SPE

Socset+Jof Petdeum

Engineers

SPE 21492
Analysis of Slug Test Data From Hydraulically Fractured Coalbed
Methane Wells
J.A. Rushing, T.A. Blasingame, B.D, Poe Jr., ROM.Brimhall, and W.J. Lee, Texas A&M U,
SPE Members

>opyright 1991, Society of Petroleunl Engineers, Inc.

rhie paper wae prepered for preaenlation al the SPE Ciae Technology Sympoeium held in Houston, Texas, January 23-25, 1991.
rhis paper waa eelactnd for presentation by an SPE Program Committee following rewew of reformation contained m an absfract submmed by !he author(s). Contents of the paper,
1s presented, have not been reviewed by the Society of Petroleum Engineere and are subject to correcl!on by the author(e). The material, as presented, does not necessarily reflect
my poeition of the Society of Petroleum Engmeera, its offm.era,or membere. Papera presented af SPE meetings are subject to publication review by Edilorlal CommNees of the Soc!ety
)1Petroleum Enginears. Permission to copy is restricted to an abstract of not more lhan 303 worde. lllustraho~a may not be copied. The abstract should conte!n conspicuous acknowladgmem
]f where and by whom th9 paper is presented. Write Publication Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836 U.S.A. Telex, 730989 SPEDAL.

ABSTRACT
This paper presents new type curves for analyzingslug tests in
hydraulically fractured coal seams. The type curves were
developed using a finite-conductivity,vertical fracture model and
are presented in terms of three parameters -- dimensionless
welltme storage coefficient, dimensionlessfracture conductivity,
and fracture-face skin. Whh these new curves, we may estimate
the hydraulic fkacturehslf-length, the fomlation permeability,and
the fracture conductivity. We also present a procedure for using
the new curves and illustrate the procedurewith an example,

viable producers, Therefore, these conventional slug test analysis

techniques cannot be used to either assess the success of the
fracture treatmentor to evaluatethe px.t-fracture potentialof these
stimulated coal seams. Karsaki, et al,g studied the pressure
response of slug tests in infinite-conductivity vertical fractures,
but they did not investigate the behavior of finite-conductivity
fractures. The purposes of this paper are to develop a male] for
slug testing in coal seamswith finite-conductivityverticalfractures
and tc illustrate application of this model to the analysis of slug
;ests,

INTRODUCTION
MATHEMATICAL MODEL
Slug testing has been proven to & an effective method for
characterizing the production potential of coal seams. A slug test
involves the imposition of an instantaneouschange in pressure (m
fluid head) in a well and the measurementof the resulting change
in pressure as a function of time. This change in pressure is
created by either injecting into or withdrawing from the well a
specific volume of fluid (i.e., a slug), From this measured
pressure response, we may estimate the permeability and nearwelltmreconditions.

We developed our slug test model using Cinco-ky, et al.sg

model which considers a well intersected by a fully-penetrating,
finite-conductivity, vertical fracture. The reservoir is assumed to
be an isotropic,,homogeneous, infinite medium having a uniform
thickness, h, permeability, k, and porosity, @. In addition, the
rescrvoir contains a slightly compressible fluid of viscosity, p,
and compressibility,c, that are independentof pressure.

Initially, slug testing methods and analysis techniques were

developed for estimating the transmissivity of shallow,
underpressured aquifersl-s, but also have found applications in
the petroleum industry, especial)j for analyzing the flow period
during drill stem tests4-6. Recently, slug testing has been
extended to the evaluation ot the production potential of coal
seams7. Since most coal seams are saturated initially with water,
slug testing provides a simple but effective method for estimating
flow properties early in the productive life before the initiation of
gas production.
Reference 7 provides an overview of
conventionalslug testing in coal seams.

Cinco-Lcy, et alsg modci assumes the fracture to be a

homogeneous, uniform slab with height, h, width, bj and half
length, L Because the fracture width is much smaller than
fracture {ength and height, the model assumes the flow in the
fracture is linear and that fluid influx at the fracture tips is
negligible, In addition, the model assumes that fluid production
from the reservoir to the wellbore occurs only through the
fracture. Further, since the fracture volume is small, the model
neglects the fracture compressibility and assumes flow within the
fractureis relativelyincompressible. Additionaldetailsconcerning
the model formulation and problem solution may be obtained in
Ref. 9 and Appendix A.

Conventional slug test analysis techniques are based on radial

flow models, However, many wells completed in coal seams
require hydraulic fracturing in order to become economically

Under these conditions, Cinco-Ley, el al.g derived art

expression for the dimensionless pressure drop at the wellbore
(i.e., XD = O) during constant rate production from a WCII
intersectedby a finite-conductivityfractureas

Referencesand illustrationsat end of paper

---

Analysisof SlugTest Data From HydraulicallyI%ctured CoalbedMethaneWells

?@(xD=o,@

SPE 21492

Pm

~wD(s) =

1 + S2cL~

~@S) ,. .,, ,., .,..., ..,,.,,.,,.,..

.,, . (6)

where the wellbore storage coefficient based on fracture halflength is

FCD
-+

J
o

lJ

f?fdxttLjD)

dx ~ +~...,,,,,,

(1)l)

where qp(x,t) is the dimensionless flow rate, per unit of fracture

length, of fluids entering the fracture from the reservoir. In
addition, the dimensionless time based on fracture half-length, Lf,
is defined as
tL~ = ~

and C is the wellbore storage coefficient for a changing liquid

level,
The conventional pressure plotting function for production
slug tests is aefined as
pDsIu8

4/4+? ............................,.,.,,,.......(2)

and, the dimensionlessfracture conductivity, FCD, is defined as

. T
&,hLJ . .... . ... ..... .... ... ... ,, .,, ,. ...,. . ... ... ... . (7)

=P-P(O
;i . P.

and, for an injection slug test,

p(t) - p[
pDslu~ = P.. pi

FCD . @

. . . . . .. . .. . . , ..,,,...,.,

,.,,.,,.,,..

....,..,,,,,

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .,, . (8)

...!.. o. ... ...! . . ... ... ....! . ...4... ....o.!...(9)

.,. ,(3)

If we assume the fracture is symmetric and homogeneous, then

qp{xDJ)
= Q@-xo,t).
Making this substitution into Eq. 1 and
taking the Laplace transformyields

where pi is the original reservoir pressure, p. is the wellbore

pressure prior to the beginning of the slug test, and p(t) is the
pressure measured during the test. According to Ramey and
Agarwal13,Eqs. 8 and 9 also define the annulus unloadingrate,

F/z)(s)=

pDsIug =@&

. . . ..! . . . . . . . . . ..! . . . . . . . ..4 . . . . . . . . . . ..(10)

which may be computedin Laplacespace as

XD

._L

FCD

~Ds/ug(d = SCL@ ~wD(s) . . . . . . . . . . . . . . ,,, ., ..,.,,.,,

~~
H

qfL)(X,S) (i.x(ix + ~FcD

%
... . .. . ... ....(4)

,,, . .

(11)

Following Ramey and Agarwal13,we also define a dimensionless

sandfacerate as

If we include a zone of altered permeability around the fracture

face, Eq, 4 becomes

qDdug = 1- cL~~

dp

, .,, ,, .,,.,, . . . . . . . . . . . . ,,, ..,,.,..,

(12)

F@(s) =

or,

~f ijfD(X,s)dX
/o

qDslug = 1 pDslug . . . . . . . . . . . . . . . . . . .. . . . . .,, .,,,,..,,

Xfj

.x

FcD

X
@S(X,S) dtsix+ -

~*
H

(5)
SFCD.

where the damage to the fracture face is qwtntified by ispositive

skin factor, $.
Recently, Cinco-Ley and Mcnglo and Mechan, et al,l 1
presented a method for solving Eqs, 4 and 5, The fracture halfh.mgthis divided into n discrete elements, and each element is
modeled as a uniform flux fracture, Since the flux is unknown,
wc write an equation describing the pressure and flux distribution
in each element, and from this system of equations, we solve for
the wellbore pressure and flux in Laplace space. We then
compute the real space solutions using the Stehfest inversion
algorithmlZ Details of this solution technique are presented in
Rcfs, 10 and 11 and Appendix B,
Before inverting the system of equations and solving for the
dimensionless pressure at the wellbore, wc include wellbore
storage as follows:

. ....

(13)

The dimensionless sandface rate defined by Eqs, 12 and 13 is

useful as a slug test plotting function to provide better resolution
of very early data,
Sl,UG TEST PRESSURE RESPONSE
In this section, we investigate the effects of various fracture
and reservoir parameters on the slug test pressure response,
Cinco.Lcy, et al.g have described four distinct flow periods that
may occur in hydraulically fractured wells -- fracture linear,
bilinear, formation linear, and pseudoradial flow. These periods
and their associatedflow patterns are illustrated in Fig, 1,
Fracture linear flow occurs at very small values of
dimensionless time, During this pericd, most of the fluid flow
comes from expansion of the fluid contained in the fracture, and
the flow pattern is essentially linear. Bilinear flow occurs when
fluid flows linearly from the formation into the fracture, During
this flow period, the fracture-tipeffects have not yet influencedthe
pressure response. For most practical values of dimensionless
time, bilinear flow appears only in fractures with FcD < 100z,
Following a transition period, the next flow pattern is formatio~,

SPE 21492

J.A. Rushing, T.A. Blasingame, B,D. PM, Jr., R.M, Brimhall, and W,J, Lee

linear flow, which occurs only in fractures with FCD> 10Qz(i.e.,

essentially infini\e-conductivityfractures). The final flow period,
pseudoradial flow, occurs with fractures of all conductivities.
However, the higher the conductivity, the later the flow pattern
can be characterizedas essentiallyradial.
Effects of Wellbore Storage Coefficient and Fracture
Conductivity. For hydraulically fractured wells in which skin
effects are negligible, the parameters controlling the pressure
response are the dimensionless wellbore storage coefficient and
fracture conductivity, The dimensionless wellbore storage
coefficient, defined by Eq, 7, is based on fracture half-length, In
this study, we investigate values of CL ranging from 104 to 1.
Small vahres of CLjD represent snlaf1 volumes of fluid in the
wellbore and/or long hydraulic fractures. Conversely, large
values of CL.. are typical of large volumesof fluid in the wellbore
and/or short fracture half-lengths, We also investigate the effects
of dimensionless fracture conductivities ranging from 0.1rr to
100z, As defined by Eq. 3, small vahsesof FCD represent long,
low-conductivity fractures and/or very permeable formations,
while values of FCD 2 100z are indicative of high or even infhrite
conductivityfractures.
Similar to the type curves presented by Ramey, et al.b, we
developed our type curves using rL@cL~
as the time Plotting
function, Figure 2 is a semilog plot ofpDs/ug vs. @CL~ for
FCD = 0.1 z and several values of CL . AU of the curves have
the characteristic S-shape exhibited r y radial flow models. In
fact, the curves do not display any distinct characteristics which
would allow us to distinguish them readily from curves generated
with radial flow models. Similarly, semilog plots showing the
pressure response-sfor values of FCD = n, 10z, and 100tr are
shown in Figs, 3-5, respectively, Note that as FcD increases, the
curve shapes become more distinct and are more sensitive to the
value of CLp. In addition, we observe that for all values of
dimensionless fracture conductivity, the level of essentially
constant or static pressure response is maintained for longer
periods for larger values of CLp, which suggests larger radii of
investigationare achieved.

plots may help reduce the ambiguity in type-curve matching.

Unfortunately, the small values of pmw and tL~/CL~ shown in
Figs. 10-13correspond to very small pressures changes and times
that may be impracticalto obtain flom fieldtests.
To estimate the time at which the slug test pressure reponse
reflects bilinear flow, we compared the senrianalytical solution
given by Eq, 4 to the approximate bilinear flow solution given by
Cinco-by, et al.ld,
~@(S)

= ~
S514~

+ ZS2cL,f)

.. ... .... ... ... . ... ... ....(14)

or, the slug test pressure response in Laplace space is

We then compute pDsIug by inverting Eq. 15 with the Stehfest12

algorithm. For the range of dimensionless times studied, we
observed bilinear flow patterns for all vahres of FCD Snd CLp,
but especially at the lower values of CLp, In general, for CLp 2
0.1, bilinear flow ends at tLp/CLp c 0.01 or much earlier for the
larger dimensionless fracture conductivities. At smaller values of
CL~, a significant portion of the pressure response represents
bilinear flow. For example, for FCD = 0.1z and CL = 104, the
end of bilinear flow occurs at approximately fL /P L~ = 1* and
as shown by Fig. 6, the pressure appears to r?c approaching a
unit-slope line within three or four log cycles after this time.
When FCD = 1O(Mand CL~ = 10-4, bilinear flow ends at
= 0,01; however, as shown by Fig. 11,
approximately tL~/CLp
the subsequent pressure reponse does not form a unit-slope line
for the range of tLp/CLp investigated, We suspect that much of
= 0.01 reflects fo~ation
the pressure behavior after tLp/cLp
linear flow. In general, we observe that the pressure responses
and large values of FCD are more
for small values of CL
sensitiveto the various t?ow patterns.
Cinco.Ley
and
Effects of Fracture-Face
Skin.
Samaniego17 have suggested that two types of fracture damage
may occur during the hydraulic fracturing process -- within the
fracture adjacent to the wellbore and in the formation around the
fracture face, The first type of damage, often described as a
choked fracture, is thought to be caused by proppant crwshingand
embedding in the formation. The second type of damage,
quantified as a fracture-face skin, is probably caused by fluid
losses into the formation, In this study, we address only the
effects of fmcturc-hcc damage on the slug test pressureresponse.
Appendices A and B describe the manner in which damage to the
fracture face is modeled as an infinitesimalskin with no storage.

The late-time pressure reponses shown in Figs. 6-9, which are

log-log plots of pDs/u vs. fLfD/CLfL3 as a function of Cr.jD,
illustrate that many oft %e curves form a unit-slope line at large
values of tLp/CL . From a comparison with the behavior of
slug tests with raifial flow, the appearance of the late-time units!ope line suggests that a stabilized pscudoradial flow period is
(cached. As we would expect, this pseudoradial flow period
iippt~s
sooner for small values of FCD, h arlditim we obseIve
that regardless of the value of FcD, the curves for CL~ z 0.1
converge to form a unit-slope line by v..fD/CLjD = 100,
Unfortunately, these same curves have similar shapes for the
range of fracture conductivites studied, which indicates that
unique type curve matches may be difficult to obtain for large
values of CLP (i.e., for short fracture half-lengths an~or We
wellbore storage cmfticicnts). Conversely, the curves for C/,p <
0,1 exhibit unique shapes for different values of FCD,

In a slug test, a damaged zone around the wellbore reduces the

rate of fluid flow into the wellbore and, therefore, tends to reduce
the rate at which the pressurechanges. We would expect a similar
reduction in fluid flow from the reservoir to the fracture when
fracture-face skin is present. However, when compared to the
curve shape for no skin, Figs, 14 and 15 show that the presence
of fmcture-face skin has little effect on the pressure response in
low-conductivity fractures, thus greatly reducing the character of
the curve shapes, Specifically,fracture-face skin factors less than
0,1 are indistinguishable from the zero skin case. In addition,
note that these skin effects are less pronouncedfor small values of
CLP and FcD (Fig, 15), which suggests that we cannot estimate
the fracture-face skin accurately in short, low-conductivity
fractures, The early-time effects shown in Figs, 16 and 17 for
= 1o-4 and 1,0, respectively, help
FcL) = O.ltr and CLp
distinguish the pressure response at early times, but the pressure
differences may be too small to be measured accurately with
conventionalpressure gauges.

The early-time prcwurc responses are illustrated in Figs, 10Por FCD =

13, which arc log-log plots of qlhlug Vs.tf.p/CLp.
0,1z and CLps 0.01, WC Observe a tsnit-slW line at very e~lY
103) For radial models Sa8eev6 has
times (i.e., tLp/CLp<
shown that in the absence of wellbore skin, the early-time
response also has a unit-slope prior to the final pressure buildup.
Note that when we use log-log plots of the early pressure
response, the turves for values of CL~ 20.1 exhibit slightly
more character thinswith scmilog plots, especially at large vahtcs
of FcD. Therefore, if the early data are availablci these 10&lo6
,-

lUI

.
SPE 21492

Analysisof SlugTest Data From HydraulicallyFracturedCoalbedMethaneWells

or, the hcture half-length~is

For highly-conductive fractures with small values of CL~

(Fig. 18), we see that fracture-face skin affects the pressure
behavior significantly. We note that larger values of skin tend to
maintain the static pressure response longer (i.e., skin reduces the
rate at which the pressure changes). The early-time pressure
responses for these same parametersare plotted in Fig. 19, As the
skin increases, wc observe the formation of the unit-slope line
earlier than when skin is not present. For high-conductivity
fractures with large values of CL~ (Fig. 20), we see that fractureface skin, especially values less than 0.1, has little effect on the
pressule response, Only the early-time pressure behavior is
affected by the presence of skin, as illustrated by Fig. 21. From
these plots, we may conclude again that fracture-face skin effec:s
are difficult to discern from slug tests in wells with short fractures
(i.e., large values of CL~). As we would expect, fracture face
skin affects the pressure response most when formation linear
flow behavior predominates (i.e., large values of FcD and small
values of C@).

f==
Note that we must have estimates of @and ct. The
wellbore storage coefficient (C, bbl/psi) for a changin2
liquid level in the wellboreis defined as

C= A&b
Pf

where A~b is the wellbore area in bbl/ft and pf is the

density of the fluid (psi/ft) used for the slug test.
6. Using the time match point obtained in Step 4, calculate
the formationpermeabilityas follows:

=33YC
(Q%@),.p

SLUG TEST ANALYSIS PROCEDURE

In this section, we present a procedure for analyzing slug iests
in hydraulically fractured wells, and we illustrate the procedure
with a simulatedinjection slug test.

If pre-fracture estimates of pemwability are available, we

may estimate the 1orosity-compressibility product using
the time match point and the correlating parameter CL~
from the type curve match, or

Analysis Procedure

1. Prepare semilog and log-log plots of the field data. We

suggest plotting both pDSIUg and 9DAig vs. time using the
same scales as the type curves. For p\o@ction slug tests,
PDShIg $ defined U Eq. 8! while for.Injection type tqsts,
PDsJ#gMdefined by Eq. ~. ln addltlon, the e~ly-time
plotting variable, qDs/~ , is defined by Eq, 13, As we
have discussed previousf y, semilog and log-log plots of
PDSIUg wc best. suited f:r analyzing intermediate-and latetime data, while semdog and log-log plots of q&lUg
provide resolution of early-time data, and under some
conditions, help to reduce the ambiguity of type curve
matching.

7. From the dimensionless fracture conductivity obtained in

Step 4, wc may estimate fractureconductivityas
k~f = (kLjJFcD
Exar@e

Problem

We illustrate the application of our type curves with a

simulatedinjectionslug test in a hydraulicallyfracturedcoal seam.
The test data were generated with the coal properties summarized
in Table 1, We have assumed the slug test was conducted in a
2. Next, we select the type curves for analysis of the slug
well completed with 7-inch casing (I.D. = 6,094 in.) and using
test. As we discussed previously, the slug test re~ponse
fresh water, The wellbore pressure at the beginning of the test
depends on three paramc{ers -- CL , FcD, and S/, , was PO = 600 psia. In addition, we note that previous fracture
Therefore, we should use all availabTe information to treatments have not suffered significant fluid losses, so we
estimate these parametersand reduce the uncertaintyin our ~ assume fracture-faceskin is negligible.
analysis, For example, if previous fracture treatments in
the particular coal seam have not resulted in significant
TABLE 1
fluid losses into the formation, we may select type curves
for no fracture-face skin.
I
Well and Coal Seam Parameters
3. Next, we find a match bctwccn the field data and type
curve plots. Because of the marmcr in which PDSIU8 and I
qmlu~ have been defined, wc simply align equal values of 1
the functions on the vertical axes and slide the graphs I
horizontrdlyuntil we obtain a match.
I
!

4, From the type curves, rcitd values of the correlating I

parameters, CLp, FCD, and $ In addition, obtain a time
match point, i.e., @/CL~ and the corresponding value I
of r from the plot of [hefie!ddata.
I

5. Using the definition of dimtmsionlcss wellbore storage !

coefficient, estimate the hydraulic fracture half-length.
From Eq, 7,
1
I

B, RB/STB
cl, psia-1
tl,
ft
$
Cp
J1,

;:, P#ft
psia
Lf, ft
k, md

Pi,

FCD

3X 1)-!
10
0.025
1.0
0.4333
0,25
500

lCNJ
5
x

The fust step is to calculate the slug test plotting functions,

] PDSIUg~d qDskg, and prepare scmilog and log-log plots of both.
~ Next, we attempt to match the log-log plot of the field data with
the slug test type curves for zero skin. Note that we may obtain
rcasonttblematches of pD$/u for several values of FCLI and CL,;
however, the best match of ~oth plo[ting functions was obtained
---

J.A. Rushing, T..4. Blasingame, B.D. Poe, Jr., R,M, Brimhall, and W,J. Lee

SFE21492

correlating parameter that could be used for all flow regimes.

Perhaps an alternative analysis technique would be to develop an
automatictype curve rndtchingtechniqueusing our slugtest model
and a multi-pammeterestimationscheme.

with type curves for FCD = m CL~ = 0.01, and Sf = O(Fig. 22).
The time match point is
t= 1 hr and t@L~

= 0.19

We also used our new slug test model to invest!yatethe effects

of various fracture and reservoir parameters on the slug test
response in hydraulically fractured coal seams. In general, we
observed bilinear flow patterns for all values of CLP and FCD,
but especially at the smaller values of CL~. However, this flow
pattern ended at L@CL~ <0.01 or much earlier for larger values
of FCD. In addition, we suspect that formation linear flow is
exhibited at late times for small values of CLfD in infiniteconductivity fractures, Further, we saw a unit-slope line at very
late times, which suggests a stabilized pseudoradial flow period.
As we would expect, this pseudoradial flow penod appears
sooner in low-conductivityfractures. Finally, we investigatedthe
effects of fracture face skin and found that these effects are less
pronounced at large values of CLfi and.FCD,which suggests that
we cannot estimatethe fracture face skin accuratelyfrom slug tests
in short, low-conductivityfractures.

Using this same match point, we also obtain a good match of the
semilog plot of the data, as shown by Fig, 23.
We estimate the wellborearea (AWb,bbl/ft) to be

&+&

Z(0.25)2
=
0.035 bbl/ft
5.615

and the wellbore storagecwfficient is

c.~.~.().(381

bbl/psi

0.433 psiflt

The hydraulicfracture half-lengthis estimatedto be

==
(o.sgd)(o.og

f=

As wc have discussed, our slug test model was developed

using Cinco-Ley, et afsg finite-conductivity, vertical fracture
model and is limited to the assumptions listed earlier in the paper,
In addition, our model assumes a constant wellbore storage
coefficient during the test. Further, our model is valid only for
single-phase flow conditions, Finally, we should note that we
assume the fracture fully penetrates a single zone and that this
single zone is the only zone in pressure communication with the
wellbore,

1)

(0.025)(3 X 103)(10)(0.01)
f=

98.3 ft

which agrees with the value used to simulate the slug test, The
permeabilityof the coal seam is estimatedto be
=339;C

NOMENCLATURE

(*~p

= area of wellbore, bbl/ft

= formation volume factor, RB/STB
= fracture width, ft
= wellbore storage coefficient based on a changing
:
liquid level, bbl/psi
CLP
= dimensionless wellhre storage coefficient based
on fracture half-length
= fluid compressibility,psi-l
c
= total system compressibility
y, psi-l
20
= dimensionlessfractureconductivity
/1
= formation thickness,ft
K.
= modified Bessel function of the second kind, zero
order
= formationpem]eability,md
= fracture permeability,md
~
= hydraulic fracture half-hmgth,ft
= measured pressure during slug test, psia
P
= dimensionless pressure
PLI
~Dslug = slug test pressure plottingfunction
PDSILW = $:;fi:
transform of slug test pressure plotting
,.
= pressure in frticturc,psia
Pj
= dimcnsiorslcssfmcturcpressure
Pp
Pp
= Lapktcc transform of dimensionless fracture
pressure
= initial pressure, psia
Pi
= measured pressure at beginningof slug test, psia
Po
= dirncnsionlcsspressureat the wclltx)re
~wD
PWD
= Laplacc transform of dimensionless wellbore
pressure
= bottomhole flowing pressure, psia
:Wj
= flow rate, bbl/day
= dimensionlessflow rate
4D
= dimensionlessfracture flow rate
6p
WMug = slug test rate plottingfunction
A ~~

~ = (3,390)(1.0)(0.081) ~
()1.0
10
k = 5,2 md
which also agrees with the value given in Table 1. The fracture
conductivityis estimatedto be
kjbf = (kLj)FcD
k~j=

(5.2)(98.3)(z)

kybf = 1,606

md-ft

SUMMARY AND CONCLUSIONS

We have developed new type curves for analyzingslug tests in
hydraulicallyfractured reservoirs rmdhave presented a procedure,
which is illustrated with an example, for applying these new type
curves, With these type curves, we may estimate the hydraulic
fracture half-length, formation permeability, and fracture
conductivity, We have also investigated the effects of fracturcface skin, and have concluded that these skin effects arc difficult
to quantify in short, low-conductivityfmcturcs.
The new type curwesare based on Cinco-Lcy, et u1s9finitcconductivity,vertical fracture model iurdarc presentedin terms of
three parameters -- CLp, FCD, and Sj, Like Ramey, et al,6 who
developed their slug-test type curves using the correlating
parameter CDe2S,we attempted to find a similar parameter which
would reduce the number of curves. However, because of the
various flow regimes that occur during pmluction from finiteconductivity fractures (i,e , fracture linear, bilinear, formation
linear, und pseudoradialflow), we were unable to develop a single
,

SPE 21492

Analysisof SlugTest Data From HydraulicallyFracturuJ Coalbed MethaneWells

Sf

s
tLp
x
XD

Y
,$G
Q=

14. Cinco-Ley, Ii, and Samaniego-V,,F.: Transient Pressure

Analysis of Fractured Wells, J, Pet, Tech. (September
1981) 1749-1766,

fracture face skin factor

Laplace transformvariable
dimensionlesstime basedon fracturehalf-length
distance along fracture,ft
dimensionlessdistancealong the fracture
= distance perpendicularto fracture
= dimensionlessdistanceperpendicularto fracture
porosity, fraction
viscosity, cp.
density of liquid in wellbore,psi/ft
=

=
=
=
=

15, Cinco-Ley, H., Ramey, H,J., Jr,, Samaniego-V,, F., and

Rcxirigucz,F.: Behavior of Wells with Low-Conductivity
Vertical Fractures, paper SPE 16776presented at the 1987
SPE Annual Meeting, Dallas, TX,, September 27-30.
16, Sageev, A.: Slug Test Analysis, Water Resources
Research J. (August 1986) 22, No. 81323-1333,
17. Cinco-Ley, H, and Samaniego-V., F.: Transient Pressure
Analysis: Finite Conductivity Fracture Case Versus
Damaged Frdcturc Case, paper SPE 10179presentedat the
1981SPE Annual Meeting, San Antonio, TX., October 5-7.

REFERENCES
1.

Ferns, J.G. and ?(nowles, D.B.: The Slug Test for

Estimating Transmissibility, U.S. Geological Survey
Ground Water Note 26 (1954), 1-7.

2,

Cooper, H.H,, Jr., Bredehoeft, J.D., and Papadopulos,

1.s.: Response of a Finite-Diameter Well to an
InstantaneousCharge of Water, Wa/er Resources Research
J. (1967) 3, No. 1, 263-269.

3.

Appendix

Papadopulos, S.S,, Bredehoeft, J.D., and Cooper, H.H.,

Jr,: On the Analysis of Slug Test Data, Wafer Resources

Kohlhaas, C.A.: A Method for Analyzing Pressures

Measured During Drillstem-Test Flow Periods, J.Pet.
Tech. (October, 1972) 793%00.

6.

Ramey, H.J., Jr., Agarwal, R,G., and Marlin, 1.:

Analysisof SlugTestor DST Flow Period Data, J. Cdn.
Pel. Tech. (July-September, 1975) 37-47.

7.

Koenig, R,A. and Schraufnagel, R.A.: Application of the

Slug Test in Coalbed Methane Testing, paper 8743
presented at the 1987 Coalbed Methane Symposium,
Tuscaloosa, AL., Nov. 16-19.

FCD

9.

PfD(xrMp)

ayD

k h@i - p/(X,t))
. . . . . . . . . . . . ,, .,, ,,, ..,. . . . . . . . (A-2)
141.2 qlip

tf#l = ~
@/JCtL~

, ..,,.,..,,,..

.... ... ...... .. ..

.$.....,,.,

(A-3)

,.

. . . . . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (A-4)

. ~

xD=A-

(A-5)

Lr`'`"` "`"`' ````""`'"`'"-`"'""""'"'`'''`"'""'''"'"'''`'''""-

YD =/f . . . . . . . . . ...,.,.,,

Karsaki, K,, Long, J.C.S., and Withcrspoon, P.A.:

Analytical Models of Slug Tests, Wu{er Resources
Research J, (1988) 24, No, 1, 115-126.

.,, .,,.,,.,,

, ...,.,,,,,,

.,, ,. .,.,,,.

. . . . (A-6)

Ihc frtcture is at initialreservoir pressurePi, or

Cinco-Ley, H., Samaniego-V., F., and Domingue~, N.:

Transient Pressure Behavior for a Well with a Finite.
Conductivity Vertical Fracture, Sot. Pet. i%?. J, (August
1978) 253-264.

PfD(xD,tL~=o)

=0, fOros~s ~D, ..... ...... ...... .......(A-7)

while the boundwy condtiondescribingconstant rate productionat

the wellbore (i.e., XD= O)is

10. Cinco-Ley, H, and Meng, HZ:

Pressure Transient
Arudysisof Wells with FiniteConductivityVerticalFractures
in Double Porosity Reservoirs,paper SPE 18172 presented
at the 1988 SPE Annual Mectirrg,Houston, TX., Ott, 5-8.
11,

Model

where the dimensionlessvariablesare defined as foilows:

F~D

8,

Finite Conductivity

b+-La&=
o.,..,,,.,..,..,,,,,.,,
dxD2

van Poollen, H.K. and Webber, J,D.: Data Analysis for

High Influx Wells, paper SPE 3017 presented at the 1970
SPE Annual Meeting, Houston, TX., Ott. 4-5.

5.

A . I)crivation of
Vertical Fracture

Following the model formulation and problem solution presented

by Cinco-Ley, ef fJf,9, the partial differential equation in
dimensionless variables describing the transient flow in the
fracture is

Research J, (1973)9, No. 4, 1087-1089.

4,

z?12=.
Jcax.

~cD..................................................(A.8)

The boundary condition describing the no flux condition at the

fracture tips (i.e., XD= *I) is

Meehan, D.N., Home, R.N., and Ramcy, H.J., Jr.:

Intcrtcrcncc Testing of Finite Conductivity Hydraulically
Fractured Wells, paper SPE 19874 prcscrrted at the 1989
Annual Meeting, San Antonio, TX., Oct. 8-11.

%-D
.()

i)xD

Numerical [nvcrstion of Laplacc

12, Stchfest, H.:
Transforms, Communicationsof the ACM (January 1970),
13, No. 1, 47-49.

,,, ,,, ,. .,,,,, .,, ,.,,,,,,,,, ,,, ,,, ,,

,0!,.,

,,,

.,,!,,.,,.

(A-9)

After integrating Eq, A-1 twice with respect to XD and applying

the initial and boundary conditions, CincoLcy, et uI,9dcnved the
pressure drop between the wellbore (i-e,, XD = 0) and any point
the fracture as

13, Ramey, H,J., Jr. and Agarwal, R.G.: Annulus Unloading

Rates as Influenced by Wellbore Storage and Skin Effect,
Sot. Pet. Ew?. J. (October 1972), 453-462.
llU

J,A, Rushing, T.A. Blasingame, B,D. Poe, Jr,, R.M. Brimhall, and W.J, Lee

SPE 21492

The Laplace transformof Eq. A-15 is

Pp(XD=OJL@ - Pp(xDJLp)
Xf)

1
fip(s) = * 1 ?fD(x,S) Ko(!-rD- X~) dx

-A

=~

FCD o

qfD(%,tLfD) & & .,,,,..,.,..,,,,...

(A-1O)

XD

-n
FcD

where the dimensionless pressure at the wellbore for constant rate

conditionsis

H
~

GfD(%,S) h

&+ ~

.. .... ...

(A-17)

and, the Laplace transform of Eq. A-16 is

k hoi
PjD(XD=OJL~)

- P~~X=O,t))

,,,,.,,.,,..,,,.,,...ll)

141.2@~

11)
FjD(s) = 1
~ -1*~P(x,s) K*D

2q~x,t)Lf

q/D(%,t) =

. . . . . . . . . . . . . . ,,, , ..,....,,.

. . . . . . . . . . . . . (A-

- xb)

and qfD is defined as the dimensionless flow rate per unit of

fracture length going from the reservoir to the fracture, or
12)

Modeling the fracture as a plane source with dimensionless flux

the dimensionless pressure drop at any poim
density qD(x,r@,
in the reservoir IS

*~D+

J
SFCD

I
.1

sfifD(x'!$)~~' ....... ... ... .. .. ....l ....l. d.(A-I8)

/@(xD,yD,f@J
=
Appendix B - Derivation of the Semianalytical
Solution
for Finite Conductivity Vertical Fracture

or, solving for the pressure drop at the face of the fracture (i.e.,
)D = o),
PD(XD,YD=OJLLD)=

In this appendix, we summarize the solution technique for Eq. A17 as presented by Cinco.Ley and Meng10and Meehan, et al. 11.
A similar technique may be used to solve Eq. A-18. rhe
technique is a semianalytical approach in which the fracture is
divided into n discrete elements of equal length Ax, and each
element is treated as a uniform flux fr~cture. If we assume the
fracture
is symmetric
and homogeneous,
then
~fD(xD!s) = 4-D(-XD$S) and % A-17 may ~ rewritten as

fwj,D,%l,t)ep[{=]l~~,~

L
40

.*

fi~(s) =

.:............(A-14)

tL~ -T

If we equate Eqs. A-10 and A-14, we may solve for the pressure
in the fracture at the wellbore,
P@(XD=O,tL@)=
1
-1qp(x,~)

lL/)

H
H

1
40

XL)

-@)-%
)2

[L@]

xl)(tf,~~
4(f- ?)-T) dxd~

~fD(x,S) d%& + ~
sFCD, . . . . . .. ..l)... (~-l)

Me first integral on the right side of Eq. B-1 maybe approximated

as

.-L

FcD

XD

.-zFCD Q

q@xJ@dx~.~

+ ;&,..,,,,,,.

(A-15)

?p(x,~)

[~o(hD

- xlfi)

+Ko(hD + xI@]dx

If we include the effects of fracture face skin (S~), Eq. A-15

becomes
pp(xD=O,lL@) =
l/#

J4 ...

qp(x,r)
XD

where ~Diand x~i+l arc the beginning and end, respectively, of

the it/~ element, We may also approximate the second integral on
the right side of F,q,B-1 as

(xL)x)*
4(tL@) Id

~,dz

(tLfl - ?)

.-lr-

FCD

~xp -

q/z)(xJ@dxdx

+!!32+
FC/j

J
.,

Sf qf/j(X,tL/D)

, .,,.,.,,..,.,,.

0.,,...0,0.

(A-16)

where XD is the midpoint of the itil clement of the discrctized

fracture, hubstituting Eqs, B-2 and B-3 into B-1, wc have
111

8.

Analysis of SIugTest Data From HydraulicallyFracturedCoalbedMethaneWells

~*-~i(S)
+ (X~j-i~)ijlli(s)+(A+

AX

FCD

~~j(s)

[ .=l

+.E2L

SFCD ""'"' "'''" '""""`'`""''"''""""""`"'`"""'""'`"`'`"-"''""""'

(B-4)

In addition, we utilize the conditionthat the summationof the flow

entering each fracture element is equal to the flow at the wellbore,
or, in Laplace space,
I

h~

,=1

~~i($) = ~ .. . . .. . .. . .. . ..

.!...,....,,.

. .

. .

. .

. .

(B-5)

We may write an equation for each fracture eIement and with Eq.
B-5 develop a system of n+l equations with n+l unknowns, i,e.,
Zjpi(s), i = 1,$,.,n and ~p(s). The unknowns are obtained by
solving the system of equations. Once we have the dimensionless
flux and pressure in Laplace space, we may then use the
Stehfest12 algortihm to obtain these variables in real space. A
similar methodology is used to solve Eq, A-18 when fracture Face
skin is included. Additional details of the solution procedure are
provided in Refs, 10 and 11.

11?

SPE 21492

1 o

Fc~=O.in

WOllbOfO

08-

==B
@

Fracture
Cj,,[,
(a)

Fracture

Linear FIOW

06-

L Fraclu10
(b)

104 lo> 102

u? 52

Bilinear

FIOW

1(11 1

10

04
Fmclute
\i/
Iilli

02

--2=,-.=

11!1!
+

ff~

Wellbore

00
10

(c) FormalIon
Fig. I

Linear

(d)

Flow

Pseudoradnal

10

Flow

I:lmv periods for Ilyckwlicdlly lrdtilurkd \Vcll$

FCD. n

qu

..

..1

-.1---1-.-1-

..L

\\\\

!0

,
..-.mL\
)
10

1
10

lLI

lb

10

-~~1
o

!0

10

10

10

! o

10

Fc[j= 10011

08

I(I4 10~ 10~ 1[1!I II

J.~

06
J
-$
j
04

02

00

~w
!0

10

!0

10

10

10

MA@

l:Ig. S - Slug Ics[ Pressure Iksponw

for ICIJ = l(Xht ml Sf = O

hg, 6. [ntcnncdi~lc.

UK! IAIIC.IIIIIC Slug lml Iksponsc for FCD = O.in imd Sr = ()

10

Fig, 7. hncnncdlak. ;md I.wc.lime

Slug ICSI RCSIJOIISCfor I;CTI = R iwd Sf = ()

IIU. II

I,uct!ncdiak.

iwd l.m.lIII1c

Slug

ICSI RCSIMJIIW fur I;C.W~. lox WI

SI A O

,.

, no-

10:

104

I(IJ

I(P

101

10

la

10

!--!..l~

0001

0 of

01

~ =

10

\oo

oL--

-J

!0

1000

10

!0

Fig.9. h]lcnllcdiatc.

and Ltm.lmx

Slug Tc!t RrspoIIsc for I;CIJ u ILYJT ml

!0

10

0,1K141>

10

! 0

lhticl.ll)
Fig. 10. F.wly.lime

SI = 0

Slug Ivsl IWqxmw

for I:CI).

L) In ml

S[ = 0

10?

1{)4

.. -. -.L_.A_

II)J

I(IZ
J----

Ill I I Ill
!-. . J-..1..

.$?
10

10

!0

-7 WTlm?---rnn

!0

Ii

= Ion
PL1)

Sj = o
. m7q--rlnTlTr(

nll~

10

h,ll(

Ilg

118

12. lhrly.1

to

!0

mw Slu~ ll\l

10

1,1)

I&hp,,nw

fur

l;CIj K I(M and SI (1

10

$0

1<

F(.D = (), I;{

!0

$?O

0.01
1)!1
.4. -- -.1.. -. -J -,
0

!0

10

l(JOn
Sf=o

FCI>=

10

.
!0

10

Wq[)

Fig. 13. I:wly.1m)c

Slug rut

RcspoIIw for I;CII = I(M

Sf = o

10

Fc-D=odllr
oa

10

on-

-.

.A-...

04-

to

02-

00

-y.-r-vf.mny-fl-fnq
0001

001

-nlnrr-l.
0!

T~
1

!00

1000

tL@+)
lIIg,

16.

f2atly.fiuic Slug ICH lfcspr)IIsc for I;CXI= (J.lrrand CI+D ~ 104

0.

1(

PCD=O,lK
CI,D= 1
00.

10

08.

gfl .10

Jj

(j

._-----_l.

10

[)

0.1

(11)1

04.

10

10

.sf

$
. ......1.. .-

1..

02,

I
10

00, , .
10

10

w -T~T6-V*
10
10

ko%)

10
0

10

10

,M 1,11

FIF,,IH Slut! Icsl I{C$IKVISC

for I:UJ r ltXlrTiwd Li ID T104

li& 17. I,mly.1imc SIUSTCSIRcspomcfor Fro = O,lx WI CL,I> = I

!o -

F(l) = loon
08-

%
00.

(l 01

A-.L.

0 I

!!
q!+

$0,.
02

00, w
!0

o@.l,l)

r-mq~-rn
!0

111
,Y_?V~717nV--*Vmq-1T~
! 0
10
!O1

! o

.,

sf5zi492

I, hours

001

0,01

$3

0.1

01

10

10UU

100

102

10:

10
J

10

1
10e

10

IG

10%

!0

10

1 o

10

t@@

Fig. 21- Early-Tin

i
!0
U

Slug Tes~ Response for F~

= 1~

h& 22.

and CL@ = 1

kq+]og

hours
001

0001

0.1

10

lCO

1000

10

08

06

04

02

00
!0

!0

to*

I 0

10

10

kid%

Fig. 23- Scmilog MaIcll of the Example Slug Test

11:

! o>

10

1
10

,d%l

hfatch
of the EX3Mpk Slug IC$I

m
1