You are on page 1of 6

[Society of Petroleum Enfrheera I

SPE 36986

Relative Permeability in Coal


Kevin Meaney, SPE, and Lincoln Paterson, SPE, CSIRO Petroleum

Ccpynght 19SS, Scaety cd Pafro!eum Engme-scs, Inc results ffom laboratory tests to field applications. Applications
TM papa was prepared fcf prescmtakm at the 1996 SPE ASWIPacfiIc 011 .S Gas Confererca occur in coalbed methane and gas drainage from coal seams
he!d m Addade, Australm 2&31 Octcbar 1996
for mine safety. 1This work is also relevant to multiphase flow
TM. pspar was selacfad for presanlatlon by an SPE Program Commmea folbwmg revmw of in other types of fractured rock.
mfofmatlon contamad m an abstract submttd by the author(s) Contents of the pape$, as
preserded, have not bean revmwad by tb SOOety of Petro!aum Englneam ard are subjscf to
carrocf!on by the author(s) The matanal, as presented, dam not necassardy refled any
positicn of tfm %cmly of Pefroleum Engmeors, its oli%xrs, or membsrs Papars presantsd al
SPE meetmgs am subpct to pubbcatton revmw by Edkmal Commdtees of the Souety cd Reservoir Properties of Coal
Petroleum Engmaers Parmissmn 10 copy IS restrcted to an abslrad of not more than 300
words Illus! rationo may not be copied The abslract should contain conspicuous Coal is generally treated as a dual porosity rock containing
acknmadedgment of where and by vdmm the paper was pfeaanted Write Lbranan, SPE, P O micropores and a network of natural fractures. The fractures
Box 833836, Rtiardson, TX 7Y3f33-3826 U.S A, fax Ot-214-9S2-9435
are known from mining as cleats. Some basic reservoir
engineering aspects of coal have been described by Gray.2’3
Coal is arguably the most difficult reservoir rock to work
Abstract with for the reasons listed below.
Relative permeability is a primary parameter for determining 1. Coal tends to be friable and very heterogeneous, so
the reservoir behaviour of coal seams. We present numerous sample selection is a major issue. Procedures for sample
relative permeability curves measured in the laboratory as well select ion have been recommended by Hyman et al.4’s
as curves from numerically matching field performance. 2. The connected fracture network of coal has low porosity.
Viscous fingering provides an explanation for the high values Laboratory testing therefore requires accurate measurements
of residual water saturation. of very small amounts of fluid.
3. Very pronounced stress-dependent permeability is
observed in coal.z Furthermore, as the stress load is cycled,
Introduction permeability tends not to return to the original value. The
Relative permeability is a primary parameter in determining reason for this has not been clearly identified, but may include
the response of coal-seam reservoirs. Despite the importance fines migration.
of relative permeability, measurements are infrequent and very 4. Many gases strongly adsorb on coal, including methane,
few recorded examples appear in the literature. The main carbon dioxide and nitrogen.
reason for this is that the properties of coal make the
measurements difficult.
In the published literature, discrepancies exist between Previous Measurements
relative permeability curves from the laboratory and the field. Recorded measurements of relative permeability in coal are
High values of residual water saturation, often in excess of sparse. It has only been in the last few years that direct
80’XO, are observed. measurements of water relative permeability during gas
In this paper we present numerous relative permeability injection have been reported. We reproduce below some of the
curves measured in the laboratory as well as curves from previously published curves. To assist visual comparison, they
numerically matching field performance. Some of the have been replotted with the same aspect ratio and with water
discrepancies in the results are postulated to be a result of the saturation on the abscissa axis.
unfavorable mobility ratios that exist when gas displaces In the early 1970s Rezrrik et al.6 reported on measurements
water. This leads to viscous fingering. Evidence to support this of relative permeabilities under steady-state conditions, with
conclusion is provided in X-ray CT images of core water saturation both increasing and decreasing. With
displacements. decreasing water saturation they were unable to measure water
Viscous fingering behaviour is very dependent on relative permeabilities, and instead computed values from the
heterogeneity. The extremely heterogeneous nature of coal is corresponding gas permeabilities by Corey’s method. They
interpreted to be primarily responsible for the variability in acknowledged that the applicability of this approach to coal is
relative permeability curves. We provide examples of how this tenuous. Relative permeability curves for decreasing water
is manifested in practice. Viscous fingering provides an saturation are shown in Fig. 1a. The relative permeability to
explanation for the high values of residual water saturation. water is shown as a broken line because it was not directly
Conclusions from this paper are important for scaling measured Different samples were used to compile the water

231
2 KEVIN MEANEY and LINCOLN PATERSON SPE 36986

and gas relative penneabilities. Samples (h) and (i) had a particularly significant fracture
Jones, Bell and Schraufnage17 have presented a relative parallel to the flow. Sample (a) was extensively fractured,
permeability curve for Fruitland coal (Fig. lb). Relative sample (b) had many mineralised fractures, sample (c) had a
permeability to gas was determined by the stationary-fluid well developed fracture network, and sample (e) had no visible
technique. Relative permeability to water was not directly fractures.
measured, but was deduced from the capillary pressure curve.
Based on this, it was noted that the gas relative permeability
would never exceed 0.1 at any stage during the producing life
of the reservoir. Field History Matching
Hyman et al. 5 have plotted gas and water relative History matching normally involves using a finite difference
reservoir simulator to reproduce gas and water production
permeabilities of Fruitland coal determined fi’om steady-state
data. Simulators have been developed specifically to account
tests (Fig. lc). They also present results from unsteady-state
tests and modified porous plate tests. for the complex characteristics of coal. For the work presented
Puri, Evanoff and Brugler8 tabulate and plot some relative here we used the code SIMED II. ““’s
permeability measurements as a function of mobile water We determined relative permeability in the Bowen and
Sydney basins by history matching field production data. All
saturation. Values from tables in their paper are plotted in
history matches shown here were conducted vertical wells,
Figs. Id and le as a fimction of total water saturation, in the
although we have obtained consistent results from horizontal
conventional sense. Here, total water saturation was calculated
boreholes drilled from mines.
by adding irreducible water saturation to the mobile water in
Relative permeability curves from history matching of field
the cleats.
Young’”” reports where the relative permeability curves production are shown in Fig. 4: (a) and (b) are tlom the Bowen
Basin; (c), (d), (e) and (f) are from the Sydney Basin. The
derived from a history match do not resemble laboratory
measured curves (Fig. 2). Young states that this is because curves in Fig. 4a are from the DR4 well, and also appear in
Ref. 16. The curves in Fig. 4d are from the same hole as Fig.
laboratory measured relative permeability obtained from small
3g. The curves in Fig. 4b are from the same seam as Figs, 3d
conventional core plugs do not adequately reflect the presence
and 3e, namely the German Creek seam,
of natural fractures. In addition, there may have been gas-
water gravity segregation. Comparison over the range of results shows that the field
Experience from the reported work has shown that testing and laboratory measurements are consistent within the sample
to sample variation. We conclude that matching heterogeneity
of small core plugs (say 25 mm diameter or less) usually gives
may be important if laboratory measurements are to be
unsatisfactory results, When extracting cores for testing, cores
representative of field values.
often disintegrate. This means that strong less fractured cores
are more likely to reach the testing laborato~. This problem is
accentuated for smaller cores.
Viscous Fingering
At discovery, it is generally assumed that the entire fracture
network in coal is saturated with water and there is no free
Laboratory Measurements
gas.’ As the reservoir pressure is reduced, methane and
Laboratory measurements of relative permeability in
Australian coal have been conducted in our laboratory. Details possible other gases desorb from the coal surface, diffuse
through the micropores, and flow to the wellbore via the
of the laboratory equipment have been reported elsewhere. ”
We used transient tests with gas displacing water to fracture network. At this last stage, gas is displacing water
measure relative ~ermeability, and analysed the results using from the fractures.
the JBN method.’ ‘“ The JBN method does not allow the full Viscous fingering occurs during flow in a porous medium
when a more viscous fluid is displaced by a less viscous one,
curves to be determined, because the Welge truncation means
that high water saturations are omitted. When relative such as when high viscosity water displaces low viscosity gas,
permeability to water becomes small, water production from Instead of a stable, planar displacement front, an uneven,
fingered front occurs.’7 These viscous fingers propagate
low-porosity coal becomes slow and difficult to accurately
measure, truncating the high gas saturation end of the curves. rapidly, causing early breakthrough and poor displacement
We used core samples 50 mm in diameter and at least 120 mm eftlciency. Injection of gas into the narrow space between two
parallel glass plates is shown in Fig. 5. The growing interface
long. Most of the samples displayed mineralisation in the
fractures. is not circular, but fingers of injected gas can be seen to be
Relative perrneabili~ curves from laboratory testing of extending from the central injection point. The parallel plate
Australian coal are shown in Fig. 3: (a), (c), (d), (e), (h), (i) apparatus serves as a simple model for a fracture, such as the
fractures found in coal. In homogeneous rocks, the severity of
and (j) are fkom the Bowen Basin; (b), (t) and (g) are from the
northern Sydney Basin. Variability in the results can be seen, the fingering is determined by the viscosity (or mobility) ratio
of the fluids.
which we attribute to variable heterogeneity in the samples. Of
the results in Fig. 3, only the curves in (h) were from a core The presence of heterogeneities dominates the growth and
development of viscous fingers,’8”g In reported studies of
with horizontal orientation, all the others were vertical cores
heterogeneous models, flow was largely determined by the
that crossed the bedding planes. Samples (f), (g), (h), and (i)
patterns of heterogeneity. Consider, for example, a porous
all had obvious fractures parallel to the flow direction.

232
SPE 36986 RELATIVE PERMEABILITY IN COAL 3

medium with uniform permeability except for a high 5. Hyman, LA. er al.: “Advances in Laboratory Measurement
permeability layer. Even if the fluids have the same viscosity, Techniques of Relative Permeability and Capillary Pressure for
there will be preferential flow in the high permeability layer, Coal Seams,” Quarterly Review o~ Methane from Coal Seams
Technology (Jan. 1992) 9-16.
and the heterogeneity will cause early breakthrough and
6. Reznik, A.A et a/.: “Air-Water Relative Permeability Studies of
reduced displacement efficiency. If the injected fluid has lower
Pittsburgh and Pocohontas Coals,” SPtiY (Dec. 1974) 556-562.
viscosity than the displaced fluid, preferential flow in the high 7. Jones, AH,, Bell, G.J. and Schraufnagel, R.A.: “A Review of the
permeability layer will be enhanced. Physical and Mechanical Properties of Coal with Implications for
An X-ray CT slice through a 50 mm diameter coal core Coal-Bed Methane Well Completion and Production,” 1988 Coal-
saturated with a sodium iodide solution is shown in Fig. 6. The Bed Methane, San Juan Basin, Rocky Mountain Association o!
sodium iodide solution, appearing dark, highlights the fracture Ceo/ogis/s, 169-182.
structure. The same coal sample is shown in Fig. 7 after gas 8 Puri, R., Evanoff, J.C., and Brugler, ML.: “Measurement of Coal
injection, Gas appears white, and has occupied portions of the Cleat Porosity and Relative Permeability Characteristics,” paper
larger fractures, The inefficient displacement efficiency is SPE 21491 presented at the 1991 SPE Gas Technology
Symposium, Houston. Jan, 23-25.
reminiscent of the viscous fingers observed in the parallel plate
9. Young, G. B.C.: “Coal Reservoir Characteristics from Simulation
model (Fig. 5), and is consistent with gas displacing water.
of the Cedar Hill Field. San Juan Basin.” Ouar!erlv Review of
The gas quickly occupies portions of the larger fractures, Methane from Coal Searns Technology (July T992) 6-~0. “
bypassing large regions of water. This water becomes 10. Young, G. B.C., Paul, G. W. and McBane, R.A,: “Cedar Hill and
effectively trapped, providing a plausible explanation for the Tiffany: Case Studies in Coalbed Methane Reservoir Simulation,”
high observed values of residual water saturation. paper presented at the 1992 Symposium on Coalbed Methane
Gravity can stabilise viscous fingers. If gas displaces water Research and Development in Australia, Townsville, Nov. 19-21.
vertically downwards, the density difference can keep the 1I. Paterson, L., Meaney, K. and Smyth, M.: “Measurements of
fluids partitioned, providing the displacement is slow enough. Relative Permeability, Absolute Permeability and Fracture
Similarly, in horizontal flow, gravity override can have a Geometry in Coal,”’ paper presented at the 1992 Symposium on
Coalbed Methane Research and Development in Australia,
similar effect to viscous tingering, resulting in bypassed fluid.
Townsville, Nov. 19-21,
The full implications of viscous and gravity instabilities need 12. Johnson, E.F., Bossier, D.P. and Naumann, V.O.: “Calculations of
further consideration, and may be related to the efficiency of Relative Permeability from Displacement Experiments,” Trarrs,
gas production. AfME ( 1959) 216,370-372.
13. Tao, T.M. and Watson, A.T.: “Accuracy of JBN Estimates of
Relative Permeability: Part l–Error Analysis,” SPEU (Apr, 1984)
Conclusions 209-214.
[. Numerous relative permeability curves for Australian 14. Stevenson, M. D., Pinczewski, WV., Meaney, K. and Paterson, L.:
“Coal Seam Reservoir Simulation, “ The Australian Petroleum
coal measured in the laboratory as well as curves from
Exploration Journa[ ( 1994) 34, I I4-120.
numerically matching field performance have been recorded.
15. Mcaney, K,, Paterson, L., Stevenson, M.D. and Pinczewski, W. V.:
2. The field and laboratory measurements are consistent “Advances in Coal Seam Reservoir Simulation for Mine Gas
within the range of sample to sample variation. The sample Drainage,” F’roc., Int. Symposium cum Workshop on
variation is attributed to the heterogeneous nature of coal. Management & Control of High Gas Emissions & Outbursts,
3. Matching laboratory to field heterogeneity may be Wollongong ( 1995) 277-282.
important if laboratory measurements are to be representative 16. Weld, M.B. et al,: “Cavity Completion for Coalbed Methane
of field values. Stimulation - An Integrated Investigation and Trial in the Bowen
4. Water bypassing by viscous fingering provides an Basin, Queensland,” paper SPE 30733 presented at the 1995 SPE
explanation for the high values of residual water saturation Annual Technical Conference and Exhibition, Dallas, Oct. 22-25.
observed. 17. Paterson, L.: “Radial Fingering in a Hele Shaw Cell, “ J, Fluid
~eCh. (1981) 113, 513-529.
5. The suppression of viscous fingering by gravity
18. Araktingi, U.G. and Orr, F. M,: “Viscous Fingering in
stabilisation may be of significance in field applications. Heterogeneous Porous Media,” SPEAT(Apr. 1993)71.
19. Brock, DC. and Orr, F. M,: “Flow Visualization of Viscous
Fingering in Heterogeneous Porous Media,” paper SPE 22614
References presented at the 199 I SPE Annual Technical Conference and
Exhibition, Dallas, Oct. 6-9.
1. Paterson, L.: “A Model for Outbursts in Coal. ” Int. J. Reek. Meclr
Min. Sci. & Geomech. Abstr. ( 1986) 23,327-332.
2, Grav. I.: “Reservoir En~ineerin~ in Coal Seams: Part l-The
Phy;ical Process of Gas ;torage a;d Movement in Coal Seams,”
SPEfW (Feb. 1987) 28-34.
3. Gray, [.: “Reservoir Engineering in Coal Seams: Part 2-
Observations of Gas Movement in Coal Seams,” SPERE (Feb.
1987) 35-40.
4. Hyman, LA. et al.: “Simultaneous Determination of Capillary
Pressure and Relative Permeability in Coalbed methane
Reservoirs,” paper presented at the 1991 Coalbed Methane
Symposium, Tuscaloosa, May 13-16.

233
4 KEVIN MEANEY and LINCOLN PATERSON SPE 36986

1 1

~1
0,8 i 0.8
i
i
~“~ (a) i ;’: (b) i
i
0.2 /“ 0,2
~~ /’
/, )’
o 0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8

L
0.81 I 0.8
- — -. Field History Match

I----2d
— Laboratory
;’; (c) ;’; (d) 0.8

0,2 0.2

0 0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1

1
0.8

~“~ (e)

0.2
E0
0.2 0.4 0.6 0.8 1

Water saturation Water saturation

Figure 1, Previous curves for relative permeability In coal. (a) Figure 2. Relative permeability in coal, redrawn from Ref 9.
redrawn from Ref. 6; (b) redrawn from Ref. 7; (c) redrawn from Ref.
s (d) drawn from data in Table 1 of Ref. 8; (e) drawn from Table 2
of Ref. 8.

234
SPE 36986 RELATIVE PERMEABILITY IN COAL 5

1 1

u
0.8 0.8

:“;(b) :“; (c)

M
~~ (a) ;“: (d)

0.2 0.2 0.2 0.2

n0 0 0 1 m
0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1
.2

::;
=
n
cd
1
‘~ k——————
a) 0.8 0.8 0.8

A
E

L k
~ 0.6
Q. ;’ (f) (h)
al 0.4 ‘e)
>
“% 0.2 0.2 0.2
6 1
Co E 0L 0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1

o.:r____l 0.8

I---__J
~“~
0.2

0
(i)

0.2 0.4 0.6 0.8 1


:’:

0.2
Eo
0)

0.2 0.4 0.6 0.8


L
1

Water saturation

Figure 3. Relative permeability In coal from transient laboratorytesting

1 1 1
0.8 0.8 0.8

0“6 (a) :’: (b) :“; (c)


@ 0.4
=
~ 0.2 0.2 0.2
b! Ill
0 n0
ED
h 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1
n
1 1 1
0.8 0,8 0.8

:“; (d) ~~ (e) :“: (f)

0.2 0.2 0.2


D0 0D n0
0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1

Water saturation

Figure 4. Relative permeability in coal from history matching of field production.

235
6 KEVIN MEANEY and LINCOLN PATERSON SPE 36986

am?.

Figure 5. Viscous fingering. Gas displacing liquid in the narrow


space between two parallel glass plates, from Ref. 17.

Figure 6. X-Ray CT slice through a 50mm diameter coal core Figure 7. Similar to Figure 10, except after gas injection. Gas
saturated with a sodium iodide solution, highlighting the fracture appeamwhite, and hasoccupied potiions of the larger fractures.
structure.

236

You might also like