is being
studied at various pilot projects in the U.S. and elsewhere.
Pressurized Gas. CO2 is used as an inexpensive, nonflammable pressurized gas.
Compressed C02 gas is used in life jackets (stored inside canisters), in air guns, in
paintball markers, and for inflating bicycle tires.
Chemical processes. CO2 is currently used for various processes in the chemical industry,
including but not limited to, the production of:
• Urea (from ammonia);
• Methanol;
• Ethanol; and
• Sodium Bicarbonate.
Pharmaceutical Processes. Pharmaceutical processes may use CO2 as a less toxic
alternative to organochlorides or more traditional solvents.
92
-------
Appendix D: Suggested Outlines for MRV Plans, MRV Plan Resubmittals,
and Annual Monitoring Reports
Suggested Outline for MRV Plan and MRV Plan Resubmittals
1) Facility Information
i) Reporter number
ii) UIC permit class
iii) UIC injection well identification numbers
iv) Authorization for CO2 injection if an offshore well not subject to SDWA
i) MRV plan identification number (for resubmittals)
ii) Date most recent MRV plan approved by EPA (for resubmittals)
iii) Reason for re-submittal (for resubmittals)
2) Project Description
a) Project characteristics
i) Estimated years of CO2 injection
ii) Estimated tons C02 received over lifetime of project
b) Environmental Setting of the MMA
i) Surface and subsurface boundary of the MMA
ii) Geology and hydrogeology
iii) Historical use of subsurface and surface
iv) Available reference sites (near but outside project area for development or
adjustments to baselines)
c) Description of the Injection Process
i) Variability of C02 composition
ii) Number, location and depth of injection wells
iii) Compression/pumping, conditioning and pipelines at the facility
d) Reservoir Characterization and Modeling
i) Simulation model(s) used
ii) Modeling objectives
iii) Modeling procedures
iv) Data inputs, sources, quality control, update process
v) Model outputs
vi) Grid size and resolution
vii)Model calibration process and sensitivity analysis
3) Delineation of the monitoring areas
a) MMA
i) Determination of free phase plume extent
ii) Determination of buffer zone
b) AMA(s)
i) Initial monitoring period, area and time frame
ii) Future monitoring periods, areas and time frames
4) Evaluation of Leakage Pathways
a) Well pathway(s)
93
-------
b) Fractures, faults and bedding plan parting pathway(s)
c) Confining system pathway(s)
d) Other identified pathways(s)
5) Detection, Verification and Quantification of Leakage
a) Leakage detection methods
i) Process for detecting leakage for each pathway
ii) Performance measures for leak detection
b) Leakage Verification and Quantification Methods
i) Process for verifying and quantifying leakage for each pathway
ii) Performance measures for verifying and quantifying leakage
6) Determination of Expected Baselines
a) Monitoring method A expected baseline method
b) Monitoring method B expected baseline method
c) Monitoring method C expected baseline method
7) Site Specific Modifications to the Mass Balance Equation
a) Equipment leaks and vented emissions from surface equipment downstream of
injection flow meter
b) Equipment leaks and vented emissions from surface equipment downstream of
production well flow meter (if applicable)
c) C02 produced in oil and other fluids
8) Estimated Schedule for implementation of MRV plan
a) Timing for expected baseline determination
b) Timing of implementation of leakage detection and quantification monitoring
c) Proposed date to begin collecting data for calculating total amount sequestered
according to equation RR-11 or RR-12 of this subpart.
9) Quality Assurance Program
10) Records Retention
11) Appendices
94
-------
Suggested Outline for Annual Monitoring Report to EPA-approved MRV plans
1) Executive Summary
a) Date MRV plan approved by EPA
b) MRV plan identification number
2) Summary Table of Monitoring Activities (suggested format Table D-l)
3) Narrative History of the Monitoring Effort Conducted
a) Listing of all monitoring equipment, period of operation, and relevant tests or
surveys conducted (see also suggested table format above)
b) Cross walk of annual monitoring activities with approved MRV plan
4) Non-Material Changes to EPA-Approved MRV Plan
5) Narrative History of Monitoring Anomalies Found
a) Monitoring anomalies detected, resulting investigations, and resolutions (see also
Table D-l)
6) Description of Surface Leakage
a) Methodologies and technologies used
i) Description of alignment with EPA-approved MRV plan
ii) Non-material deviations from the leakage detection and quantification
approaches specified in the MRV plan
iii) Measurement records showing detection of monitoring anomalies
b) Assumptions Involved in Calculating the Quantity of CO2 Emitted
i) Assumption identification and supporting information
c) Uncertainties Involved in Calculating the Quantity of CO2 Emitted
Table D-l: Suggested Format for Listing 0
Monitoring Equipment
Monitoring
Equipment
Period of
Operation
Tests/Surveys
Conducted
Test/Survey
Results
Monitoring
Anomaly
Investigation
Resolutions
After
Monitoring
Anomaly
[Equipment
technology type,
make and model
number]
[Dates of
operation]
[Identification of
tests/surveys
conducted]
[Brief summary
of survey results
and identification
of any anomalies
detected]
[If anomalies
were identified, a
brief summary of
action taken to
evaluate and
investigate the
anomaly and the
related results
(i.e., if it is
concluded that
the anomaly
represents a
leak)]
[A brief
summary of the
follow-up
actions taken
after anomalies
are further
evaluated an
investigated]
Monitoring well
Number
T31NR14WS23-
IZ, Paine model
213-36-940
Pressure
Transducer
Continuous
from August
1,2011
through
December
31,2011
Hourly recording
of pressure in IZ
(a), 5020 feet
September 23-
28,2011,
increased
pressure (up to
2180 psi)
detected at this
location.
Exceeded
baseline
expectation by
21 psi
Secondary
pressure
transducer at this
location did not
read elevated
pressure.
Investigation
revealed that
vent line was
obstructed.
Vent line
obstruction
corrected on
September 29,
2011.
95
-------
Appendix E: Sampling Considerations for Designing a Monitoring Strategy
A consideration in designing a leakage detection strategy and also in establishing an
expected baseline is ensuring adequate data resolution to cover a range of points in space
and time. A non-representative data set that contains too few samples at too few times of
day/year from too few locations will generate a biased picture and result in erroneous
decision making. A sampling program may be designed to ensure consideration of the
appropriate frequency of sampling, aerial extent of sampling, and sample size.
Frequency of Sampling
If the environmental factor being measured incurs considerable intra-day and/or seasonal
variation, the sampling program may be designed to capture these fluctuations. An
example of a parameter that shows high intra-day and seasonal variation is near-surface
CO2 flux. A program based on CO2 flux monitoring would need a high frequency of
sampling across the year. On the other hand, if the environmental parameter being
monitored is not known to show a substantial intra-day or seasonal variation, the
sampling program could rely on a lower frequency of sampling. An example of such a
parameter may be fluid pressure in a deep confined reservoir to be used for ACZ
monitoring. To account for different monitoring technologies or variability in the
monitoring area, the reporter may employ different frequencies within the monitoring
area to achieve adequate baseline definition. The MRV plan should clearly define the
frequency of baseline sampling and provide a rationale for the chosen frequency.
Areal Extent of Sampling and Subdividing Maximum Monitoring Area
The specific method used for the measurement of the monitored environmental parameter
will produce values representative of a certain area. Eddy covariance towers are known to
produce flux estimates representative of a radius equal to approximately 50 times the
height of the tower.127 If cost-effective, it would thus be possible to construct as many
eddy covariance towers as necessary to ensure coverage of the entire monitoring area. On
the other hand, if the method employed produces representative values for a small range,
it may be necessary to rely on statistical techniques to balance cost and coverage.
Accumulation chambers are an example of a technique that measures a very small area
(tens of cm2). Expert judgment based on modeling results may be used to focus sampling
in higher C02 leakage risk areas and limit sampling in lower risk areas. Modeling inputs
and sensor placement should be updated periodically with the results from the monitoring
program to redefine the areal sampling plan as necessary. The MRV plan should present
the rationale for selecting the areal extent of sampling and demonstrate that it provides
adequate coverage to determine a representative baseline. If the monitored parameter is
expected to vary by subdivisions of the MMA, then the proposed subdivisions and the
process for preparing expected baselines for each should be presented in the plan.
1 27
Campbell Scientific, Inc. 2004-2006. Open Path Eddy Covariance System Operator's Manual CSAT3,
LI-7500, and KH20. Available at: http://www.campbellsci.com/documents/manuals/opecsvstem.pdf
96
-------
Sample Size
An adequate sample size will ensure greater ability to detect differences between
prevailing conditions and the expected baseline. Approximate knowledge of sampling
variation in the measured parameter (potentially based on prior sampling experience)
could be used to estimate the number of samples needed to limit the occurrence of false
positives and false negatives to the desired level. Frequently, cost considerations place
limitations on the number of samples that may be gathered such that reporters should
estimate the probability of occurrence of false positives and false negatives to better
guide decision making. For instance, if calculations show a high probability of a false
negative at a given location for a given sample size, the reporter could initiate a program
for securing more samples from that area before taking a decision. Knowledge of the
expected sampling variation in a parameter will improve with each successive sampling
phase, which in turn will improve the accuracy of the predicted sample size for the
specified error rates. The MRV plan should demonstrate that the sample size is adequate
to account for the observed variability in the measured parameter.
Coverage Adequacy of a Sampling Program
The MRV plan is likely to include one or more leakage detection or quantification
monitoring techniques that involve collecting data at one or more locations within the
MMA. This appendix explains how the statistical coverage adequacy of a given sampling
program (i.e., the probability a given size of leak will be detected) can be computed and
presented in an MRV plan.
The overall probability that a given leak will be detected is the probability that the leak
will be physically encountered by the sampling program times the probability the leak
will be recognized once encountered. In equation form this is:
P,=PxP
a e r
(Equation 1)
Where:
Pd = overall probability leak will be detected
Pe = probability leak will be encountered by a sampling program
Pr = probability leak will be recognized once encountered
The probability that the leak will be recognized after it is encountered (Pr) can be
computed given the mean and standard deviation in the expected environmental or
operating baseline, and the mean and standard deviation of the measured parameter. As
noted earlier, Pr is equal to one minus the chance of false negatives. As discussed below,
the probability of any given monitoring element encountering the leak, Pe, is a function of
the size of the monitoring area, the size and shape of the leak area, the size and shape of
the sample coverage, and the number and pattern of sampling.
97
-------
The "leak area" may be specified in terms of how much total area the expected leak will
encompass. It may also be useful to know, if possible, how many separate pieces or
segments the leak area will be in at the surface (or other monitoring zone), and what
shape (maximum length of the leak in one dimension) those segments might take. The
leak area can be estimated by hypothesizing a size of leak to be examined and estimating
the expected dispersion pattern to the surface (or to another monitoring zone such as a
shallow USDW) for a given leakage pathway. The dispersion pattern for leaks along
artificial penetrations and faults would be expected to be more easily defined, and
probably have smaller footprints compared with CZ leaks, which can meander and break
apart traveling upward. The dispersion modeling would indicate the expected area of the
leak in units such as square meters and the surface flux in units such as micrograms of
C02 per square meter per second (ng m"2 s"1). The probability of encountering the leak is
related to the leak area, while the probability of recognizing the leak once encountered is
often related to the flux.
The term "sample coverage" refers to the physical area over which the monitoring
method provides a representative measurement of the parameter. Common types of
sample coverage are: point, linear, and area. For example, a soil gas sample typically
would provide a "point sample coverage" for a limited area without a significant length
or width. A closed-path remote sensing sample consisting of a laser beam bounced off a
mirror would have a "linear sample coverage," which is an area with length but no
significant width. An eddy covariance survey would typically have an "area sample
coverage," which is a roughly circular area with a given radius. An example of the effect
of leakage footprint size on eddy covariance detection station density is presented by
Cortis et al.128
128 Cortis, A., Oldenburg, C.M., and Benson, S.M. 2008. The role of optimality in C02 seepage detection
for geological carbon sequestration sites. International Journal of Greenhouse Gas Control, Vol. 2(4)
Special Issue: Sp. Iss. SI, 640-652. doi:10.1016/j.ijggc.2008.04.008.
98
-------
|