0% found this document useful (0 votes)
61 views35 pages

Enhanced Coal Bed Methane Recovery and CO Sequestration in The Powder River Basin

Enhanced coal bed methane recovery
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
61 views35 pages

Enhanced Coal Bed Methane Recovery and CO Sequestration in The Powder River Basin

Enhanced coal bed methane recovery
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 35

INL/EXT-10-18941

Enhanced Coal Bed


Methane Recovery and
CO2 Sequestration in the
Powder River Basin

Eric P. Robertson

June 2010

The INL is a U.S. Department of Energy National Laboratory


operated by Battelle Energy Alliance
INL/EXT-10-18941

Enhanced Coal Bed Methane Recovery and CO2


Sequestration in the Powder River Basin

Eric P. Robertson

June 2010

Idaho National Laboratory


Idaho Falls, Idaho 83415

http://www.inl.gov

Prepared for the


U.S. Department of Energy
Assistant Secretary for Fossil Energy
Under DOE Idaho Operations Office
Contract DE-AC07-05ID14517
Big Sky Carbon Sequestration Partnership Phase II

Enhanced Coal Bed Methane Recovery and CO2 Sequestration in the


Powder River Basin Deliverable Gd10
Eric P. Robertson, Idaho National Laboratory

June 2010
U.S. Department of Energy (DOE)
National Energy Technology Laboratory (NETL)
DOE Award Number: DE-FC26-05NT42587
INL/EXT-10-18941

Submitted to: William Aljoe, DOE Project Officer


Julie Heynes, DOE Administrator

Submitted by: Dr. Lee Spangler


BSCSP Principal Investigator and Director
Montana State University
P.O. Box 173905
Bozeman, MT 59717
[email protected]
http://www.bigskyco2.org
Phone: (406) 994-1658
Fax: (406) 994-3745
Big Sky Carbon Sequestration Partnership Phase II

Enhanced Coal Bed Methane Recovery and CO2 Sequestration in the Powder River Basin

Eric P. Robertson
Idaho National Laboratory

DOE Award Number: DE-FC26-05NT42587


Montana State University
P.O. Box 173905
Bozeman, MT 59717-3905
Abstract

Unminable coal beds are potentially large storage reservoirs for the sequestration of
anthropogenic CO2 and offer the benefit of enhanced methane production, which can offset
some of the costs associated with CO2 sequestration. The objective of this report is to
provide a final topical report on enhanced coal bed methane recovery and CO2 sequestration
to the U.S. Department of Energy in fulfillment of a Big Sky Carbon Sequestration
Partnership milestone.
This report summarizes work done at Idaho National Laboratory in support of Phase II of the
Big Sky Carbon Sequestration Partnership. Research that elucidates the interaction of CO2
and coal is discussed with work centering on the Powder River Basin of Wyoming and
Montana. Sorption-induced strain, also commonly referred to as coal swelling/shrinkage,
was investigated. A new method of obtaining sorption-induced strain was developed that
greatly decreases the time necessary for data collection and increases the reliability of the
strain data. As coal permeability is a strong function of sorption-induced strain, common
permeability models were used to fit measured permeability data, but were found inadequate.
A new permeability model was developed that can be directly applied to coal permeability
data obtained under laboratory stress conditions, which are different than field stress
conditions. The model can be used to obtain critical coal parameters that can be applied in
field models.
An economic feasibility study of CO2 sequestration in unminable coal seams in the Powder
River Basin of Wyoming was done. Economic analyses of CO2 injection options are
compared. Results show that injecting flue gas to recover methane from CBM fields is
marginally economical; however, this method will not significantly contribute to the need to
sequester large quantities of CO2. Separating CO2 from flue gas and injecting it into the
unminable coal zones of the Powder River Basin seam is currently uneconomical, but can
effectively sequester over 86,000 tons (78,200 Mg) of CO2 per acre while recovering methane
to offset costs. The cost to separate CO2 from flue gas was identified as the major cost driver
associated with CO2 sequestration in unminable coal seams. Improvements in separations
technology alone are unlikely to drive costs low enough for CO2 sequestration in unminable
coal seams in the Powder River Basin to become economically viable. Breakthroughs in
separations technology could aid the economics, but in the Powder River Basin, they cannot
achieve the necessary cost reductions for breakeven economics without incentives.

Disclaimer

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

ii
CONTENTS
ACRONYMS .............................................................................................................................................. vii

1. Introduction ........................................................................................................................................ 1

2. Study goals and objectives ................................................................................................................. 2

3. Measurement of sorption-induced strain in coal ................................................................................ 3


3.1 Coal collection ......................................................................................................................... 3
3.2 Axial strain cell description ..................................................................................................... 3
3.3 Sample preparation for use in axial strain cell apparatus ......................................................... 6
3.4 Use of equipment and collection of appropriate data ............................................................... 6
3.4.1 Locating reference end points of samples ................................................................... 6
3.4.2 Measuring the distance between sample end and reference mark .............................. 7
3.4.3 Temperature control .................................................................................................... 7
3.4.4 Initialization pressure cycles ....................................................................................... 8
3.5 Data reduction: working with strain versus time data .............................................................. 8
3.6 Plotting strain versus pressure curves ...................................................................................... 9

4. Modeling sorption-induced coal permeability changes .................................................................... 10


4.1 Measurement of permeability data ......................................................................................... 10
4.2 Comparison of permeability data to common models ........................................................... 10
4.3 Development of new permeability model for laboratory conditions ..................................... 12
4.3.1 Discussion of new model .......................................................................................... 13
4.3.2 Use of model as tool to predict key coal properties .................................................. 13
4.3.3 Relative contribution of terms in permeability model .............................................. 13
4.3.4 Summary of new permeability model ....................................................................... 14

5. Economic evaluation ........................................................................................................................ 14


5.1 Approach ................................................................................................................................ 16
5.2 Economic model description .................................................................................................. 17
5.2.1 Forecast of fluid production rates ............................................................................. 17
5.2.2 Mineral rights ............................................................................................................ 18
5.2.3 Injection and production wells .................................................................................. 18
5.2.4 Produced water disposal costs................................................................................... 18
5.2.5 Transportation costs for injection-gas ....................................................................... 19
5.2.6 CO2 separation costs ................................................................................................. 19
5.2.7 Costs to separate nitrogen from produced gas .......................................................... 19
5.2.8 Other input variables ................................................................................................. 19
5.3 Results and discussion ........................................................................................................... 19
5.4 Deterministic comparisons ..................................................................................................... 20
5.5 Monte Carlo analysis of CO2 injection scenario .................................................................... 20
5.5.1 Sensitivity analysis.................................................................................................... 21
5.5.2 Total CO2 cost to sequestration projects ................................................................... 21
5.6 Summary of economic evaluation .......................................................................................... 22

iii
6. References ........................................................................................................................................ 23

FIGURES
Figure 1. Map of Powder River Basin showing geographic location to INL in Idaho Falls
and surrounding basin borders. ................................................................................................ 2
Figure 2. Photograph of the axial strain pressure cell without the temperature contol and
accompanying insulation applied. The micrometer microscope is attached and a
reflected light can be seen shining from behind through the sample beds cut into
the 1/4-inch rod. .......................................................................................................................... 5
Figure 3. Components of the digital filar microscope used for measuring sorption-induced
strain in coal. ............................................................................................................................... 7
Figure 4. Average axial strain changes with respect to time for five coal samples from the
Trapper mine in Colorado under various CO2 gas pressures at 80 F. ................................. 8
Figure 5. Axial strain for coal from the Trapper mine in Colorado for various gases as a
function of pressure at a constant temperature of 80 F. For each of the gas
datasets, the data are fitted with the Langmuir equation. ...................................................... 9
Figure 6. Photograph of trimmed ends from Gilson-seam coal cores 2 in. in diameter. Note
the irregular cleat system associated with this coal. .............................................................. 10
Figure 7. Permeability as a function of pore pressure for three different pure gases for (a)
Anderson coal and (b) Gilson coal. ......................................................................................... 11
Figure 8. Model comparisons for two coal cores and three flowing gases. For all
experiments, confining pressure was 1000 psia and temperature was 80F. ...................... 12
Figure 9. Plots showing the relative contributions of model terms to model output. ........................ 13
Figure 10. Wyodak-Anderson net coal isopach map. The area encompassed by the white
line represents the approximate unminable coal area overlain by over 1000 ft
(304.8 m) of overburden. .......................................................................................................... 15
Figure 11. Gas capacity isotherms for carbon dioxide, methane, and nitrogen for the
Wyodak-Anderson coal seam used in the reservoir simulation. .......................................... 17
Figure 12. Methane production curves for the three scenarios analyzed. .......................................... 18
Figure 13. Probabilistic net present value (NPV10) of the CO2 injection scenario. ............................ 21
Figure 14. Correlation coefficients for the five input variables with the most impact on the
net present value of the CO2 injection scenario. .................................................................... 22
Figure 15. Distribution and mean value of the cost of CO2 separation/capture required to
yield a 10% rate of return. ....................................................................................................... 23

TABLES
Table 1. Properties of coal samples collected and used in this research as ascertained from
various analyses on an as received basis. .............................................................................. 4
Table 2. Wyoming coal-fired power plants. ........................................................................................... 14
Table 3. Flue gas composition of Wyodak PC power plant.................................................................. 15

iv
Table 4. Base values of input parameters used to compare economics of carbon dioxide
sequestration scenarios. ............................................................................................................ 16
Table 5. Key parameters used for simulation of CO2-sequestration in PRB unminable coal
seams. ......................................................................................................................................... 17
Table 6. Deterministic economic results using most likely values for input parameters for
three scenarios analyzed........................................................................................................... 20

v
vi
ACRONYMS
CBM coal bed methane
CH4 methane
CO2 carbon dioxide
GHG green house gas
PC pulverized coal
N2 nitrogen
ECBM enhanced coal bed methane
PRB Powder River Basin
RECOPOL a CO2 sequestration in coal project in Poland
INL Idaho National Laboratory
O&M operation and maintenance costs
NPV10 net present value at a discount rate of 10%
ROI return on investment
GOIP gas originally in place
OIP originally in place

vii
Error! Reference source not found.
1. Introduction
Sequestration of greenhouse gases (GHG) will likely be required in order to mitigate the
climatological effects of rising GHG concentrations in the atmosphere. Carbon dioxide (CO2) is a
commonly emitted GHG resulting from the combustion of fuels such as coal, natural gas, oil, wood, and
other similar fuels. Retrofitting stationary point sources such as pulverized coal (PC) power plants with
technology to separate and capture CO2 from flue gas emissions would allow the produced CO2 to be
collected and sequestered. In 2006, the U.S. generated 49% of its total electricity from coal-fired power
plants (U.S. DOE, 2007a). These plants emitted roughly 1.91 Gt CO2 in the same year (U.S. DOE,
2007b).
Coal seams have the capacity to adsorb large amounts of gases (especially carbon dioxide) because of
their typically large internal surface area (30 to 300 m2/g) (Berkowitz, 1985). Some gases, such as CO2,
have a higher affinity for the coal surfaces than others, such as nitrogen (N2). Knowledge of how the
adsorption or desorption of gases affects coal permeability is important, not only to operations involving
the production of natural gas from coal beds but also to the design and operations of projects to sequester
greenhouse gases in coal beds.
Injecting CO2 into unminable coal seams has been proposed as a method to permanently sequester
anthropogenic CO2 emissions (Mathews et al., 2001; Steinberg, 2001; Schroeder et al., 2002; Zutshi and
Harpalani, 2004; White et al., 2005). In addition to being potential sinks for CO2 sequestration, many
coal beds contain commercial quantities of adsorbed natural gas. Injecting CO2 has been proposed as a
method to enhance the production of methane from coalbed methane (CBM) operations in a process
called CO2-enhanced coal bed methane (CO2-ECBM) (Puri and Yee, 1990; Pekot and Reeves, 2003;
Busch et al., 2003; Prusty and Harpalani, 2004). The produced methane would create revenue that can
offset the costs associated with the injection and sequestration of CO2 in coal beds. In the CO2-ECBM
process, CO2 is used to displace the adsorbed methane molecules and increase methane production
without lowering reservoir pressure. As carbon dioxide is injected into a coal seam containing methane,
the CO2 molecules compete with the methane molecules for adsorption sites. Methane molecules detach
from adsorption sites as CO2 is injected into the coal due to a decrease in methane partial pressure in the
free gas phase. The displaced methane (CH4) is then free to flow to a production well while the injected
CO2, which has a greater adsorption affinity than CH4, is adsorbed onto the coal.
Fluid movement in coal is described by Darcys Law as the gases flow through the natural fracture
(cleat) system of the coal. Adsorption of gases by the internal surfaces of coal causes the coal matrix to
swell and desorption of gases causes the coal matrix to shrink. The swelling or shrinkage of coal as gas is
adsorbed or desorbed is referred to as sorption-induced strain. Sorption-induced strain causes the width
of the coal cleats to change, thus changing the permeability of the coal. Modeling changes in coal
permeability as a function of gas pressure has been the object of a number of papers (Gray, 1987; Sawyer
et al. 1990; Seidle and Huitt 1995; Palmer and Mansoori 1998; Pekot and Reeves 2003; Shi and Durucan
2003), but modeling permeability changes in coal remains problematic (Robertson and Christiansen
2007).
Worldwide CO2 storage capacity estimates in unminable coal seams range from 272 Gt CO2 (Hen-
driks et al., 2004) to 350 Gt CO2 (Stevens et al., 1998). Reeves (2003) estimated that the CO2 storage
capacity of unminable coal seams in the U.S. is 90 Gt CO2 and he also estimated that unminable coal
seams within the Powder River Basin of Wyoming and Montana are capable of storing 14 Gt CO2. Ac-
cording to these estimates of current output and storage capacity, the unminable coal seams within the
U.S. can provide 47 years of storage capacity for PC power plant CO2 emissions, a significant potential
storage medium for CO2 emissions.

1
There have been a handful of field tests demonstrating the technical feasibility of using unminable
coal beds to sequester CO2 and stimulate the production of methane (White et al., 2005). These include
the Tiffany unit in southern Colorado (Liang et al., 2003), the Allison unit in northern New Mexico
(Reeves et al., 2003), the Medicine River coal seam in Alberta (Mavor et al., 2004), and the RECOPOL
field project in Poland (van Bergen et al., 2003). These tests were not designed to evaluate the economic
viability of sequestering CO2 emitted from large anthropogenic point sources such a PC power plants. An
economic analysis of CO2 injection into unminable coal seams that accounts for the added benefit of
increased methane production to help offset the cost of sequestering CO2 emitted from PC power plants is
needed.
2. Study goals and objectives
The objective of this report is describe work at Idaho National Laboratory to study the interaction of
CO2 and coal and the applicability of using CO2 to enhance study the economic feasibility of CO2
sequestration in unminable coal seams in the regional setting of the Powder River Basin (PRB) of
Wyoming and southern Montana (Figure 1). This specific setting was selected because of its large CO2
sequestration potential, its national importance as the largest coal-producing basin in the United States,
and its location within the Big Sky Regional Carbon Sequestration Partnership area.

Figure 1. Map of Powder River Basin showing geographic location to INL in Idaho
Falls and surrounding basin borders.

The Powder River Basin is located largely in northeastern Wyoming with a small portion extending
into southeastern Montana (see Figure 1). The basin is a deep, northerly trending, asymmetric, mildly

2
deformed trough, approximately 400 km long and 160 km wide. Its axis is close to its western margin,
which is defined by the Bighorn Mountains uplift and the Casper arch. It is bordered on the south by the
Laramie and Hartville uplifts and on the east by the Black Hills uplift. The northern margin is defined by
the subtle northwest trending Miles City arch.
3. Measurement of sorption-induced strain in coal
Sorption-induced strain can contribute up to 60% of the total change in coal permeability that occurs
during the production of methane from coal beds (Robertson and Christiansen, 2008). Sorption of CO2
can account for even more of the total permeability change during CO2 sequestration or CO2-ECBM
operations. Published sorption-induced strain data is very sparse because the common method of
collecting such data has been very cumbersome and time-consuming. INL pioneered the development of
a new technique that is much quicker and reliable than the old method that obtained strain data by
attaching strain gauges onto large blocks of coal.
A detailed methodology for taking the sorption-induced strain measurements is described in this
section. The methodology is divided into 4 sections: 1) preparing coal samples and equipment, 2) use of
equipment and the steps used to collect the appropriate data, 3) data reduction, and 4) plotting strain
versus pressure curves.
3.1 Coal collection
A number of large coal blocks, roughly 1 ft3 in size, were collected from coal mines in Utah and
Wyoming. High-volatile bituminous coal from the Uinta-Peceance Basin was collected from the Gibson
seam of the Book Cliffs coal field from an underground mine near Price, Utah. Subbituminous, low-
contaminant coal from the Powder River basin was collected from the Anderson seam from an open-pit
mine near Gillette, Wyoming. At the mine location, the Anderson seam was more than 100 feet thick.
Proximate, ultimate, and heating-value analyses were subsequently performed on samples of the collected
coal and are shown in Table 1.
Because most coals are sensitive to oxidation and because exposure of freshly mined coal to air at
ambient temperatures for as little as a few days can adversely affect some properties such as heating value
and tar yield (Berkowitz, 1994), care was exercised during the collection of the coal to limit exposure to
air and to ensure that the samples remained as pristine as possible by following sampling procedures
recommended by Gash (1991).
The Utah coal was taken from the conveyer belt carrying recently mined coal (less than one minute
after contact by continuous miner machinery) as close to the active mine face as possible to limit oxygen
exposure. Immediately after being taken from the conveyer, each sample was double wrapped in plastic
bags and sealed by tape. Transporting the sample from the mine face to the surface via vehicle took from
5 to 20 minutes depending on the collection site. Upon reaching the surface, the samples were removed
from the bags and placed under de-ionized water inside sealed containers for transport to INL facilities in
Idaho Falls, Idaho. The Wyoming coal was collected from recently exposed walls (less than one day)
from an open-pit mine where large boulders were broken open by hand to expose fresh coal. Smaller
samples (roughly 1 ft3 in size) of fresh coal were taken from these large boulders and immediately placed
under water in sealed containers for transport to Idaho Falls, Idaho. Once at INL, the blocks were kept
under water until needed.
3.2 Axial strain cell description
Previously used approaches for measuring coal strain are very time consuming and the resulting data
can be difficult to interpret (Harpalani and Schraufnagel, 1990; Seidle and Huitt, 1995). Consequently,
the data needed for understanding the change of coal permeability due to gas sorption are scarce. This
section details INLs development of an optical approach for measuring the strain of coal during gas
sorption tests. With an optical approach, challenges of attaching strain gauges and passing electrical

3
connections to a high-pressure chamber are entirely avoided. Furthermore, the size of the sample can be
adjusted to minimize equilibration time, and multiple samples can be easily tested simultaneously. With
this optical approach, we should be able to gather a suitable suite of strain data for testing permeability
correlations.

Table 1. Properties of coal samples collected and used in this research as ascertained
from various analyses on an as received basis.
Anderson Gilson
Location of seam Wyoming Utah
High-volatile
Coal rank Subbituminous
bituminous

Proximate Analysis wt%:


Moisture 26.60 7.52
Ash 6.18 2.99
Volatile Matter 30.99 37.42
Fixed Carbon 36.23 52.07
Total 100.00 100.00

Ultimate Analysis wt%:


Moisture 26.60 7.52
Hydrogen 2.08 3.86
Carbon 50.57 71.66
Nitrogen 0.43 1.36
Sulfur 0.27 0.49
Oxygen 13.87 12.12
Ash 6.18 2.99
Total 100.00 100.00

Heating Value, Btu/lb


Measured 8,514 12,437

Vitrinite Reflectance, % 0.24 0.53

Figure 2 is a photograph of the axial strain vessel developed at INL (Robertson and Christiansen,
2007 (a) and (b)). Based upon the problems encountered using the proof-of-principle apparatus, a
modified strain measurement apparatus was constructed. The main components of this refined system
included a Jerguson cell, a removable multi-sample holder, and a digital filar microscope mounted
directly to the pressure cell. The Jerguson cell was modified slightly by drilling-out the entrance and exit
ports to allow for the passage of a -inch rod. The multi-sample holder was fabricated out of a -inch
stainless steel rod and is easily removable from the pressure cell when samples need to be changed. A
digital filar microscope manufactured by Gaertner Scientific, Inc. was mounted directly onto the pressure
cell to more accurately measure changes in coal strain under different pressures and gas compositions.
Mounting the microscope directly to the pressure cell eliminated the vibrational problems encountered
with the proof-of-principle apparatus.


4

Figure 2. Photograph of the axial strain pressure cell
without the temperature contol and accompanying
insulation applied. The micrometer microscope is attached
and a reflected light can be seen shining from behind
through the sample beds cut into the 1/4-inch rod.


The rod holding the samples was centered in the pressure vessel by passing it through both the inlet
and outlet ports of the pressure cell. The test gas could enter and exit the vessel through holes drilled
through the center of the rod. A thermocouple was placed into the pressure vessel through the gas ports in
the rod. Temperature in the vessel can be controlled with input from the thermocouple.
Six coal samples can be tested at the same time with this apparatus. The size of the samples is
constrained by the dimensions of the sample beds machined into the rod. The maximum size is 1-inch
in length by 5/32-inch in width by 1/8-inch in height. The sample beds have a solid bottom on which the
coal samples rest except for the upper portion, which is machined completely through the rod to allow
light to shine up from below and illuminate the coal sample.
A filar micrometer microscope with electronic readout was mounted directly to the Jerguson pressure
cell to eliminate vibrations and to more accurately measure small movements. The microscope has a
cross hair that moves across the field of view by means of a precision micrometer screw and reads to
0.0001mm with a 2.5mm range, which will accommodate up to a 10% elongation of the coal sample. For
a sample of one inch in length, a 0.5% growth equates to an actual growth of 0.127mm, well within the
specifications of the electronic readout.

5
A microscope holder was designed and fabricated that clamps directly to the pressure vessel. The
microscope was mounted to the holding device; and by means of a hand-adjusted rotating screw, the
microscope can be precisely moved along the entire length of the cell window. In this manner, multiple
coal samples can be monitored during the same experiment. In Figure 2, one can see a number of the
sample beds along with the microscope mounted onto the pressure cell.
3.3 Sample preparation for use in axial strain cell apparatus
The small samples used to measure strain were taken from these larger blocks using a rock saw
cooled by deionized water. Coal pieces that fit in the machined beds must be cut and clipped from fresh
coal chunks. This is done by first cutting thin rectangular strips of coal inches tall by  inches wide by
however long they can be cut using a water cooled tile saw.
Once these strips are cut, smaller match-stick-size pieces are clipped so that they fit into the beds
machined into a -inch diameter solid stainless steel rod that is inserted into the axial strain vessel. The
steel rod holding the coal samples can be seen inside the axial strain vessel in Figure 2. A pliers-like,
hand-held tool is used for clipping these sample pieces. A control sample (cut from -inch stainless steel
tubing makes an excellent control sample) usually occupies the lower-most sample bed and is used as a
quality control sample. No measureable strain should be seen in the control.
Before the samples are inserted into the cell but after it is known that they are all the right size to fit
into the small opening of the axial strain pressure vessel, the length of each sample is directly measured
using a caliper and recorded in a laboratory notebook. Once the coal samples are loaded in the axial
strain pressure vessel and before any other action (including temperature modification), measurements are
collected on each of the samples to identify their initial end location with respect to a standard measuring
mark.
A clear plastic strip with periodic marks about 1 mm apart is used as a standard against which sample
strain are measured. The standard strip is laid on the outside of the pressure vessel directly above the coal
samples on the sight glass and firmly secured to prevent subsequent movement. This strip is called the
measurement standard.
After the measurement standard is placed securely on the outside of the axial strain vessel and the
samples (e.g. coal, stainless steel control, etc.) are placed within the vessel, the axial strain pressure vessel
should be tilted to a fairly high angle about 70 degrees from horizontal to allow gravity to keep the
lower end seated in the sample bed while keeping the upper end free to move when conditions are
induced to cause strain.
3.4 Use of equipment and collection of appropriate data
A strain versus pressure curve is obtained by collecting sample length data at various gas pressures,
converting the length data into strain, and then plotting strain versus gas pressure. Gases of interest
include helium, nitrogen, methane, and carbon dioxide (others may be used also). Data points are
collected at zero pressure (vacuum) and at a minimum of four other pressure values somewhat evenly
spaced over the pressure range desired for the strain versus pressure curve.
3.4.1 Locating reference end points of samples
At this point, the location of each sample with respect to certain marks on the measurement standard
is measured. This assumes that the filar microscope has already been mounted correctly to the pressure
vessel. The microscope is focused on the free end of a sample, beginning with the lower-most sample.
The crosshairs of the filar microscope are placed at the end of the sample by adjusting the horizontal
adjuster (see Figure 3 for components of the filar microscope).

6
Horizontal DC power
Eyepiece crosshairs source
adjuster

Vertical
microscope
adjuster

Microscope Digital Readout

Figure 3. Components of the digital filar microscope used for


measuring sorption-induced strain in coal.

The ends of the samples will not be smooth when seen through the microscope,; they will be irregular
or jagged. Prominent points on each of the samples should be noted as reference end points. Drawing
this reference end point for each sample in the laboratory notebook is a helpful way to remember it. The
same reference point must be located and used repeatedly when measuring the strain throughout the strain
experiments.
3.4.2 Measuring the distance between sample end and reference mark
The digital readout is zeroed when the crosshairs are on the reference end point of the sample. The
microscope is then adjusted vertically (perpendicular to the plane of the direction of strain measurement)
until it is refocused on the measurement standard on the outside of the pressure vessel.
The crosshairs are then adjusted until they are placed on the nearest mark of the measurement
standard. This standard mark will be used for all subsequent measurements associated with this sample.
The standard mark will be wavy, but a specific wave must be noted and used repeatedly for each
sample. As with the sample reference end point, a drawing of the standard mark wave in the laboratory
notebook is helpful in finding the point again.
The distance in millimeters between the end of the sample and the standard mark is read from the
microscopes digital readout and recorded in laboratory notebook. This distance will be used to correlate
subsequent changes in sample length to the original length measured by the calipers before the sample
was put into the axial strain cell.
Once one of the sample measurements is taken, the entire microscope assembly is moved to the next
sample by turning a threaded rod until the sample end comes into view through the microscope.
3.4.3 Temperature control
Strain measurements must be done at a constant temperature. Normal fluctuations in room
temperature can cause significant errors in the collected data if temperature is not controlled. To control
the temperature within the axial strain pressure cell, the cell is wrapped with an electrical heating strip and
then with insulation. A thermocouple inserted into the axial strain pressure cell is used to record and

7
control the current in situ temperature. A temperature setting of 80 F has been found to be high enough
over room temperature of around 70 F to overcome the effects of room temperature fluctuations. About
6 hours are required to stabilize the temperature inside the axial strain pressure cell from room
temperature up to 80.0 F.
3.4.4 Initialization pressure cycles
After the temperature has stabilized, the end location of each sample is read and recorded. Following
this data reading, a hard vacuum is applied to the pressure chamber for 24 hours followed by pressurizing
the chamber with helium. Strain measurements should be taken during these pressure cycles, which are
repeated until constant strain values are obtained for each pressure state. When beginning a test with a
different gas, a hard vacuum is applied to the samples for 24 hours and then pressurized with helium as a
precautionary purge for 24 hours followed by another 24-hour vacuum period.
3.5 Data reduction: working with strain versus time data
After the strain data has been collected for each of the samples for a given gas, the strain versus
pressure curve can be constructed. When the data are inputted into specially designed Excel
spreadsheets, two plots are created: one plots the strain equilibration for each sample and the other plots
the average strain versus time for all the samples. Figure 4 shows data for the average strain versus time
for different adsorption pressures for methane and coal from the Trapper mine.

2
Coal from Trapper mine
(CO2 pressure at 80 F)
Axial strain, % of original length

1.6

1.2

Langmuir-type fit
Measured data
0.8

0.4
140 psig 330 psig 550 psig 830 psig vacuum
vacuum
0

0 20 40 60 80 100 120 140


Equilibration time, hr
Figure 4. Average axial strain changes with respect to time
for five coal samples from the Trapper mine in Colorado
under various CO2 gas pressures at 80 F.

The measured data shown in this figure are shown as open circles, while the solid lines are best fits
of the data, fit separately for each pressure regime. The best fit lines were obtained using a least squares
subroutine for a modified form of the Langmuir equation that accounts for non-zero starting points:
SL t  t1
S S1  ,
t L  t  t1
where S is the calculated strain, S1 is the initial strain at the beginning of the pressure regime, SL is the
Langmuir strain, t is the time in hours, t1 is the time in hours at the beginning of the pressure regime, and

8
tL is the Langmuir time. Both SL and tL were obtained by fitting the data using least-squares analysis with
a Langmuir curve when both S1 and t1 are zero.
3.6 Plotting strain versus pressure curves
Langmuir parameters for the strain versus pressure curve are obtained by fitting the end point values
of the data in Figure 4 for each pressure regime using the following Langmuir-type equation.
p
S SL ,
p  pSL

where S is the freestanding axial strain, SL is the Langmuir strain, p is the gas pressure, and pSL is the
Langmuir strain pressure.
Figure 5 shows an example of the axial strain for coal from the Trapper mine in Colorado for various
gases as a function of pressure at a constant temperature of 80 F. The calculated data shown in this
figure are shown as open circles, while the solid lines are best fits of the data obtained using a least-
squares approach with the above Langmuir-type equation.

0.02
Coal from Trapper mine
(80 F)
CO2
Axial strain, fraction of original length

0.016

0.012

0.008

CH4
0.004

N2
0
He

-0.004

0 200 400 600 800 1000 1200


Gas pressure, psia

Figure 5. Axial strain for coal from the Trapper mine in


Colorado for various gases as a function of pressure at a constant
temperature of 80 F. For each of the gas datasets, the data are
fitted with the Langmuir equation.

Description of the procedures and equipment used to obtain free standing, axial strain, sorption-
induced strain was included in this report in order to show the rigor used during data acquisition.
Obtaining the raw data necessary to construct the relationships shown in Figure 5 requires about five
weeks. Constructing this same plot from data obtained using traditional strain gauges would require
about one year due to the long stabilization times associated with the required larger samples. These data
are necessary to accurately model permeability changes in coal due to sorption (desorption and/or
adsorption) of gases that occurs during geologic CO2 sequestration activities.

9
4. Modeling sorption-induced coal permeability changes
All permeability measurements were made in INL laboratories at 80F with gas as the flowing fluid
using 2-in. diameter cores cut parallel to the bedding planes. The cores were cut from the larger blocks of
coal using deionized water as the cooling/lubricating fluid. Some of the cores fell apart, either while
cutting the cores or while attempting to remove them from the core bit, but persistence and care while
drilling resulted in many good cores to use for the flow tests. The cores were stored in small, sealed
containers under deionized water until used in the experiments. Figure 6 is a photograph of the trimmed
ends of some of the Gilson-seam cores and is shown to give the reader an idea of a typical cleat system
associated with these cores.




Figure 6. Photograph of trimmed ends from Gilson-seam
coal cores 2 in. in diameter. Note the irregular cleat
system associated with this coal.

A hydrostatic-type core holder was used for the permeability experiments. The flowing gas was
supplied by pressurized gas cylinders, and the flow rate was controlled by adjusting both the cylinder
regulator and a metering valve upstream of the core holder. A backpressure regulator was used to apply a
minimum of 100 psia of pore pressure to minimize the Klinkenberg effect of gas-permeability
measurements. Upstream and differential pressures were made using pressure transducers. A flow meter
was placed directly downstream of the backpressure regulator.
4.1 Measurement of permeability data
A description of the experimental procedure used to collect permeability data is discussed by
Robertson and Christiansen (2007). Using the permeability apparatus, permeability was measured using
nitrogen, methane, and carbon dioxide as the flowing fluids for two coals described in Table 1. The
results of these permeability tests are shown in Figure 7 (a) and (b).
4.2 Comparison of permeability data to common models
Being able to model and accurately predict permeability changes is an important and vital capability
that is necessary to the proper design of CO2-ECBM and for CO2 sequestration in unmineable coal seams.
Three well-known permeability models were selected from the literature to model the laboratory
permeability data (see Figure 7): Seidle and Huitt (1995), Palmer and Mansoori (1998), and Shi and
Durucan (2003).


10
2.0
Gas k 0 (mD) Anderson Coal
N2 257 (Subbituminous)
T = 80o F
1.6 CH 4 147 Pob = 1000 psia

Permeabiity ratio (k/k0)


CO2 86

1.2

0.8

0.4

0.0
0 200 400 600 800 1000
Average pore pressure, psia 
(a)

3.0
Gas k0 (mD)
Gilson Coal
N2 0.0372 (H.V. Bituminous)
CH 4 0.0289 T = 80 oF
Permeabiity ratio (k/k0)

Pob = 1000 psia


CO 2 0.0226
2.0

1.0

0.0
0 200 400 600 800 1000
Average pore pressure, psia 
(b)

Figure 7. Permeability as a function of pore pressure for three
different pure gases for (a) Anderson coal and (b) Gilson coal.


Figure 8 shows the three models applied to the measured permeability data. None of the three coal
permeability models did a very good job matching actual permeability data for all six cases. However,
two of the models (Shi-Durucan and Palmer-Mansoori) did relatively well matching the nitrogen
permeability data.
It should be noted that the laboratory permeability measurements were made under hydrostatic
conditions where the confining pressure was equal in all directions, whereas the models selected to match
that data were derived for matchstick-type fractured geometry under uniaxial strain conditions, thought to

11
more closely describe field conditions. Even though the models might not be totally applicable to
laboratory permeability measurements, they are applied in this work to define a starting point for potential
modifications or improvements. If one could develop a model to accurately predict laboratory
permeability changes in sorptive-elastic material, such as coal, the model could be used as a tool to
determine coal properties of use in field simulations.

Shi-Durucan model Palmer-Mansoori model Seidle-Huitt model

2 2 2
Flowing fluid = carbon dioxide Flowing fluid = methane Flowing fluid = nitrogen
Core = Anderson 01 Core = Anderson 01 Core = Anderson 01
1.6 1.6 1.6
Permeability ratio (k/k0)

Permeability ratio (k/k0)

Permeability ratio (k/k0)


1.2 1.2 1.2

0.8 0.8 0.8

0.4 0.4 0.4

0 0 0
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Average pore pressure, psia Average pore pressure, psia Average pore pressure, psia

2 2.5 2.5
Flowing fluid = carbon dioxide Flowing fluid = methane Flowing fluid = nitrogen
Core = Gilson 02 Core = Gilson 02 Core = Gilson 02
1.6 2 2
Permeability ratio (k/k0)

Permeability ratio (k/k0)


Permeability ratio (k/k0)

1.2 1.5 1.5

0.8 1 1

0.4 0.5 0.5

0 0 0
0 200 400 600 800 1000 0 200 400 600 800 1000 0 200 400 600 800 1000
Average pore pressure, psia Average pore pressure, psia Average pore pressure, psia


Figure 8. Model comparisons for two coal cores and three flowing gases. For all experiments, confining
pressure was 1000 psia and temperature was 80F.

4.3 Development of new permeability model for laboratory


conditions
A new model was derived (Robertson and Christiansen, 2008) that describes the permeability
behavior of a fractured, sorptive-elastic media, such as coal, under typical laboratory conditions where
common radial and axial pressures are applied to a core sample during permeability measurements. The
new model can be applied to fractured rock formations where the matrix blocks do not contribute to the
porosity nor to the permeability of the overall system, but where adsorption and desorption of gases by
the matrix blocks cause measurable swelling and shrinkage and thus affect permeability. The derived
model is represented by the following equation:
D pp  pp0
3 1 2Q pL  pp
1 e
3 c0  S p
pp  pp0  max L ln


k D I0 E

pL  pp0 pL  pp0

e ,
k0

where k is measured permeability, ko is the initial permeability of the coal core, c0 is the initial fracture
compressibility,  is the fracture compressibility change rate, pp is the pore pressure, pp0 is the initial pore
pressure, I0 is the initial porosity, Q is Poissons ratio, E is Youngs modulus, SL is the Langmuir strain,
and pL is the Langmuir pressure.

12
4.3.1 Discussion of new model
Understanding the dynamics of the physics involved with changes in permeability in sorptive-elastic
media, such as coal, is of vital interest to those involved with optimizing production from coal bed natural
gas fields, sequestering carbon dioxide in coal beds, or producing natural gas from some shale formations.
Laboratory experiments can be designed to enlighten the engineer as to what processes contribute to
project success and how they can best be manipulated to increase recovery or economic viability. The
equations derived above are based on the conditions encountered in laboratory experiments designed to
calculate permeability. An accurate understanding of how permeability can change during production and
injection operations is very important. This present model should help researchers to better understand
the processes that influence permeability and to derive realistic values of important parameters, such as
fracture compressibility, initial porosity, and elastic mechanical moduli, that need to be included in field-
wide reservoir simulations.
4.3.2 Use of model as tool to predict key coal properties
This model has been appropriately derived for conditions frequently used in laboratory-measured
permeability test on coal samples. Key, hard-to-measure coal properties can be determined by fitting
laboratory-measured permeability data with this model. For example, Youngs modulus and Poissons
ratio are important parameters needed to accurately forecast permeability behavior during coalbed
methane operations; yet, these values are difficult to measure and general values are typically used. If
initial porosity, fracture compressibility, and sorption-induced strain were all fairly well known, the
elastic moduli (Youngs modulus and Poissons ratio) could be determined by varying their values until
the model reached a reasonable fit of the permeability data. These model-determined values could then
be used in field-scale models to improve permeability forecasts.
4.3.3 Relative contribution of terms in permeability model
The permeability model given above contains three terms representing the contribution of fracture
compressibility, mechanical elastic matrix strain, and sorption-induced matrix strain to fracture width and
permeability change. One might suspect that one factor could be more important to permeability change
than the others or that one of the three could be a relatively small contributor to permeability change.
Figure 9 (a) shows the change in permeability ratio caused by contribution of the three components of the
permeability model and Figure 9 (b) shows the relative contribution of each of the terms to the total
permeability ratio calculated by the model.

8 1.2
Relative contribution of terms in model, fraction

Methane is flowing fluid


for both (a) and (b) plots
1
6 Total contributions
Permeability ratio, k/k 0

Fracture compressibility 0.8


Mechanical-elastic properties
Sorption-induced strain
4 0.6

0.4
2
0.2

0 0

0 200 400 600 800 1000 1200 0 200 400 600 800 1000 1200
Average pore pressure, psia Average pore pressure, psia
(a) (b)

Figure 9. Plots showing the relative contributions of model terms to model output.

13

4.3.4 Summary of new permeability model
The new permeability model can be used to model permeability changes in coal (and other substrates,
such as sorptive shale) as stresses change. In addition, the model can be used to model permeability
changes caused by the injection of other gases, such as carbon dioxide for sequestration in coal.
Sensitivity analysis of the model found that each of the input variables can have a significant impact
on the outcome of the permeability forecast as a function of changing pore pressure (Robertson and
Christiansen, 2008). However, the permeability model can be used as a tool to determine some of the
parameters by curve-fitting laboratory-generated permeability data. These model-determined values
could then be used for field simulations with a greater degree of confidence.
The new model reduces the effect of sorption-induced strain on permeability compared to two field
permeability models.
5. Economic evaluation
The methodology for performing an economic evaluation of CO2 sequestration/CO2-ECBM outlined
in this section of the report could be applied to other basins around the world using costs and coal
characteristics specific to the basin of interest. The Wyodak PC power plant located near Gillette,
Wyoming, in the PRB was utilized as the CO2 point source and the deep, unminable Wyodak-Anderson
coal zone some 80.5 km from the power plant was chosen as the hypothetical injection site. Unminable
coal was defined as coal below a depth of 304.8 m (Nelson et al., 2005).
Wyoming has several large point source carbon dioxide emitters with total emissions of about 57 Mt
CO2 per year (Robertson, 2007). Table 2 lists the coal-fired power plants in the state, which account for
over 80% of Wyoming's total point source emissions with the balance coming from trona processing,
petroleum refining, and cement manufacture.

Table 2. Wyoming coal-fired power plants.


Capacity CO2 Emissions
Plant County
(MW) (Tons/yr)
Jim Bridger Sweetwater 2,120 18,576,558
Laramie River Station Platte 1,650 14,442,863
Dave Johnston Converse 762 7,362,207
Naughton Lincoln 700 6,012,586
Wyodak Campbell 335 3,762,075
Neil Simpson II Campbell 114 1,264,726
Wygen1 Campbell 90 900,000
The Wyodak facility was selected as the specific CO2 point source for this analysis because of its
location in the Powder River Basin, its proximity to potential geologic sequestration sites, and its
relatively average size compared to other power-generation facilities within the state. The Wyodak PC
power plant generates 335 MW of net power and consumes approximately 5500 Mg/D of Powder River
Basin coal. The plant was built in 1978, with an expected life of 45 years and a thermal efficiency of
29.3%. The coal used in the plant has an average heating value of 7727 Btu/lb (17.934 kJ/g) and an
average carbon content of 46.2 wt.% (Robertson, 2005).
About 613.6 m3/s of flue gas at atmospheric pressure and 85 C flows through the emissions stack.
Nitrogen and carbon dioxide respectively account for 67.0 mol% and 11.8 mol% of the flue gas
discharged from the Wyodak PC power plant (Table 3). On average, 9344 Mg CO2 are emitted each day
from the Wyodak facility (Robertson, 2007).

14
Table 3. Flue gas composition of Wyodak PC power plant.
Flue gas component Concentration
N2 67.0%
CO2 11.8%
O2 12.0%
H2 O 8.0%
CO 300 ppm
SO2 180 ppm
NOx 150 ppm
The Wyodak-Anderson subbituminous coal zone is the largest coal zone in the Powder River Basin.
The approximate extent of the unminable portion of the Wyodak-Anderson coal zone in the PRB is the
area encompassed by the white line drawn in Figure 10. The maximum depth of the Wyodak-Anderson
coal zone is about 2500 ft (762 m) near the basin axis. Defining unminable coal as all coal below 1000 ft
(305 m), the average depth of the unminable coal in the Wyodak- Anderson coal zone is 1500 ft (457 m).



Figure 10. Wyodak-Anderson net coal isopach map. The area
encompassed by the white line represents the approximate unminable coal
area overlain by over 1000 ft (304.8 m) of overburden.

15
The thickness of the unminable coal is between 75 ft (23 m) and 200 ft (61 m) (Figure 10). A value of
100 ft (30 m) was used in the analysis as a representative thickness. Nelson et al. (2005) estimated the
CO2 storage capacity for the unminable coal in the Wyodak-Anderson coal zone of the PRB to be 101.2
trillion cubic feet (5.34 Gt CO2). Based on this capacity, the unminable portion of the Wyodak coal zone
would be capable of sequestering over 3000 years of CO2 from the Wyodak power plant at current output
levels.
5.1 Approach
The majority of coal bed reservoirs are initially saturated with water and whether the adsorbed
methane is produced by pressure depletion or by the aid of gas injection (CO2 sequestration), much of the
water will be pumped from the coal before the majority of the adsorbed gas can be produced. As a base
case, the economic feasibility of traditional methane production without any gas injection to enhance
production was analyzed. Two gas injection scenarios were also analyzed. These are broadly described
as (1) injecting a separated CO2 stream from emissions from the Wyodak PC power plant into an
unminable coal seam in the Powder River Basin; and (2) injecting unseparated flue gas (including the
entrained CO2).
Each scenario was analyzed assuming a 5-spot pattern on 320-ac (129.5-ha) well spacing. For the
two injection scenarios, a 1:1 ratio of injection and production wells was employed as well as a 50-mile
(80.5-km) pipeline to transport the injection gas from the Wyodak plant to the sequestration site.
Additionally, each scenario required the drilling of shallower wells for the disposal of produced water.

Table 4. Base values of input parameters used to compare economics of carbon dioxide sequestration
scenarios.
Value
Input Variable Units
Base Minimum Maximum
Length of pipeline to transport gas to
50 5 50 miles
sequestration site
Depth from surface to coal seam 1,500 1,000 2,000 ft
Capacity of nitrogen separation facility 300 270 330 Mcf/D

Capital costs
Capex associated with water disposal 36,400 35,000 40,000 $/well
Capex for nitrogen separation facility 500 400 550 $/Mcf/D
Mineral rights and permitting 120,000 108,000 132,000 $/320 ac

Operating costs
Injected gas transportation tariff 9.17E-3 8.25E-3 10.09E-3 $/Mcf/mile
carbon dioxide/flue gas separation 42.00 20.00 50.00 $/ton
Water disposal 0.10 0.09 0.11 $/bbl
O&M for nitrogen separation 0.40 0.35 0.50 $/Mcf

Miscellaneous items
Natural gas price 8.00 6.00 12.00 $/Mcf
Price differential between national average
-1.00 -1.50 -0.50 $/Mcf
and Powder River basin wellhead
Inflation rate 0.03
Royalty rate 0.125
Ad valorem & severance tax rate 0.12
Federal income tax rate 0.35
Discount rate 0.10

16
5.2 Economic model description
The economics of each scenario were compared using probabilistic discounted cash flow, after-tax
analysis for the life of the projects. A probabilistic approach was employed to evaluate critical and
significant input variables. Input parameters were assigned triangular distributions by assigning
minimum values, most likely values, and maximum values. Revenues for the scenarios were generated
based on the sale of produced methane from the coal seams. Costs include well drilling costs, injection
gas purchase cost, water disposal costs, taxes, etc. Table 4 is a summary of the input parameters used in
the economic evaluations showing their base values as well as their minimum and maximum values used
for the triangular distributions. These parameters are further discussed in the following section.
5.2.1 Forecast of fluid production rates
Commercial CBM reservoir simulation software was used to simulate the production response from
an unminable coal seam in the Powder River Basin resulting from the injection of carbon dioxide and flue
gas. Important reservoir parameters and their values used for the simulation are shown in Table 5.
Isotherms for carbon dioxide, methane, and nitrogen (Figure 11) are used to calculate production and
storage as a function of in situ pressure within a particular coal seam.

Table 5. Key parameters used for simulation of CO2-


sequestration in PRB unminable coal seams.
Reservoir parameter Value Units
Depth to top of coal seam 1500 ft
Coal thickness 100 ft
Well spacing 320 acre
Initial reservoir pressure 650 psia
Permeability 50 md
Porosity 0.02 fraction



Figure 11. Gas capacity isotherms for carbon dioxide, methane,
and nitrogen for the Wyodak-Anderson coal seam used in the
reservoir simulation.

17
Simulated methane production curves for the three scenarios (Figure 12) show that production is
retarded in the two ECBM scenarios. This retardation is primarily due to the large volume of water being
driven from the injector towards the producing well. The water inhibits methane desorption from the coal
matrix and also impedes the flow of gas through the cleat system.




Figure 12. Methane production curves for the three scenarios
analyzed.

5.2.2 Mineral rights


Mineral rights and permitting for the 320-acre (129.5-ha) spacing was estimated to be $120,000
following the example of Bank and Kuuskraa (2006).
5.2.3 Injection and production wells
The capital costs for these wells were calculated according to the methodology outlined by Bank and
Kuuskraa (2006), for CBM wells in the Powder River Basin. For well depths (Dw) between 1000 ft
(305 m) and 3000 ft (914 m), drilling cost (CD) was calculated using:

At a depth of 1500 ft (457 m) to the coal seam, the cost of the wells were estimated to be $159,500
per well.
5.2.4 Produced water disposal costs
Options for produced water disposal include surface discharge, infiltration impoundment, shallow
underground injection, and deep underground injection. Treatment options that would allow surface
discharge of the produced water include technologies such as reverse osmosis and ion exchange. A
shallow underground injection scheme was chosen as a suitable option for the disposal of produced water
for the three scenarios considered in this paper. Capital costs for this water disposal option were
estimated to be $36,400 and operating and maintenance (O&M) costs were estimated to be $0.10/bbl
($0.63/m3) of produced water (Bank and Kuuskraa, 2006).

18
5.2.5 Transportation costs for injection-gas
The transportation cost for the injection gas was incorporated into a pipeline tariff and includes the
gas compression cost. The pipeline tariff was calculated Thomas et al. (1996):



where T is the pipeline tariff in $/m3/km, Cp is the capital cost of the pipeline construction in $/km, and V,
in m3, is the total volume of fluid transported through the pipeline over a 10-year period. The tariff
calculation applies to both injection scenarios. Robertson (2007) reported that a CO2 pipeline
transporting gas separated from the Wyodak power plant would cost about $522,000/km with a
throughput of 2,383,000 m3/D. Using these costs, the tariff for the CO2 injection scenario was calculated
to be $0.000201/m3/km.
5.2.6 CO2 separation costs
To obtain a CO2 stream for the CO2 injection scenario, a CO2 separations unit was placed at the
Wyodak plant. Three CO2 separation technologies were analyzed for the treatment of the Wyodak flue
gas stream, which were chemical absorption technology, separation by adsorption, and membrane
separation. The analysis concluded that chemical absorption technology was the most economic
separation technology for the specific requirements and captured roughly 50% of the CO2 emitted
(Robertson, 2007). The breakeven sale price of the CO2 separated from the plant was determined to be
$42/ton ($46.3/Mg).
5.2.7 Costs to separate nitrogen from produced gas
Under the flue gas injection scenario, both nitrogen and carbon dioxide were injected into the coal
seam. The nitrogen has a lesser affinity for adsorption onto the coal surfaces than either methane or
carbon dioxide. As a result, the nitrogen component in the flue gas travels through the coal seam and
reaches the producing well sooner than the carbon dioxide necessitating the installation of a nitrogen
separations unit at the production well to remove the nitrogen from the produced gas to improve the
product to pipeline quality. Bomberger et al. (1999) discussed various technologies for nitrogen removal
from natural gas and reported that capital costs for nitrogen removal from natural gas streams were about
$17.66/m3/D with operating costs of about $0.0141/m3. These values were used in the economic
evaluation of the flue gas injection scenario.
5.2.8 Other input variables
The natural gas price forecast according to the Energy Information Administration (EIA, 2007) is
projected to be fairly flat. The analysis assumes a flat natural gas price of $8.00/Mscf ($0.2825/m3) in the
U.S., with a $1.00/Mscf ($0.0353/m3) discount that accounts for the price differential between wellhead
gas in the Powder River Basin and the national average.
The total income tax rate for the analyses was estimated to be 35%, the royalty rate was estimated to
be 12.5%, and the combined state severance and ad valorem tax rate was estimated to be 12%. Inflation
was estimated to be 3% and a discount rate of 10% was used for the discounted cash flow analysis.
5.3 Results and discussion
Economics of the three production scenarios (no gas injection, flue gas injection, and CO2 injection)
were compared using deterministic discounted cash flow analysis. Further analysis of the CO2 injection/
sequestration scenario was done using probabilistic or Monte Carlo techniques.

19
5.4 Deterministic comparisons
Two economic parameters were used to compare the scenarios: net present value at a discount rate of
10% (NPV10) and rate of return on investment (ROI). Economic analyses using the most likely values
show large disparities between the results of the three evaluated production scenarios (Table 6). The
NPV10 of the no-gas-injection scenario was $1.55 million, the NPV10 of the flue-gas-injection scenario
was $0.81 million, and the NPV10 of the CO2-injection scenario was $36.2 million. The most
economically advantageous avenue would be to produce the methane from the coal bed without any
additional enhancement by the injection of either flue gas or carbon dioxide. At current natural gas
prices and costs to capture CO2 from anthropogenic sources, CO2-ECBM is not expected to be profitable
for scenarios similar to those analyzed in this paper.

Table 6. Deterministic economic results using most likely values for input parameters for three
scenarios analyzed.
Methane CO2
Length of NPV10,
Scenario ROI, % recovered, sequestered,
project, yr $(millions)
% GOIP tons
No gas injection
26 1.55 24.3 71.7 0
(pressure depletion)
Flue gas injection
17 -0.81 5.4 70.2 147,000
(flue gas-ECBM)
CO2 injection
19 -36.2 0 88.2 6,860,000
(CO2-ECBM)
The projected termination date of the no-gas-injection scenario is 26 years after project initiation with
recovery of 72% of the methane originally in place (OIP) and no carbon dioxide sequestered. Because of
large amounts of nitrogen in the produced gas, the flue-gas-injection scenario is terminated after 17 years
during which 70% of the methane OIP were recovered and 147,000 tons (133,600 Mg) of CO2 were
sequestered. The CO2-injection scenario was terminated after 19 years due to CO2 breakthrough at the
production well with recovery of 88% of the methane OIP and 6,857,000 tons (6,233,600 Mg) of CO2
sequestered.
Although the economic results suggest that the flue gas-ECBM scenario is close to being profitable,
the amount of CO2 sequestered is small and does not significantly contribute to the need to sequester CO2
in large quantities. Therefore, flue gas injection scenarios should be viewed as an enhanced CBM method
and not as a realistic option for CO2 sequestration. Out of the two injection scenarios analyzed, only the
scenario injecting pure CO2 into the coal seam is a true CO2 sequestration option. This option, although
uneconomic without forced or voluntary subsidies, effectively recovers methane and also sequesters 6.9
million short tons (6.2 Mt) of CO2 per quadrant of a 320-ac (129.5-ha) well-spacing pattern.
5.5 Monte Carlo analysis of CO2 injection scenario
Monte Carlo analysis was used to determine the distribution of net present value for the CO2 injection
scenario by employing the triangular distribution assigned to each input variable. The resulting
distribution of NPV10 for the CO2 injection scenario is shown in Figure 13, which ranges from
-$43 million to -$6 million and has a mean value of -$29.3 million.


20


Figure 13. Probabilistic net present value (NPV10) of the CO2
injection scenario.

5.5.1 Sensitivity analysis


A sensitivity analysis was performed on the CO2 injection scenario to determine the input parameters
causing the most variability in the net present value. Correlation coefficients, which represent the degree
to which an input variable and net present value change together, were calculated for each variable and
normalized with respect to 100%. A negative correlation coefficient means that as the input variable
increases, the resulting net present value tends to decrease; while a positive coefficient signifies an
increase in NPV with an increase in the input variable. The correlation coefficients of the five input
variables with the most impact on the net present value are plotted in Figure 14. The input parameter with
the greatest effect on economic viability is the cost to separate CO2 from flue gas.
5.5.2 Total CO2 cost to sequestration projects
The total cost of the CO2 to the sequestration project is the combination of costs to separate it from
flue gas and costs to transport it to the sequestration site (including compression costs). Data shown in
Figure 14 identify the cost to separate and capture CO2 from the flue gas stream of the Wyodak PC power
plant as the most critical input variable for the economic evaluation of the CO2 injection project in the
Powder River Basin because it has the greatest impact on the net present value of the scenario.
CO2 separation and capture has the greatest impact on net present value not only because of its large
cost, but also because of its large uncertainty range. The most likely value of $42/ton ($46.2/Mg) for this
cost was based on a site-specific estimate of retrofitting currently commercial CO2 separation technology.
The minimum value for this cost of $20/ton ($22/Mg) was based on future improvements to CO2 capture
and separation technology that would dramatically reduce this cost to about one-half of its current value;
while the maximum value of $50/ton ($55/Mg) was selected assuming no future technological
improvements.
Assuming that the CO2 separation cost is $42/ton ($46.2/Mg) and also including the transportation
cost of $0.46/Mscf ($0.0162/m3) for a 50-mile (80.5-km) pipeline, the total cost of the CO2 is $3.03/Mscf
($0.107/m3). If the separation cost were reduced to $20/ton ($22/Mg) of CO2, the total cost of the CO2
would be $1.68/Mscf ($0.0593/m3).


21


Figure 14. Correlation coefficients for the five input variables with the
most impact on the net present value of the CO2 injection scenario.

The economic model was evaluated to determine the effect of reducing the CO2 separation cost to
$20/ton ($0.706/Mg) through technological innovation. The resulting NPV10 distribution of the scenario
was still negative for all cases. This information indicates that, at least for this specific location, injecting
CO2 into an unminable coal seam would most likely never be profitable without some additional
economic driver being present. Powder River Basin coal has low methane content typically around 30
scf/ton (0.9365 m3/Mg) of coal (ALL, 2004) compared to other basins containing higher rank coal such
as the San Juan Basin with methane content of around 400 scf/ton (12.48 m3/Mg) of coal (ALL, 2004).
The low volumetric recoveries for the Powder River Basin are not sufficient to offset the costs of CO2
capture and injection. The economic profitability of CO2 sequestration in coal in other basins containing
coal with higher methane contents is not addressed in this paper.
Because CO2 capture/separation cost is the biggest economic driver, knowing what its value would
need to be in order to create a favorable economic incentive to inject CO2 is extremely important. To
arrive at this important information, an economic simulation of 1000 Monte Carlo trials was set up and
run using the input variable distributions. After each trial, the CO2 cost yielding a rate of return of 10%
(NPV10 equal to $0) was calculated and tabulated. The cost of separating CO2 from the Wyodak PC
power plant would need to be less than zero for there to be an economic incentive to inject the CO2 into
an unminable coal seam in the Powder River Basin (see Figure 15). This condition could not be achieved
from technological improvements alone, but must be imposed by other outside entities such as a tax on
CO2 emissions or some other incentive. The distribution of CO2 separation cost required for economical
CO2 sequestration in the Powder River Basin ranges from a low of $11/ton (12.1/Mg) to a high of $3/ton
($3.3/Mg) with an average separation cost of $4.36/ton ($4.80/Mg). A CO2 separation cost of $42/ton
corresponds to a total CO2 cost of $3.03/Mscf ($0.107/m3). The breakeven CO2 separation cost of
$4.36/ton ($4.806/Mg) corresponds to a total cost for CO2 at the injection site equal to $0.19/Mscf
($0.0067/m3).
5.6 Summary of economic evaluation
Although injecting flue gas to recover methane from CBM fields may be economic, this method will
not significantly contribute to the solution needed to sequester CO2 in large quantities. A flue gas
injection project should be viewed as an enhanced CBM method and not as a realistic option for CO2
sequestration.
This study suggests that separating CO2 from flue gas and injecting it into the unminable coal zones
of the Powder River Basin seam, while currently uneconomical, can increase recovery of methane by
17% and can sequester over 86,000 tons/ac (192,800 Mg/ha) CO2.

22




Figure 15. Distribution and mean value of the cost of CO2
separation/capture required to yield a 10% rate of return.

The cost to separate CO2 from flue gas was identified as the major cost driver associated with CO2
sequestration in unminable coal seams in the Powder River Basin. Using existing commercial separations
technology, the cost to separate CO2 from flue gas is currently $42/ton ($46.3/Mg) CO2, which yields a
total cost of CO2 at a sequestration site 50 miles (80.5 km) from the point source equal to $3.03/Mscf
($0.107/m3).
Improvements in separations technology alone are unlikely to drive costs low enough for CO2
sequestration in unminable coal seams in the Powder River Basin to become economically viable. The
average CO2 separation cost would need to be $4.36/ton ($4.806/Mg) corresponding to a total cost of
CO2 of $0.19/Mscf ($0.0067/m3) at a sequestration site to achieve economic viability for this basin with
relatively low gas-content coal.
In order to achieve economic viability, some form of government incentive would need to be imposed
to drive the cost to separate CO2 from flue gas to less than zero. Technological breakthroughs could aid
the economics, but would not be expected to accomplish all that is necessary.
6. References
ALL Consulting, Montana Board of Oil and Gas Conservation, 2004. Coal bed methane primer: new
source of natural gasenvironmental implications. Contract No. DEFG26-02NT15380, US DOE,
National Petroleum Technology Office, Tulsa, Oklahoma, February.
Bank, G.C., Kuuskraa, V.A., 2006. The economics of Powder River Basin coalbed methane
development. Prepared for the US DOE by Advanced Resources International, January.
Berkowitz, N., 1985. Coal Science and Technology 7 The Chemistry of Coal, 86. New York City:
Elsevier Science.
Bomberger, D.C., Bomben, J.L., Amirbahman, A., Asaro, M., 1999. Nitrogen removal from natural gas:
phase II final report. Contract No. DE-AC21-95MC32265, SRI International, Menlo Park California.

23
Busch, A., Gensterblum, Y., Siemons, N., Krooss, B.M., 2003. Investigation of preferential sorption
behaviour of CO2 and CH4 on coals by high pressure adsorption/desorption experiments with gas
mixtures. 2003 International Coalbed Methane Symposium, University of Alabama, Tuscaloosa,
Alabama, 5-9 May, paper 0350.
EIA, 2007. Annual energy outlook 2007 with projections to 2030. US DOE, Energy Information
Administration. Report No. DOE/EIA-0383(2007).
Ellis, M.S., Gunther, G.L., Ochs, A.M., Roberts, S.B., Wilde, E.M., Schuenemeyer, J.H., Power, H.C.,
Stricker, G.D., Blake, D., 1999. Chapter PN: coal resources, Powder River Basin. US Geological
Survey Professional Paper 1625-A.
Harpalani, S. and R.A. Schraufnagel, 1990: Influence of Matrix Shrinkage and Compressibility on Gas
Production from Coalbed Methane Reservoirs, SPE paper 20729, presented at the SPE Annual
Technical Conference, New Orleans, LA.
Hendriks, C., Graus, W., van Bergen, F., 2004. Global carbon dioxide storage potential and costs.
ECOFYS, Utrecht, The Netherlands. Report No. EEP-02001.
Liang, J.T., Raterman, K.T., Robertson, E.P., 2003. A mechanistic model for CO2 sequestration in tiffany
coal bed methane field. 2003 International Coalbed Methane Symposium, University of Alabama,
Tuscaloosa, Alabama, 5-9 May, paper 0339.
Mathews, J.P., Karacan, O., Halleck, P., Mitchell, G.D., Grader, A., 2001. Storage of pressurized carbon
dioxide in coal observed using X-ray tomography. First National Conference on Carbon
Sequestration, Washington, D.C., 14-17 May.
Mavor, M.J., Gunter,W.D., Robinson, J.R., 2004. Alberta multiwell micro-pilot testing for cbm
properties, enhanced methane recovery and CO2 storage potential. 2004 SPE Annual Technical
Conference and Exhibition, Houston, Texas, 26-29 September, paper SPE 90256.
Nelson, C.R., Steadman, E.N., Harju, J.A., 2005. Geologic CO2 sequestration potential of the Wyodak-
Anderson coal zone in the Powder River Basin. Plains CO2 Reduction (PCOR) Partnership Topical
Report for U.S. Department of Energy; Energy and Environmental Research Center: University of
North Dakota, August.
Palmer, I. and J. Mansoori: How Permeability Depends on Stress and Pore Pressure in Coalbeds: A New
Model, paper SPE 52607, SPE Reservoir Evaluation & Engineering (December 1998) 539-544.
Pekot, L.J., Reeves, S.R., 2003. Modeling the effects of matrix shrinkage and differential swelling on
coalbed methane recovery and carbon sequestration. 2003 International Coalbed Methane
Symposium, University of Alabama, Tuscaloosa, Alabama, 5-9 May, paper 0328.
Prusty, B.K., Harpalani, S., 2004. A Laboratory study of methane/CO2 exchange in an enhanced CBM
recovery scenario. 2004 International Coalbed Methane Symposium, University of Alabama,
Tuscaloosa, Alabama, 12-14 May, paper 0426.
Puri, R., Yee, D., 1990. Enhanced coalbed methane recovery. 1990 SPE Annual Technical Conference
and Exhibition, New Orleans, Louisiana, 23-26 September, paper SPE 20732.
Reeves, S.R., 2003. Assessment of CO2 sequestration and ECBM potential of US coalbeds. US DOE
topical report No. DE-FC26-00NT40924, February.
Reeves, S., Taillefert, A., Pekot, L., Clarkson, C., 2003. The Allison unit CO2-ECBM pilot: a reservoir
modeling study. US DOE topical report No. DE-FC26-0NT40924, February.
Robertson, E.P., 2005. Measurement and modeling of sorption-induced strain and permeability changes
in coal. Ph.D. Thesis, Colorado School of Mines, Golden, Colorado, December; also published as an
external report from Idaho National Laboratory, INL/EXT-06-11832, October.

24
Robertson, E.P., 2007. Analysis of CO2 separation from flue gas, pipeline transportation, and
sequestration in coal. Idaho National Laboratory external report No. INL/EXT-08-13816, Idaho
Falls, Idaho, September.
Robertson, E.P. and Christiansen, R.L., 2007. Modeling laboratory permeability in coal using sorption-
induced strain data. SPE Reservoir Evaluation & Engineering 10 (3), 260269.
Robertson, E.P. and Christiansen, R.L., 2007 (a). Methods and Apparatus for Measurement of a
Dimensional Characteristic and Methods of Predictive Modeling Related Thereto. U.S. Patent No.
US 7,224,475.
Robertson, E.P. and Christiansen, R.L., 2007 (b). Methods for Measurement of a Dimensional
Characteristic and Methods of Predictive Modeling Related Thereto. U.S. Patent No. US 7,284,604
B2.
Robertson, E.P. and Christiansen, R.L., 2008. A permeability model for coal and other fractured,
sorptive-elastic media. SPE Journal, September, 314-324.
Schroeder, K., Ozdemir, E., Morsi, B.I., 2002. Sequestration of carbon dioxide in coal seams. J. Energy
& Environment Research 2, 5463 1.
Seidle, J.P. and L.G. Huitt, 1995: Experimental Measurement of Coal Matrix Shrinkage Due to Gas
Desorption and Implications for Cleat Permeability Increases, paper SPE 30010, presented at the
SPE international Meeting on Petroleum Engineering, Beijing, China (14-17 November).
Shi, J.Q. and S. Durucan: Changes in Permeability of Coalbeds During Primary Recovery Part 1:
Model Formulation and Analysis, paper 0341 proc. at the 2003 International Coalbed Methane
Symposium, University of Alabama, Tuscaloosa, Alabama (May 2003).
Steinberg, M., 2001. Decarbonization and sequestration for mitigating global warming. First National
Conference on Carbon Sequestration, Washington, D.C., 14-17 May.
Stevens, S.H., Spector, D., Riemer, P., 1998. Enhanced coalbed methane recovery using CO2 injection:
worldwide resource and CO2 sequestration potential. 1998 SPE International Conference and
Exhibition in China held in Beijing, China, 2-6 November, paper SPE 48881.
Thomas, C.P., Doughty, T.C., Hackworth, J.H., North, W.B., Robertson, E.P., 1996. Economics of
Alaska north slope gas utilization options. Idaho National Laboratory external report No. INEL-
96/0322, Idaho Falls, Idaho, August, B-8.
U.S. DOE, 2007a. Electric power annual with data for 2006. US DOE EIA, 22 October. Available
online at: http://www.eia.doe.gov/cneaf/electricity/epa/epa_sum.html.
U.S. DOE, 2007b. 2007 coal power plant database. US DOE National Energy Technology Laboratory,
16 November. Available online at: http://www.netl.doe.gov/publications/ press/2007/07079-
Coal_Plant_Database_Available.html.
van Bergen, F., Pagnier, H.J.M., van der Meer, L.G.H., van den Belt, F.J.G., Winthaegen, P.L.A.,
Krzystolik, P., 2003. Development of a field experiment of ECBM in the Upper Silesian coal basin
of Poland (RECOPOL). 2003 International Coalbed Methane Symposium, Tuscaloosa, Alabama, 5-9
May, paper 0320.
White, C.M., Smith, D.H., Jones, K.L., Goodman, A.L., Jikich, S.A., La Count, R.B., DuBose, S.B.,
Ozdemir, E., Morsi, B.I., Schroeder, K.T., 2005. Sequestration of carbon dioxide in coal with
enhanced coalbed methane recoverya review. Energy & Fuels 19 (3), 659724.
Zutshi, A., Harpalani, S., 2004.Matrix swelling with CO2 injection in a cbm reservoir and its impact on
permeability of coal. Paper 0425 presented at the 2004 International Coalbed Methane Symposium,
University of Alabama, Tuscaloosa, Alabama, 12-14 May.

25

You might also like