SPE

Society of Petroleum Engineers

SPE 20732

Enhanced Coal bed Methane Recovery

R. Puri and D. Yee, Amoco Production CO.
SPE Members

Copyright 1990. Society of Petroleum Engineers Inc.

This paper was prepared for presentation at the 65th Annual Technical Conference and Exhibition of the Society of Petroleum Engineers held in New Orleans, LA, September 23-26, 1990.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract submitted by the author(s). Contents of the paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessanly reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

METHANE IS ADSORBED TO COAL SURFACE

ABSTRACT
At present, coal bed methane is being recovered
by means of reservoir pressure depletion. While
this method is simple and effective, it is not effi-
cient. Reduction in reservoir pressure deprives the
fluids of the energy necessary to flow to the well-
bore. Furthermore, there is a practical and eco-
nomic limit on the extent to which reservoir
pressure can be reduced. It is estimated that res-
ervoir pressure depletion strategy of coalbed meth-
ane production will permit the recovery of 50% or
less of the gas-in-place.
This paper discusses an alternate method of
coalbed methane recovery which consists of nitrogen
flooding a coal seam. Lab research has shown that
essentially all methane sorbed on coal can be
stripped by nitrogen without necessarily reducing
the total system pressure. That is, methane desorp-
tion from coal 1S achieved by reducing the partial
pressure of methane rather than merely the total
pressure. Lab and modeling results are presented to
demonstrate how nitrogen injection in coal can
accelerate methane production rates and enhance
reserves. Scoping process economics of enhanced
coal bed methane production also appear to be attrac-
tive.
INTRODUCTION
Coal bed methane gas-in-place is estimated to be
400 TCF, of which about 95 TCF is considered to be
economically recoverable with current technologyl.
This represents a significant resource Slnce the
total U.S. proved gas reserves are estimated to be
187 TCF. Efforts to commercially recover coalbed
methane have begun with the drilling of over 2000
wells, primarily in the San Juan and Warrior Basins.
References and illustrations at end of paper.
Unlike conventional gas reservoirs, methane in
coal is not stored as a free gas but rather as
sorbed gas, at near liquid densities, on the inter-
nal surface area of the microporous coal. As shown
in Figure 1, the methane storage capacity of coal is
given in terms of a sorption isotherm where the gas
content in SCF/ton is plotted verses pressure at a
constant temperature. For example, if the reservoir
pressure of a coal seam were 500 psia at discovery,
the coal would be capable of holding a maximum of
about 270 SCF of methane gas per ton of coal.
LOW GAS RECOVERY ANTICIPATED WITH CURRENT TECHNOLOGY
At present, coalbed methane is recovered using
a reservoir pressure depletion strategy. That is,
reservoir pressure is reduced by removing water
which causes some gas to be desorbed from coal.
Methane diffuses through the microporous matrix
blocks and upon reaching the cleat system, a network
of natural fractures, flows to the wellbore along
with water as per Darcy's Law. Figure 2 shows a
schematic view of the coal bed methane recovery mech-
anism.
While the pressure depletion method of coal bed
methane production is simple and effective, it is
not efficient. Loss in pressure deprives the reser-
voir fluids of the energy necessary to flow to the
wellbore. Consequently, gas production rates suffer
and the ultimate methane recovery by means of pres-
sure depletion 1S generally not expected to be
greater than 50% of the gas-in-place, even after
several decades of production. For example, as
shown in Figure 3, to recover 50% of the gas-in-
place from coal, the reservoir pressure would have
to be reduced from 500 psia to less than 150 psia.
Since reduction in reservoir pressure much below 150
psi is generally not considered to be practical or
economic, a substantial amount of methane in coal is

193
2 ENHANCED COALBED METHANE RECOVERY
expected to be left behind by the current operating
strategy of pressure depletion.
COAL DESORPTION SIMILAR TO SORBENT BED REGENERATION
Adsorbent beds are commonly used in the chemi-
cal process industry for gas separation or purifica-
tion. Examples of adsorbent beds are silica gels,
charcoals, zeolites, catalysts, molecular sieves,
etc. Once these adsorbent beds become saturated,
they must be regenerated before they can be reused.
The desorption of methane from coal is actually no
different from regenerating an adsorbent bed. In
other words, a coal seam is like a giant adsorbent
bed that we wish to regenerate by removing methane
adsorbed to the surface. Therefore, the methods
used to regenerate adsorption beds should be equally
applicable to the recovery of methane from coal
seams. In the literature, adsorbent bed regener-
ation methods are classified as (1) Pressure
Depletion, (2) Inert Gas Stripping, (3) The2m~1
Desorption, and (4) Displacement Desorption ' •
Pressure depletion is a common adsorbent bed
regeneration method in use today because of its sim-
plicity and cost effectiveness. Inert gas stripping
of an adsorbent bed is achieved by reducing the par-
tial pressure through the introduction of a low-ad-
sorbing gas at constant pressure. The partial
pressure of a component is equal to the total system
pressure multiplied by its mole concentration in the
gas phase. Inert gas stripping methods are pre-
ferred only when the adsorbed species is held weakly
to the adsorbent bed. Thermal desorption uses heat-
ing to increase the temperature of the adsorbent bed
so as to reduce its adsorption capacity. In dis-
placement desorption, the adsorbed gas is displaced
by the introduction of another gas which is competi-
tively adsorbed. In practice, usually two or more
of these methods are combined.
This paper discusses the potential application
of the inert gas stripping method for recovering
coalbed methane 4 • Figure 4 shows a schematic view
of the enhanced coal bed methane recovery mechanism
by means of nitrogen gas injection. Nitrogen
injected into coal reduces the concentration of
methane in the gas phase while maintaining, or even
increasing, the total system pressure. Conse-
quently, methane desorbs from the coal matrix and
diffuses to the cleat network. As mentioned ear-
lier, methane, nitrogen, and water in the cleat net-
work flow to the producing wellbores as per Darcy's
Law.
LABORATORY EXPERIMENTS
Laboratory experiments have been performed to
understand methane sorption and desorption on coal
in the presence of other gases such as nitrogen and
helium. Furthermore, lab tests were conducted to
determine the feasibility of stripping methane from
coal by nitrogen and helium gases. Some of the lab
test results are briefly presented in this paper.
METHANE PARTIAL PRESSURE CONTROLS ADSORPTION TO COAL
Sorption equilibrium measurements made in the
lab using mixtures of methane and helium, the latter
being a nonsorbing gas, have shown that adsorption
(increasing pressure) of methane on coal is con-
194
SPE 20732
trolled by the partial pressure of methane gas
rather than the total pressure. In other words, the
amount of methane sorbed on coal is not only depend-
ent on the total system pressure, but also on the
concentration of methane in the gas phase.
In one test, dry finely ground (SO-230 mesh
size) coal from the San Juan Basin was placed in a
pressure cell and its sorption isotherm measured
with pure methane at room temperature (77°F). The
gas content verses pressure data is shown are
Figure S. Note that both the adsorption and desorp-
tion gas contents lie on the same isotherm indicat-
ing that the methane sorption is reversible.
Furthermore, the methane sorption isotherm fits a
Langmuir isotherm given by:
where V is the gas content in SCF/ton, P is the
pressure in psia, VM is the Langmuir sorption isot-
herm constant in SCF/ton L and b is the Langmuir
pressure constant in psi 1
The adsorption of three different
methane/helium mixtures were measured. The adsorp-
tion capacity of helium on coal has been shown to be
negligible, and should not interfere in any way.
For each mixture, the measurement started with an
evacuated sample cell. Next, a methane/helium mix-
ture was injected, and the sample cell was allowed
to reach pressure equilibrium. The final equilib-
rium pressure was approximately equal for all three
tests, but the methane partial pressure was changed
for each test by varying the injected gas composi-
tion.
Aside from the final sample cell pressure, it
was also necessary to know the final gas composition
in order to calculate the methane sorption and
verify that helium sorption was negligible. This
was done by collecting small gas samples from the
void volume and having them analyzed by mass spec-
troscopy. With this information, the real gas law
was used to determine the quantity of free methane
in the sample cell; and by difference from the quan-
tity injected, the amount of methane sorbed on coal
was calculated.
The methane/helium adsorption results are shown
in Figure S. Only the methane partial pressure and
gas contents are presented, since there was no
appreciable helium adsorption, as expected. The
methane/helium mixture gas contents fall exactly on
the pure methane isotherm when plotted on a methane
partial pressure basis. This shows that the methane
adsorption is determined by the methane partial
pressure. If the total pressure controlled methane
adsorption to coal, the methane gas contents would
have to be much higher.
PARTIAL PRESSURE ALSO CONTROLS DESORPTION FROM COAL
Tests were also performed to show that the
desorption (decreasing pressure) of methane is
determined by the partial pressure of methane. This
behavior was expected since the previously measured
pure methane isotherm given in Figure S showed that
the methane sorption was reversible.
SPE 20732 RAJEN PURl AND DAN YEE
A moist, finely ground (50-230 mesh size) coal
from the San Juan Basin was placed in a pressure
cell and its isotherm measured at reservoir temper-
ature (115°F). The pressure and gas content data
are shown in Figure 6. Unlike the previously
described methane/helium adsorption tests which
began with no methane initially sorbed on the coal
sample, this test started with the coal sample fully
sorbed with methane at 626 SCF/ton at 1614 psia.
This pressure represents both the total and methane
partial pressure since methane is the only component
in the gas phase.
Helium was injected into one end of the cell
while collecting all the gas produced from the other
end. In so doing, it was possible to maintain the
sample cell at a constant (approximate) total pres-
sure of 1614 psia, even though the injection of
helium continually reduced the methane partial pres-
sure. After each of the four helium injection
steps, the sample cell was shut-in and allowed to
reach pressure equilibrium. At each shut-in, the
pressure would rise signifying the desorption of
methane due to the reduction in the methane partial
pressure.
An overall mass balance was determined by know-
ing the quantity of injected helium and produced
gas. The composition of the produced gas was deter-
mined by mass spectroscopy. By assuming that helium
does not sorb, the real gas law could be used to
determine the free and sorbed methane, since the
amount of methane and helium in the sample cell was
known from the overall mass balance.
The methane desorption results are presented in
Figure 6. The methane partial pressure was reduced
from 1614 psi to 209 psi over the four steps, and
the sorbed methane was reduced from 626 to 253
SCF/ton. Nearly 60% of the original methane on the
coal desorbed, though the total pressure in each
step was kept greater than, or equal to, the start-
ing value of 1614 psia. When plotted in terms of
methane partial pressure, the gas contents lie on
the pure methane isotherm, signifying that methane
desorption, just as adsorption, is controlled by
methane partial pressure. Water in excess of the
equilibrium moisture was used for this test, but
that did not hamper the methane desorption or
adsorption.
METHANE ADSORBED ON COAL CAN BE STRIPPED BY NITROGEN
Since it does not sorb on coal, helium is ideal
for demonstrating that the methane partial pressure
controls the adsorption and desorption of methane.
Still, helium is not a practical gas to use in real
applications due to its high cost and scarcity.
Nitrogen, because it is cheap and abundant, would be
a better choice even though it does sorb on coal but
to a lesser extent than methane. Because of its low
sorption characteristics, nitrogen would be expected
to act much like helium. In fact, some additional
benefit occurs since nitrogen actually releases
methane by both sorption displacement and partial
pressure reduction.
Test 1: To confirm that nitrogen is less
strongly sorbed than methane, a sorption isotherm
was measured for pure nitrogen. A moist finely
ground coal sample was used. The lab results are
195
3
presented in Figure 7. At 908 psia, the nitrogen
sorption appears to be about 40% that of methane,
confirming that coal has a lower capacity to sorb
nitrogen as compared to methane.
An inert gas stripping test was performed by
continuously injecting nitrogen into the sample
cell. A back pressure regulator was used to main-
tain a constant total pressure in the sample cell.
A schematic of the experimental set-up is shown in
Figure 8.
The sample cell was filled with methane and
allowed to come to equilibrium at 908 psia and
115°F. From the measured isotherm, the sorbed meth-
ane was calculated to be 517 SCP/ton. At the ini-
tial conditions of 908 psia and 115 0 p, the sample
cell void of 93.1 cc contained 5541 cc (STP) of free
methane in addition to the 898 cc (STP) of sorbed
methane in 55.67 gm of coal. The sample cell void
volume represented an equivalent porosity of 69.2%
The nitrogen was injected into the sample cell
at the average rate of 126.9 cc (STP)/min for a
total time of 119 minutes. The nitrogen injection
was equivalent to 2.93 pore volumes. Produced meth-
ane was calculated from the measured volume and com-
position of the produced gas. The calculated
results are shown in Figures 9 and 10, methane con-
centration verses time and cumulative produced meth-
ane verses time, respectively. While it is likely
that methane was desorbed as soon as the nitrogen
was injected, it could only be determined that
sorbed methane was produced after all the free meth-
ane, 5541 cc (STP), in the sample cell was taken
into account. This occurred roughly 50 min after
nitrogen injection was initiated. By the end of the
test, all the free methane plus 394 SCF/ton of
sorbed methane had been produced, which represents
76.2% of the total sorbed methane or 684 cc out of
989 cc (STP).
To see if the remaining sorbed methane could be
recovered, nitrogen was injected for two more peri-
~ds, consisting of 117 and 119 minutes. The second
injection occurred after the sample cell had been
shut-in for 66.4 hours. An additional 173 cc (STP)
of methane was measured, increasing the total recov-
ery of sorbed methane to 857 cc (STP) or 493
SCF/ton. The third injection took place after a
shut-in of 50.7 hours. Another 58 cc (STP) was col-
lected, increasing the total recovered sorbed meth-
ane to 915 cc (STP) or 526 SCF/ton. The overall
methane mass balance was good since 6456 cc (STP)
was measured compared to 6438 cc (STP) that was cal-
culated to be in the sample cell at the start. The
total time for all three nitrogen injection periods
was 355 minutes which is typical of the time
required for the finely ground coal sample to come
to equilibrium by diffusion. The nitrogen sorbed
during this experiment was not measured, but if the
coal had been stripped of all methane (517 SCF/ton),
it would have adsorbed 198 SCF/ton of nitrogen gas
at 908 psia and 115°F.
Test 2: Since finely ground coal sample does
not represent the length and time scales typical of
an actual coal seam, a second test was conducted
using a core sample with 2.13" diameter and 6"
length. The wet weight at the beginning of the test
was 554 gm. After completely drying at the end of
4 ENHANCED COALBED METHANE RECOVERY
the test, the weight of the core sample was 542 gm
which means that the moisture was 2.2% at the begin-
ning. The sample was placed into a 2.5" diameter by
6.38" deep cell and packed in 0.0625" diameter glass
beads. The glass beads and the associated cell void
volume act like the cleat system in a real coal
seam, while the sample acts like the coal matrix
blocks. Darcy flow occurs through the cell void
volume filled with glass beads, while gas desorption
and diffusion take place in the coal sample. The
cell void volume was determined to be 116.65 cc or
26.6%.
The coal in the cell was completely saturated
with methane at 899 psia and 115°F. There was 337
SCF/ton or 5705 cc (STP) of methane adsorbed to coal
and 6402 cc (STP) free methane contained in the cell
void volume. That is, 47.2% of the methane in the
cell was sorbed to coal. The test set-up schematic
is illustrated in Figure 11.
Injection of nitrogen gas into the test cell
was initiated at a rate of 25 cc(STP)/min. A back
pressure regulator on the cell maintained a constant
pressure of about 900 psi at all times. Effluent
gas was analyzed for its composition, and a complete
record of gas injection and production was main-
tained. The nitrogen injection was performed in
four periods. The first period lasted 3469 min,
after which the cell was shut-in for 3722 min. The
second period lasted 470 min, and the subsequent
shut-in was only 976 min. The last two periods were
450 and 441 min, separated by shut-ins of 2407 min
each. To complete the mass balance, the pressure
was reduced to atmospheric while maintaining the
temperature at 115°F. This blowdown step occurred
after the 5599 min shut-in period following the end
of nitrogen injection. To remove residual methane,
four separate helium pressure cycles were performed.
Finally, the cell was depressured to room condi-
tions.
The produced methane concentration versus time
is presented in Figure 12. The gas is all methane
for the first 125 min. This methane bank is due to
the displacement of the free methane by the injected
nitrogen. The concentration falls rapidly as nitro-
gen breakthrough occurs. Methane concentration is
less than 10% by 728 min and falls to approximately
2% by 2469 min, the end of the first injection
period. In the subsequent three injection periods,
methane concentration peaks at 4.6%, 2.1%, and 1.7%,
respectively.
The cumulative produced methane versus time is
presented in Figure 13. The cumulative methane goes
up linearly with time until nitrogen breakthrough
occurs at 125 min, after which the curve begins to
flatten. Even so, it takes only 324 min to produce
the 6402 cc (STP) equal to the free methane in the
cell at the start. After 728 min, 8279 cc (STP) of
methane was recovered. This corresponds to all the
free methane in the void volume of the cell and 111
out of the 337 SCF/ton of sorbed methane (i.e., 33%
of the methane sorbed on coal). After the end of
the first injection period (3469 min), roughly 250
SCF/ton or 74.2% of the methane was desorbed from
coal. When the test was terminated, almost 289
SCF/ton or 85.6% of the methane was recovered from
coal. To achieve the same amount of methane by
means of pressure depletion, the total pressure
196
SPE 20732
would have to reduced from 900 psi to about 52.5
psi.
The overall nitrogen injection time of 4830 min
is similar to the time required for core desorption
to occur by diffusion when the total pressure is
reduced. The diffusivity of methane in this core
was 2.01 E-4 min- 1 which corresponds to a time of
4975 min for complete (100%) methane desorption to
occur by pressure depletion. That is, the diffusiv-
ity for inert gas stripping is similar to that for
pressure depletion. Since coal bed methane recovery
by pressure depletion is most often limited by Darcy
flow in the cleat system rather than diffusion of
gas from the matrix, the same would be true for
nitrogen stripping.
The methane produced following nitrogen
injection was used to complete the mass balance.
The blowdown step reduced the total pressure to
atmospheric and produced an additional 236 cc (STP)
of methane. Afterwards, pressure cycling with
helium produced an extra 152 cc (STP) of methane.
Including these two contributions, the total meas-
ured methane is equal to 11672 cc (STP) compared to
the 12107 cc (STP) initially placed in the cell. As
a result, 96.4% of the initial methane was measured
giving a reasonable mass balance. The difference is
attributed to experimental errors. No attempt was
made to measure the nitrogen adsorbed to the coal at
the end of the test. However from the pure nitrogen
isotherm available on this type of coal, 123 SCF/ton
of nitrogen would be lost to the coal at 900 psia,
but only after all (322 SCF/ton) of methane had been
recovered.
NUMERICAL MODELING
Substantial effort has been made to accurately
model the recovery of methane by the injection of
nitrogen gas in coal. A state-of-the-art numerical
simulator was used along with the assumption that
the binary Langmuir theory adequately models the
methane-nitrogen sorption characteristics on coals.
This assumption is currently being verified in the
laboratory. Therefore, the modeling studies took
into consideration the loss of nitrogen to coal due
to adsorption. Results of one modeling study are
presented here primarily to demonstrate that methane
recovery from coal can be accelerated and enhanced
by means of nitrogen flooding. The reservoir
description used in the following example does not
necessarily represent an actual field, nor is the
operating strategy optimum. Therefore, only the
differences in model response during pressure
depletion and nitrogen flooding should be noted, and
not their absolute values.
NITROGEN FLOODING ENHANCES METHANE RATE AND RECOVERY
The example field consists of 66 producers on
320 acres/well spacing. The model assumed that 46
ft of coal is contained in four layers ranging in
thickness from 4 to 18 ft, at an average depth of
2800 ft. The permeabilities of the model layers
ranged from 1.75 md to 7.5 md, with an average value
of 4.5 md. The coal seams are assumed to be uniform
and isotropic, with no permeability in the vertical
direction. The simulation study presented here also
disregards any change in permeability of the coal
seam as a result of changes in reservoir pressure.
SPE 20732 RAJEN PURl AND DAN YEE
Lab measured gas-water relative permeability curves
obtained from coal cores samples from an actual
field were used, along with the assumption that the
net coal porosity of the coal cleat network is 17..
The initial reservoir pressure of the coal seam in
the model was approximately 1400 psia. The coal
seam was assumed to be completely saturated with
about 590 SCF/ton of methane gas at the initial res-
ervoir conditions. The production wells are assumed
to be unstimulated, while a -3.0 skin was used for
the injectors.
In Run #1, the 66 well field was placed on
pressure depletion with a constant bottomhole pres-
sure in every well of 200 psia at all times. In Run
#2, after 6 years of pressure depletion, infill
injection wells were drilled so as to change the
spacing to 160 acres/well. That is, in Run #2,
there are 66 injectors and 66 producers in a 5-spot
pattern, while the location of the producers
remained uncaanged. Nitrogen injection is con-
trolled by maintaining a fixed injection pressure of
2000 psia, well below the formation parting pres-
sure. The model predicted total gas and water rates
from the field are shown in Figures 14 and IS for
both Run #1 and Run #2. Clearly, the dewatering of
the coal is accelerated by nitrogen injection and
there is a very large increase in the total gas pro-
duction rate from the field. The composition of the
produced gas is shown in Figure 16, and the net
methane recovered as a function of time is shown in
Figure 17. Figure 18 shows the cumulative methane
recovered for the first 20 years of well life.
Whereas a pressure depletion strategy led to the
model predicted recovery of about 157. gas-in-place
after 20 years, injection of nitrogen increased the
methane recovery to 387.. Note that since the perme-
ability of coal in the model is low, and the wells
are assumed to be unstimulated, the gas-in-place
recovery predicted by the model, even after 20
years, is also unusually low. However, acceleration
in methane production by means of nitrogen injection
is clearly evident. Since methane production rate
from this hypothetical field is high even after 20
years of nitrogen flooding, the separation and rein-
jection of nitrogen can be continued.
Extensive lab and modeling studies have conclu-
sively demonstrated the technical viability of the
enhanced coalbed methane recovery process. Scoping
economic studies of nitrogen flooding have been per-
formed and were found to be encouraging. The
expense of nitrogen generation, injection, and sepa-
ration is more than offset by the benefits of rate
acceleration and reserve enhancement. However,
pilot testing in the field would be required to
obtain a performance history match and make accurate
predictions.
CONCLUSIONS
Lab, modeling, and economic studies have shown
that:
1. The adsorption capacity of coal depends on the
component type and its partial pressure in the
gas phase.
2. Methane is desorbed from coal when the partial
pressure of methane is reduced.
197
5
3. Methane adsorbed to coal can be stripped by a
less-sorbing gas such as nitrogen. Although
there is some loss of nitrogen to coal, most
of the more valuable methane can be recovered
from a coal seam by means of nitrogen
injection, without necessarily depleting the
total reservoir pressure.
4. Modeling studies show that methane recovery
rates from a field can be significantly accel-
erated by means of nitrogen injection. Fur-
thermore, nitrogen stripping can be expected
to recover significantly more gas than reser-
voir pressure depletion alone.
S. Pilot testing is recommended to reduce some of
the technical and economic uncertainties asso-
ciated with full field implementation.
METRIC CONVERSION
°c (OF-32)/1.8
cm 3 /g 32.037 SCF/ton
kPa 6.8948 psia
m 0.3048 ft
MPa 0.0068948 psia
REFERENCES
1. Ayers, W. B., and Kelso, B. S., "Knowledge of
Methane Potential for Coal bed Resources Grows,
But Needs More Study," Oil & Gas Journal,
p. 64, Oct. 23, 1989.
2. Ruthven, D. M., Principles and Practice of
Adsorption and Adsorption Processes, 1984,
pp. 336-342.
3. Yang, R. T., Gas Separation by Adsorption Proc-
esses, 1987, pp. 202-204.
4. Puri, R., and Stein, M. H., "Method of Coal bed
Methane Production", U. S. Patent 4,883,122.
S. Seidle, J. P., and Arri, L. E., "Use of Con-
ventional Reservoir Models for Coal bed Methane
Simulation", CIM/SPE 90-118, Presented at the
Petroleum Society of CIM/Society of Petroleum
Engineers International Technical Meeting,
June 10-13, 1990, Calgary, Alberta, Canada.
 • Reduce Cleat Pressure by Producing Water
• Methane Desorbs from Matrix and Diffuses to Cleats
• Methane and Water Flow to Wellbore

100 200 300 400 500 600

Pressure. psla

Rgure 1 Gas Holding Capacity of Coal is Pressure Dependent Rgure 2 Current Coal bed Methane Recovery Mechanism

• Inject Nitrogen in Cleats

• Keep Total Cleat Pressure High
• Reduce Partial Pressure of Methane
• Methane Desorbs from Matrix and Diffuses to Cleats
• Methane and Nitrogen & Water Flow to Wellbore
300

100 200 300 400 500 600

Pressure. psis

Rgure 3 Reservoir Pressure Depletion Limits Gas Recovery Rgure 4 Enhanced Coal bed Methane Recovery Mechanism
Rates & Reserves
 600 1000
LANGMUIR ISOTHERM LANGMUIR ISOTHERM

V = V.[b P/(l + b P)j 900 V = V.[b P/(l + b P)j

500
800
,.-..
------
"'6 0 700
0 0
0 400 U
>-
L
>-
L-
600
0 a
'-"
'-"
c 300 c 500
0 0

S- ~ 400
Legend
O Legend U 23.9% MIN MAT V.=797 scr/lo" b=0.00203 psi"
If) 200 Vl
+ AOSORP 100r. METHANE +
-> X OESORP 1007. "ETHANE
-> 300
o
AOSORP 100r. METHANE

100 o
V.=622 scrflo" b=0.0045S psi-' 200 o MIX DESORP ,I
MIX OESORP 10 1007. METHANE
48.77. METHANE
MIX AOSORP II 13.87. METHANE 6. "IX DESORP 12 22.6% METHANE
o MIX AOSORP 12 43.3% METHANE 100 o MIX DESORP 13 16.57. METHANE
o MIX AOSORP P 68.8% METHANE \.l "IX DESORP 14 13.6% METHANE

100 200 300 400 500 600 700 800 900 1000 600 900 1200 1500 1800 2100 2400 2700
P(Total or Partial), psia P(Total or Partial), psia

FIGURE 5. METHANE SORPTION ISOTHERM FOR SAMPLE A FIGURE 6. METHANE SORPTION ISOTHERM FOR SAMPLE B

:g

1000
LANGMUIR ISOTHERM
900
V =VM[b P/(1 + b P))
GAS
800 CHROMATOGRAPH
,.-..
0 700
0
U
>-
L- 600
a
'-"
c 500
0
~
u:: 400
U
Vl NITROGEN
NITROGEN +
-> 300
Legend FLOW-IN METHANE - .
c:!:!.....Y.=797 ~~:.£QIQ3~'
FLOW-OUT
200
o CH, AOSORP DATA 7.10% MOIST
100 N, V.=561serflo" b=0.000603 psi-'

0
o N, ADSORP DATA 7.06% MOIST
SINTERED FILTER SINTERED FIlTER
0 300 600 900 1200 1500 1800 2100 2400 2700
P, psia

FiGURE 7. NITROGEN SORPTION ISOTHERM FOR SAMPLE B FIGURE B. EXPERIMENTAL SETUP FOR INERT GAS STRIPPING
OF FINELY GROUND SAMPLE
 ocm:o 7000

-----------------~--------------------------.--------- ___________________ J
CL

o tn 6500
0.9
c o ~ 6000 SORBED METHANE
898 CC STP
:2
()
0.8
o -0
Q)
5500 -------------~~b-------------------------l-----------------------------

u
o ::J 5000
Lt 0.7 o -0
o
Q) e 4500
o
76% RECOVERY
OF SORBED METHANE
o 0.6 o CL
Q) 4000 o
~ c
Q) 0.5 o -<=
o 3500 o
c Q;
o
2 0.4
o :E 3000
FREE METHANE
Q; o .~
Q)
2500 o 5541 CC STP

~ 0.3 (; 2000 o
:;
"<t
I 0.2 E 1500 o
u ::J
o
>-
0.1
U
.
J:
U
1000

500
0
>
04-----,-----,------,--~~~~L,-----.----_,----_, 0
o 25 50 75 100 125 150 175 200 0 25 50 75 100 125 150 175 200
t N2 , Cumulative Nitrogen Injection Time, min t N2 , Cumulative Nitrogen Injection Time, min

FIGURE 9. METHANE CONCENTRATION FROM INERT GAS STRIPPING OF FINELY GROUND SAMPLE FIGURE 10. CUMULATIVE METHANE PRODUCED FROM INERT GAS STRIPPING
OF FINELY GROUND SAMPLE

0.9
c
0
§
:;: 0.8 0

,I -------____________ _ _ ()
0
Lt 0.7
8
0
/ GLASS BEADS __
Q)
0 0.6 §
/ ~ 3469 MIN 5599 MIN
SHUT -IH
Q) 0.5 -foIIf-Y-------- CO NTI NU0 US ---------I.~ BEFORE
c INJECTION BLOWDOWN
0 <is
J:: 0.4
NITROGEN +- 2407 MIN

1
Q)
NITROGEN SHUT -IN Legend
FLOW-IN MET~ANE --.. ~ 0.3
flOW-OUT :;;. 976 MIN 0 PERIOD 1
I
u 0.2 3722 M~NHUlT-IN + PERIOD 2
>- SHUT-IN 0 PERIOD 3
0.1
+. x PERIOD 4

COAL CORE 0 L-~~~~~~~~

0 1000 2000 3000 4000 5000 6000
t N2' Cumulative Nitrogen Injection Time, min

FIGURE 11. EXPERIMENTAL SETUP FOR INERT GAS STRIPPING FIGURE 12. METHANE CONCENTRATION FROM INERT GAS STRIPPING
OF WHOLE CORE SAMPLE OF WHOLE CORE SAMPLE
 14000
Il... BLOWDOWN + RESIDUAL
l- 388 CC STP
V>
U
U
12000 -----------f- ----------------------- ------------------------_-----L---._
-0 SORBED METHANE ~
II>
()
:J 10000
5705 CC STP ~ -- I I

"0 86%
0
'- RECOVERY
Il...
4883 CC STP
II> 8000

~-
c
0
..c
Q; - - --- - ---- - - -- - - - - - - - -- - - -- - -- - - - - - - - - - ------- - - - - - - - - - - - 96%
~ 6000 MASS
BALANCE
II>
> 11672 CC STP
:;=
0 4000 Legend
:J
E
FREE METHANE
6401 CC STP
o PERIOD 1
:J
U + PERIOD 2
2000
~
:I:
o PERIOD 3
()
> X PERIOD 4
0
0 1000 2000 3000 4000 5000 6000
t N2 , Cumulative Nitrogen Injection Time, min

FIGURE 13. CUMULATIVE METHANE PRODUCED FROM INERT GAS STRIPPING

OF WHOLE CORE SAMPLE

20~----------------------------------------------------~
1~~----------------------------------------------------~

It
,
~ II
0 15
Ii: II
()
II)
100 I~
:IE
:IE If','.
.i
II:
C
Run #1: Pre'lure Depletion Only i
t;.
10
I f
,'1.
I
Run #1: Prauura Depletion Only
0
·u o !!.\In.l2.i..NUI~dwg ~ft!! 8_Vr! 'tJ ,
, 0 !!.\In.l2.i.Nt.lne.ct.i2n~ft!! 6_Vr!

"0 £
'tJ

.•
It ~

I
• ",. """ .
CI
iii
;§ I
~ •

0
0 10 12 14 18 18 20 22
O~---T~--~--~--~----~---r----~--~--~----~--~
o 10 12 14 18 18 20 22
Time. Yea,s Time. Yoa,s

FIGURE 14: FULL FIELD TOTAL GAS PRODUCTION RATE FIGURE 15: FULL FIELD WATER PRODUCTION RATE
 loo~---------------------------------------------------------------, loo~---------------------------------------------------------------,

80 80 ~
~.-_j
..
. .0.....,.,.•.
...,
. ~

~ .~.
• ""-0
J'•
c: >-
OJ
CD
CO Run .1: Pra8lur. Deplation Only c
g eo il: eo
Z
E
CD
o !!!Jn~2~NtJnt!.ct!2n .aft!! tL,Y'! 0
Ul
::E
::E
,•"
!:! Run #1: Pr8aaure Depletion Only
:. 40 ! 40 • o ~n.l2.;.NtJn~cr!2n .eft!,!; ,_Y,!
,•
CD
E a:
.2
~ ..
CD
c:
.c:
a; I
::E
20 20
"
"
10
Time. Yea,s
12 1. Ie 18 20 22 o 10
"
0+---~----~---'----~--~----~--~~--~----~--~---4
12 18 18 20 22
Time. Yea,s
,.

FIGURE 16: NITROGEN PERCENT IN PRODUCED GAS FIGURE 17: FULL FIELD METHANE PRODUCTION RATE

600,---------------------------------------____________________-,

400
38% Gas-in-Pieca
RECOVERED
U.
0
III

.,; 300
!
•>
0 Run #1: Pre.lur. Depletion Only
u
CD
a: o !!!,n..l~NUnl!.ct12n ~ft![ fLY'!
..
CD
c: 200

~
::E
E
:>
0 100

o~==~--~----~--~--~----~--~--_r--~----~--~
o 4 10 U M " ,. W u
Time. Yea,s

FIGURE 18: CUMULATIVE METHANE RECOVERED FROM FIELD