An Investigation of Wellbore Storage and Skin Effect

.
In
Unsteady Liquid Flow: I. Analytical Treatment

RAM G. AGARWAL PAN AMERICAN PETROLEUM CORP.

TULSA, OKLA.
RAFI AL-HUSSAINY MOBIL RESEARCH 8. DEVELOPMENT CORP.
JUNIOR MEMBERS AIME DALLAS, TEX.
H. J. RAMEY, JR. STANFORDU.
MEMBER AIME STANFORD, CALIF.

ABSTRACT INTRODUCTION
Due to the cost of extended pressure-drawdown Many practical factors favor shore-duration well
or buildup well tests and the possibility of acquisition testing. These include loss of revenue during shuc-in,
of additional information from well tests, the modem costs involved in measuring drawdown or buildup
trend bas been toward development of well-test daca lor extended periods, and limited availability
analysis methods pertinent for short-time data. of boccomhole-pressure bombs where ic is necessary
"Short-time" data may be defined as pressure co survey large numbers of wells. On che ocher hand,
information obtained prior to the usual straight-line reservoir engineers are well aware of the desirability
portion of a well test. For some time there bas been of running long-duration tests. The result is usually
a general belief that the factors affecting short-time a compromise, and not necessarily a satisfactory
data are too comp! ex for meaningful interpretations. one. This situation is a common dilemma for che
Among these factors are wellbore storage, various field engineer who muse specify che details of special
skin elfects such as perforations, partial penetration, well tests and annual surveys, and interpret the
fractures of various types, the effect of a finite results. For chis reason, much effort has been given
formation thickness, and non-Darcy flow. A number co the analysis of short-time tests. The tenn
of recent publications have dealt with short-time "shorc-cime" is used herein to indicate either
well-test analysis. The purpose of this paper is to drawdown or buildup cescs run for a period of time
present a fundamental study of the importance of insufficient to reach the usual straight-line portion.
wellbore storage with a skin elfect to short-time Nevertheless, ic is rare that either buildup or
transient flow. Results indicate that proper interpre• drawdown data taken before the traditional scraighc-
tations of short-time well-test data can be made line portion are ever used in analysis of oil or gas
under favorable circumstances. well performance. Well files often contain well-cesc
Upon starting a test, well pressures appear data chat were abandoned when ic was realized that
controlled by wellbore storage entirely, and data che straight line had not been reached. This situation
cannot be interpreted to yield formation flow is particularly odd when ic is realized chac early
capacity or skin effect. Data can be interpreted to data are used commonly in ocher technologies which
yield the wellbore storage constant, however. After employ similar, or analogous, transient tests.
an initial period, a transition from wellbore storage It is che objective of this study to investigate
control to the usual straight line takes place. Data techniques which may be used to interpret infonnacion
obtained during this period can be interpreted to obtained from well tests at times prior to the nonnal
obtain formation flow capacity and skin effect in straight-line period.
certain cases. One important result is that the
steady-state skin effect concept is invalid at very THEORY
short times. Another important result is that the time
The problem to be considered is the classic one
required to reach the usual straight line is normally
of flow of a slightly compressible (small pressure
not affected significantly by a finite skin effect.
gradients) fluid in an ideal radial flow system. That
is, flow is perfectly radial co a well of radius 'w
Oriclnal. manuscript received In Society of Petroleum Ena;lneers in an isotropic medium, and gravitational forces are
office Jan. 18, 1969. Revised manuscript received July 15, 1970.
Paper (SPE 2466) was presented at SPE 44th Annual Fall neglected. We will consider that the medium is
Meetlna;, held In Denver, Colo:, Sept. 28-Oct. l. 1969. re',CopyrfcJ,t infinite in extent, since interest is focused on times
1970 American Institute of Mlnlnc, Metallurclcal, anaPetroleum
Enclneers, Inc. short enough for outer boundary effects not to be
lRelerences slven at end of paper. felc at che well. The initial condition is taken as
This paper will be printed In Tr.,nsactlona volume 249, which
will cover 1970.
constant pressure, P;, for radii greater than or equal

SEPTEMBER. 1970 2T9

 to rw-
The inner boundary condition will be taken as
production at constant surface rate from a wellbore
of finite volume, and it will be assumed that a
steady-state (zero storage capacity) skin effect
exists at the sand face. This boundary condition
c dpwD
dtn -( :~) rD=l
= 1 . (4)

without a skin effect was first introduced by van and

Everdingen and Hurstl and is sometimes called the
wellbore storage, unloading, or afterflow problem.
Later, van Everdingen2 and Hurst3 extended the
problem to include a steady-state skin effect. Both
van Everdiogen and Hurst presented solutions to the
problem for the special case of a Hoe-source well.
Solution, including a skin effect, was presented in
Pwn =
[ %- ·C~~)] r D=l
' . (5)

the form of the real inversion integral of a Laplace Eq. 4 states that the dimensionless wellbore
transform solution, but only the long-time approxi- unloading rate plus the dimensionless sand face
mation has been published. flow rate must equal unity, the surface flow rate.
The inner boundary condition described is actually The wellbore unloading or storage constant, C, 1 s
a special case of Jaeger's "general boundary that defined by van Everdingen and Hurst.1 That is,
condition" applied to unsteady beat-conduction
problems. 4-6 In the "general boundary condition," c =
C
2:1t¢hcr 2 (6)
Jaeger considered the cylindrical core (region w
between r = 0 and r = r111 ) to contain a solid of perfect
conductor or well-stirred fluid in which (1) heat C represents the volume of wellbore fluid unloaded
co·utd be generated at a constant rate per unit volume, or stored, cc/atm. Storage may be by virtue of either
(2) heat could be transferred to the surrounding compressibility or a changing liquid level.
cylindrical solid through a film resistance, and (3) Eq. 5 introduces a steady-state skin effect and,
a mass of the well-stirred fluid could be withdrawn thus, a pressure drop at the sand face which is
at constant rate. The first edition of the book on proportional to the sand-face flow rate. Note from
heat conduction by Carslaw and J aeger6 contained Eq. 4 that
a rigorous solution to a heat-conduction problem. It
was analogous to the fluid-flow problem originally qsf'
posed by van Everdingen and Hurst 1 for wellbore
unloading without a skin effect for a finite radius
q = ( :~)
well. It is interesting that the real inversion integral rn=l
published by Carlslaw and Jaeger has been evaluated
in several publications in connection with problems - dpwD
other than the van Everdingen-Hurst problem. 7, 8 = 1 - C dtn I • • • • • (7)
Jaeger also evaluated the integral and presented
useful long-time approximations. 6,9 In the second
where q is the constant surface flow rate and q I
edition of their book on heat conduction, Carslaw • th • s
1s e sand-face flow rate. Finally, the dimensionless
and Jaeger6 presented a review of the problem that
is peninent to the present study. We will present the p
flowing pressure, PwD• is the same as 8 (t0 ) used
previously by van Everdingen2 and Ramey.11 Thus,
fluid-flow analog briefly. The procedure is similar
to the heat flow problem originally presented by
PwD represents the pressure within the wellbore
Blackwell. lo while PD represents pressures on the formation sid;
of the skin effect.
The diffusivity equation for fluid flow in terms
of dimensionless variables is Solution of Eqs. 1 through 5 follows readily using
the Laplace transformation as shown by Blackwell. lo
We will review only portions of the solution of
(1) interest to this study. The transform of the
+ dimensionless flowing pressure may be written as

The initial and outer boundary conditions at:e

=
(2)
Pn (rD, o) = o

lim
(3) p[ypK1(yp) + Cp{K0 (yp) + s {pK
1
({1,)}]
rn--
(B)
while the inner boundary condition is
 where Ko and K1 are the modified Bessel functions although there has been some overlap in evaluations,
of the second kind of zero and unit orders. An there has been little duplication because various
identical transform was published recently by authors have been interested in different ranges of
Russell. 15 storage constant, C, values. Note also that the
Jaeger also considered determination of the special case of zero storage leads to the well-known
pressure distribution within the radial system, the constant rate solution. (Jaeger a and Hurst a values
sand face flux, and the pressure drop across the of infinity), Jaeger9 has presented the most complete:
skin (within the heat conduction analog, of course). information for finite skin effect values, but the
However, these quantities are only of incidental range of both dimensionless times and storage
interest to the present purpose of this study, constants is far too limited to be of use in well-test
Blackwell and Jaeger presented the heat conduction analysis.
analog to the following real inversion integral Another related solution bas appeared in the
solution to Eq. 8: petroleum litcracure. 1· 3 By using the superposition
principle and representing the well as a continuous
Hoc source, the Laplace transform of the dimension-
less well pressure may be shown to be:

f[
00

_4_
1(2 (1
- e -u tD)du 4
2 L
{Pwn} =

0
2 Ko (.JP)+ s
• (10)
u3{ [ ufu0 (u)-(l-Csu )J1 (u)]
2
p [ 1 + C p K0 (FP)+ s c P]
}]. (9) This same result may be obtained from Eq. 8 by
noting that the product [-.ff K1 (,[ti)] approaches
unity as the argument p becomes smaller. Thus Eq.
Mtere Jo(u) and Ji(u) are the Bessel functions of
10 is a long-time approximation for Eq. 8.
the first kind of zero and unit orders, and Yc/.u) and Eq. IO produces the real inversion integral:
Y1(u) are the Bessel functions of the second kind
of the respective orders. Both numerical evaluations
and short- and long-time approximations for the
integral in Eq. 9 have been presented by a number
of authors in connection with ocher problems.
However, the relationships between them seem to
2
have been largely overlooked. Table 1 presents a
comparison of the symbols used by various authors
-n t D) J (u}du ¾
0
who have presented pertinent evaluations of the
the integrals or related derivations. Table 2 presents
the ranges of parameters in numerical evaluations
considered by various investigators.
It is apparent from Tables 1 and 2 that most
evaluations of the integral in Eq. 9 have concerned
the special case of a zero skin effect - no-flow
resistance at the sand face. It is also apparent that,

TABLE 1 - CORRELATION OF VARIOUS DIMENSIONLESS TABLE 2- RANGES OF PARAMETERS CONSIDERED BY

GROUPS USED BY OTHER INVESTIGATORS VARIOUS INVESTIGATORS IN NUMERICAL EVALUATIONJ
Lesem • C _to
__ Comments
6 8
This Study Blockwell
10
Jaeger •9 et a/.7 Hurst
Blockw•IIIO Analytical
stvdy only•
41c2
PwDCs,C,tD) e1c,, 27rG(fi,a,r) -11
,r2
uN(u,,0 ) Jaeg,...6.,9 D V&.1/6, 1/4,
V2, 213, I, 2
D.2-2.0 Tal,ular
Iona•
I O, 1/2, I, 2, l, 1/2, I 0,2-:20 G.ophical
c -a
2
Va Vic Vu
Leaem et ol. 7
4,5,7, 10,20
D 1/6, v,, 1/12,
fom1•
I - 1,000 Guphlcal

ID
T, t r e 'o
V15, 1/11, 1/21,
1/24, 1/27, 1/30 "'""·
• 1/H h s=O s=D Hunt 1 D 2.5, S, 10, \2.5,
1fr2/3, 25, SO
,.,.,,
10-1_,04 C.cpl.lcol

von EYetding•n 0 1,000-SO,OOO 10 3 - 107 Lln .. wwc•

Note: For the fluid flow analog of Blackwell'• study, f,~ h, an<I H... atl ... 11 oppf'o••
K and Q' were all taken as unlry (In his sym!.allsm). Both 111Gtion.
2
van Everding en 2 ond Hurst 3 used sym!.olism similar to this study grophicol••
In their line-source opproximatfon, except fi was used In place • Both aho,t.. on4 l0ft9""tim• opp,ox1mat1nt fonna given.
of PwD• ••Long•trm• c,ppro•imotill"f1 form gl,ren.

SEPTEMBER, 1910 281

 Eq. 11 was evaluated numerically and presented Short-time approximations are
in graphical form for the special case of zero skin
effect by van Everdingen and Hurst 1 and later by t 2
D
Chatas.12 Storage constant, C, range was 1,000 to Pwn( s,c,~)=-=-
- 1
{ 11> -
50,000 in Ref. 1 and 1,000 to 7.5.000 in Ref. 12. C 2cs
Originally, both Blackwell and Jaeger solved Eqs.
1 through .5 in the classical manner. That is, the 8 t 5/2
Laplace transforms of the equations were taken and
the resulting equations solved simultaneously to
+
D
15 Ji ~2
Cs
+ 0( tD3) } ,. s,C ro
provide the transform space solution given by Eq. 8.
Although Jaeger mentioned that classical super- • • • • (14)
position could have been used, be did not employ
this essentially simpler technique; van Everdingen
and Hurst did employ superposition in their - 1 { 4 ~3/2
development of Eq. 10. Thus, one solution leading Pwn(o,c,tn)= c "tn- cf;c
to Eq. 8 can be developed employing superposition:
3
+ o( tn2 > } ; s = o, c ,. o ... (15)
where y is Euler's constant 0.57722. The well known
constant rate solution (no storage or skin effect)
was shown by van Everdingen and Hurst 1 to approach
the following form at long times:

1
2
dt D' + s [ J. - -C _ _ tD)~
dpwD( _ • • (12)
dtD
Throughout Eq. 12, the dimensionless time within
y ], t D
> J.00 • • C16)
brackets simply indicates that the preceding term is
a function of time. The prime mark indicates a Comparison of Eqs. 13 and 16 indicates that at very
long times
variable of integration. Eq. 12 is an
integro-differential equation of the convolution type.
The Laplace transform can be taken directly. If the
constant-rate line source is used to obtain the Laplace
transform needed for the second term in brackets in as was concluded by van Everdingen 2 and Hurst.3
the integral, Eq. 10 results. If the constant-rate, This can be seen a little easier if Eq. 13 is
finite radius cylindertransform(Ref. l)is substituted, rearranged and the substitution for PD(to) made from
Eq. 8 results. This approach is enlightening because Eq. 16:
the true nature of the wellbore storage problem is
indicated clearly by Eq. 12. Reference to Eq. 7 PwD (s,C,~) =
shows that the first term within brackets in the
integral in Eq. 12 is the sand-face production rate
at any given time. Thus, Eq. 12 expresses the
wellbore pressure caused by a changing production
rate which results from the wellbore storage condition.
More will be said about this point later.
2~
1
- t! l
(2s - 1) + 1 o( -2)
Both Jaeger9 and Blackwell lo presented short- and 2
long-time approximate inversions for Eq. 8. Chang-
4~ tD '
ing their solutions into the fluid-flow nomenclature, • • • • • • • • • • • • • • (18)
the long-time approximation is:
In this form, several interesting features of the
long-time solution can be recognized. First, the
term [1/(2t0 )] within brackets will always be
negligible for to greater than 100. Terms of the
- 'Y + 2s + 2 ![D
1n 4tD - Y+ 1
order of 10 ·2 will be negligible. Rameyll pointed
out that the effect of wellbore storage essentially
dies for zero skin effect cases by a dimensionless
time of
2
-2C ( 1n 4tD - -Y + 2s) ] + o( 11>- ) }
tD > 60 C . . . . . . . . . (19)
• • • • • • • • • • ( 13)
 le is clear that this approximation will also hold for some t!wd storage capacity (pore space) or they
finice skins (either positive or negative in sign) for could not permit fluid flow. Thus there must always
all practical purposes. That is, when Eq. 19 is be a period of unsteady-state flow through che skin
at shore times for the physical problem posed by
valid, the term (C/t0 ) is only 0.017 as compared to
well behavior. But nothing ver:y general can be said
unity, and the next to lase term involving "s" will
abouc che potential duration of chis transient period
always be negligible compared to s. Of course, it is
without specifying che nature of the skin effecc.
not apparent that Eq. 18 is a valid approximation for
Hawkins13 pointed out thac one physical incerpre-
times specified by Eq. 19. This can only be
tacion of che skin effect is .an annular region of
established by comparison of results from Eq. 18
permeability k 1 and radius r 1 immediately adjacent
wich the rigorous solution provided by Eq. 9. This
to che well. Hawkins showed that
comparison will be discussed later. Jaeger9 offered
two interesting limiting forms correct for any value 1
of time: ( : ) , • • • • (23)
tD w
= --, for s = 00 where /,, is che penneabilicy of the formation proper.
C If k1 is assumed to be infinitely greater chan k
(perhaps 500 times as great), Eq. 23 becomes:
• • • • • (20)

+
s = -in(::), · · ••· · •· .. (24)

or
for C = 0 • • • (21)
= r e -s ' . . . . . . . . . . (25)
w
Eq. 20 would represent the condition of the sand
face completely blocked to flow, or depletion of the a relationship which is often cited to indicate chac
wellbore volume only. Note that both short-tim~ the skin effect can be interpreted simply in terms
approximating forms (Eqs. 14 and 15), contain tolC of an "effective" well radius, r1 •
as the first term of the series. Eq. 21 presents The point we wish to make is that, even if an
constant-rate production with a skin effect but no annular region of altered permeability is visualized
storage. This solution is in fact the basis for most as physically responsible for the appearance of a
of the pressure buildup and drawdowo testing skin effect, the finite dimensions of the annular
commonly used coday. There is one aspect of this region would lead to unsteady flow within che skin
solution which has not been discussed. Since p0 (t0 ) ac short times. This is the "composite reservoir"
approaches zero as cime approaches zero, Eq. 21 problem described by J aeger9 in connection witb
indicates thac there would be a finite pressure heac conduction and by Louckes and Guerrero14 iIJ
difference between the formation and the wellbore connection wich fluid flow.
as time approaches zero. Although the solution is The steady-state skin effect concept is a useful
mathematically correct, this condition does not one, and it has been established in previous studies
represent the physical fluid-flow problem. That is, that flow in the region near the well does become
steady a short time after production begins.
lim However, there still is a basic problem. As Eq. 23
+0 indicates, there is an infinite number of pairs of
values (k 1,r1) that will produce the same constant
Eq. 21 indicates thac the previous solutions do not skin effect, s. Furthermore, there is undoubtedly an
satisfy Eq. 22 for the special case of zero wellbore infinite variety of conditions near a wellbore which
storage. Eq. 22 is satisfied if the storage constant can also lead to the appearance of a skin effect -
is finite. This results because the sand-face flow whether positive or negative in sign. The problem
rate increases gradually from zero to a conscanc regarding che variety of annular configurations which
value for finite storage, but increases instantaneously can produce a specific skin effect value is not quite
from zero to a constant rate if wellbore storage as bad as it first appears. This problem is discussed
(unloading) is non-existent. by Wattenbarger and Ramey.22 It Is beyond the
The fact that solutions with finite storage purpose of this paper to pursue further che problems
constants appear to make more physical sense at involved in the skin-effect concept. It is important
short times does not indicate validity of the chat che inherent nature of the skin effect be realized
solutions. The problem lies in the basic definition so chat appropriate methods will be used in well-test
of the skin effect as a pressure drop across an analysis.
infinitesimal skin (zero storage capacity, or Finally, Jaeger9 also presented the solution to a
steady-state flow in the skin region). Obviously, heat-conduction problem analogous to that of che
real skins which affect well performance must have fluid-flow problem. The solution involved che

SEPTEMBER, 1970 2B3

sudden opening of a bottom-hole choke in a well
where the tubing had been pressured initially and
C
the wellhead valve closed. .A. skin effect was
included; and Jaeger's solution would permit
estimation of the dissipation of the wellbore pressure lnt
to the formation. Although this problem has no + 0 (-1!) . , (29)
common useful significance in well-test analysis, t 3
the solution is related to the time derivative of D
PwDCtD); thus can be used to compute the wellbore
unloading flow rate, and consequently the sand-face Eq. 29 may also be written in terms of the Pv(tD)
flow rate for the subject problem of this paper. function as per Eq. 16:
Moreover, there is a useful physical interpretation.
We can picture the flow rate variation caused by the
wellbore storage problem to be tbe superposition of
tbe depletion of well casing fluid to a formation
upon the behavior of a well produced at constant
rate.
Converting Jaeger's solution to the fluid-flow +....£._
nomenclature of the subject problem and multiplying
2tD2

1-
by appropriate constants yields:
• • • • • (30)

- dpwD 4c Comparison of Eqs. 26 through 29 with Jaeger's

C d~ = lf2 [
0
-u2ti, Cu + Eqs. 6 through 10 indicates that:9

0
2 C = F (h, a , ,. ) • • (31)
2
u{ [ uCJ0 (u\-(l-Csu )J1 (u)]
where F (b, a, r) is a function evaluated by Jaeger,
2 2 and b, a, and T have the same significance as
+ [ UCY0 (u)-(1..'csu )Y 1 (ul] }] , (26)
indicated in Table I.

Reference to Eq. 7 indicates that Eq. 26 provides DISCUSSION

the fraction of the total surface flow rate produced EVALUATION OF INTEGRALS
from the casing at any time (and thus the sand-face
flow rate by difference), It is clear that Eq. 26 is
In order to determine useful engineering approaches
to interpretation of well-flow tests, the real .inversion
simply the product of C and the time derivative of
integral presented as Eq. 9 was evaluated numerically
Eq. 9. Of greater interest, Jaeger provided short- and
for_ a range of values of time, storage constant, and
long-time approximating forms, as well as limited
skin effect. The line-source approximation given by
tabulations and graphical solutions, for Eq. 26:
Eq. 11 was also evaluated for a range of conditions
to determine the validity of the approximation for
For short time and skin effect, s, not zero:
ranges of parameters met in well testing. This
dpwD tD information is presented in Tables 3 through 8.
C
dtn =
l - -
Cs
+ O(tD 2
":l
), C f-
-
0 Evaluations were made with a high-speed digital
computer using "Richardson's Deferred .Approach to
the Limit", 20 which is a modified form of Simpson's
•• (27)
rule. !ntegrations were performed repeatedly, reducing
the Simpson's rule interval until values of the integral
For short time and skin effect, s, of zero: agreed to within 10·8• Results were also obtained
by a more sophisticated method of numerical
integration, the Romberg21 method, to provide a
C l - (~ -½) check of the computed results. It is not possible to
specify rigorously the accuracy of integration, but
it is believed that results are good to at least three
J significant figures.
2
+ o( tD ) , c Io . . c2s>
RESULTS

For long times, Tables 3 and 4 present results obtained from Eq.
9 for the cylindrical source. Table 3 presents results
for a skin effect of zero, while Table 4 presents
 results for a skin effect of +20. Tables 5 through 8 TABLE 5 - PwD(s, C, tr:) VS t0 FOR s = 0, LINE SOURCE
present results obtained from Eq. 11 for the WELL, P...o(0, c; tx,) FOR ~ OF
line-source well, for skin effects of zero, +5, +10
and +20. Comparison of Tables 3 and 5 and 4 and 8 10 2 10l 10 4 105
indicates thac the results for the cylinder source 100 0.7938 0.09758 0.00998 0.00100
and line source agree within 0.5 percent for all 200 1.3671 0.1918 0.01992 0.00200
500 2.4299 0.4582 0.04956 0.00500
times, wellbore storage conscancs and skin effects 1,000 3.2640 0.8580 0.0984 0.00999
considered. It is common procedure to produce 2,000 3.9254 1.5292 0.1944 0.01995
shore- and long-time approximations such as those 5,000 4.5579 2.8824 0.4697 0.0497
given by Eqs. 13 through 15. In general, it is not 10,000 4.9564 4.0321 0.8925 0,0989
20,000 5.3286 4.9347 1.6274 0.19S8
possible to establish the applicable ranges of the 5.8026 5.6761 3.2108
50,000 0.4765
approximations unless a comparison is made with 100,000 6.1548 6.0940 4.6772 0.9141
the general solution. Eq. 11 is, in a sense, a 200,000 6.5043 6.4736 5.8870 1.6931
long-time approximation for Eq. 9. We have just 500,000 6.9643 6.9515 6.7894 3.4571
1,000,000 7.3116 7.3049 7.2309 5.2163
shown that ic is an exceedingly good approximation 2,000,000 7.6585 7.6S50 7.6185 6.7731
over the ranges of parameters of interest in well 5,000,000 8.1169 8.1154 8.1104 7.8983
testing. Short-time approximations such as Eqs. 14 10,000,000 8·.4635 8.4627 8.4550 8.3701
and 15 are usually given in ascending powers of the 20,000,000 8.8101 8.8097 8.8057 8.7663
50,000,000 9.2683 9.2681 9.2664 9.2523
dimensionless time and, strictly speaking, converge
100,000,000 9.6149 9.6148 9.6139 9.6082
only for values of the dimensionless time less than
unity. Bue comparison with the general solution often
will indicate that the first few terms may provide an
excellent approximation for a much greater time
TABLE 6 -PwnCs, C,t_nl VS_!0 FOR s=..t5, LINE SOURCE
WELL, pwD(5, C,tD) FOR C OF
TABLE 3 - PwD(s, C. tiJ) VS tD FOR s =
O, CYLINDRICAL
'n 102 10 3 104 10s
SOURCE WELL, pwD (0, C, tD) FOR C OF
100 0.9319 0.09929 0.009993 0.00100
ro2 ro3 ro4 10s
200 1.7512 0.1973 0.01997 0.00200
100 0.7975 0.09763 0.00998 0.00100 500 3.6982 0.4843 0.04984 0.00500
200 1.3724 0.1919 0.01992 0.00200 1,000 5.7984 0.9410 0.0994 0.00999
500 2-43S7 0.4585 0.04956 0.00500 2,000 7.8403 1.7820 0.19n 0.01998
1,000 3.2681 0.8585 0.0984 0.00999 5,000 9.3823 3.8349 0.4863 0.0499
2,000 3.9274 1.S298 0.1944 0.01995 10,000 9.8913 6.1533 0.9480 0.0995
S,000 4.S585 2.8832 0.4697 0.0497 20,000 10.300 8.5524 1.8062 0.1979
10,000 4.9567 4.0328 0.8925 0.0989 50,000 10.792 10.436 3.9463 0.il878
20,000 5.3288 4.93S0 1.627S 0.1958 100,000 11.150 11.025 6.4558 0.9536
50,000 5.8027 S.6762 3.2109 0.4765 200,000 11.502 11.445 9. 1982 1.8256
100,000 6. 1548 6 0940 4.6n3 0.9141 500,000 11.963 11.941 11.488 4.0388
200,000 6.5043 6.4736 S8871 1.6931 l,000,000 12.311 12.300 12.156 6.7163
500,000 6.9643 6.951S 6.7895 3.4S71 2,000,000 12.658 12.652 12.589 9.7845
1,000,000 7.3116 7.3049 7.2309 5.2164 5,ooo,ooo 13. 117 13.114 13.090 12.517
2,000,000 7.6585 7.6S50 7.618S 6.nJl 10,000,000 13.463 13.462 13.450 13.286
5,000,000 8.1168 8.1154 8.1004 7.8983 20,000,000 13.810 13.809 13.803 13.734
10,000,000 8.4635 8.4627 8.4550 8.3701 50,000,000 14..268 14.268 14.265 14..239
20,000,000 8.8101 8.8097 8.80S7 8.7663 100,000,000 14.615 14.615 14.613 14.601
S0,000,000 9,2683 9.2681 9.2664 9.2523
100,000,000 9.6149 9.6148 9.6139 9.6082

TABLE 4-PwD(s,C,t0 ) VS tD FOR s=+20, CYLINDRICAL

SOURCE WELL, pw0 (20, c;, TABLE 7 - C.
Pwrf.s, to) VS to FOR•=+ 10, LINE SOURCE
0 ) FOR 't'"OF WELL, PwDCI0, t'; to) FOR l'-OF
tD 102 10 3 104 10 5 103 104 10 5
10 2
100 o.9m 0.09978 0.01000 0.00100
100 0.9594 0.09958 0.01000 0.00100
200 1.9132 0.1991 0.01999 0.00200 1.8463 0.1984 0.01998 0.00200
200
500 4..4900 0.4946 0.0499 0.00500 4..1401 0.4904 0.04990 0.00500
500
1,000 8.1220 0.9787 0.0998 0.0100 1,000 7.0124 0.9629 0.0996 0.0100
2,000 13.479 1.9172 0.1992 0.0200 2,000 10.487 1.8587 0.1985 0.0200
5,000 21.102 4.512S 0.4948 0.0500 5,000 13.852 4.2027 0.4911 0.0499
10,000 24.241 8.1987 0.9797 0.0998 10,000 14.797 7.2010 0.9658 0.0997
20,000 25.186 13.710 1.9209 0.1993 15.269 10.955 1.8693 0.1986
20,000
50,000 25.758 21.786 4.5333 0.4953 15.781 14.811 4.2568 0.4918
50,000
100,000 26.134 25.271 8.2698 0.9810 100,000 16.144 15.917 7.3677 0.9683
200,000 26.494 26.324 13.925 l.92S2 16.499 16.413 11.382 1.8785
200,000
500,000 26.960 26.907 22.443 4.5545 16.962 16.930 15.737 4.3043
500,000
1,000,000 27.310 27.284 26.286 8.3394 17.311 17.295 17.031 7.5162
1,000,000
2,000,000 27.657 27.645 27.460 14. 133 2,000,000 17.658 17.650 17.556 11.nl
S,000,000 28.116 28.112 28.055 23.085 18.117 18.113 18.079 16.631
5,000,000
10,000,000 28.463 28.461 28.434 27.297 10,000,000 18.463 18.462 18.445 18.138
20,000,000 28.810 28.809 28.795 28.606 20,000,000 18.810 18.809 18.801 18.699
so,000,000 29.268 29.268 29.262 29.216 50,000,000 19.268 19.268 19.264 19.227
100,000,000 29.61S 29.615 29.612 29,596 19.615 19.613 19.59S
100,000,000 19.61S

SEPTEMBER, 1910
 TABLE 8 - p.,,c,(s, C, tn) vs tn FOR s =+20, LINE SOURCE
WELL, p (20, c; tD) FOR C°OF • • • • • (32)
wD
fD 102 10 3 10 4 105 C
100 o.9n6 o.099n 0.01000 0.00100 Fig. 1 indicates that the approximation given by
200 1.9130 0.1991 0.02000 0.00200 Eq. 32 holds to dimensionless times on the order of
500 .C..CS96 0.49'6 0.0499 0.00500
1,000 8.1212 0.9787 0.0998 0.0100
0.5 C: to C:. This indicates that early performance
2.000 13.478 1.9172 0.1992 0.0200 is controlled entirely by wellbore unloading. It
5,000 21.101 4.5125 0."948 o.osoo further indicates that the storage constant (but only
10,000 2.C.241 8.1986 0.9797 0.0998 the storage constant) might be determined from early
20,000 25.186 13.709 1.9209 0.1993 field performance data. That is, a plot of the logarithm
50,000 25.758 21.786 .C.5333 0.4953
100,000 26.134 25.271 8.2698 0.9810 of (P; - Pwjl vs the logarithm of flowing time should
200,000 26.494 26.324 13.925 1.9252 have a un1t slope and a value of (q/C) at t = 1.
500,000 26.960 26.907 22.443 .C.5545 .Alternatively, the flowing pressure could be plotted
1,000,000 27.310 27.28" 26.286 8.3394 vs flowing time on Cartesian coordinates and the
2,000,000 V.657 V.6"5 V.'60 1.C.133
5,000,000 28.116 28.112 28.055 23.085 slope would be -{q/C). These interpretations result
10,000,000 28.'63 28."61 28.-434 V.297 because we may substitute Eq. 6 for C: in Eq•. 32,
20,000,000 28.810 28.809 28.795 28.606 and the usual .identities for dimensionless pressure
S0,000,000 29.268 29.268 29.262 29.216 and time:
100,000,000 29.615 29.615 29.612 29.596
2'rkh(pi - Pw:r)
range. The range of validity of long-time approxima- = • • • • • (33)
tions such as Eq. 13 can only be established by
q I'
comparison with the general solution.
Portions of the solutions arc shown on Fig. 1. •
= - -kt
- - • • • • • • • • • (34)
Fig. 1 is a log-log plot such as would be required
if "type-curve" matching of field perfoanance were
¢"' crw2
to be used. Beck et al. 1 6 illustrated type-curve
matching of borehole unsteady-temperature data to to achieve
detennine thermal constants of well systems. The
similar appearance of the computed curves for
various storage constants and skin effects indicates
= <!> t · · ...... (35)

that this sort of performance data interpretation

This approach may be useful to determine the storage
might be rather poor for fluid flow systems. There constant where specific information is not available.
is one useful technique illustrated by Fig. 1, Perhaps more important, this analysis indicates that
however• .All curves approach a unit slope at small
well pressure data cannot be used to determine
times for finite, positive skin effects, and a skin
either s]cjn effect or flow capacity before times
effect of zero. This condition results because only
when the dimensionless pressure on Fig. 1 begins
the first tenn of the series in Eqs. 14 and 15 is
to bend below the early straight line dictated by
important at very small times. Thus at very early
Eq. 32.
times,
A pertinent question is what is the duration of
this initial period? If we assume that a time
•A full•alze plot of Fie, l (with • cdcl) will be avallabla tluous!,
the American Doi:..-entatlon Inatltute, Llbr..,, of Concre••• sufficient for the sand-face flow rate to become 20
Waahlncton, D.C. At preaent, lt may be obtained by writlnc to percent of the surface flow rate is adequate, Eq. Tl
H.J. Ramey, Jr., Dept. of Petrol•-= Enclneedna;, Stanford U.,
Stanford, Calif. 94305. can be solved approximately to provide

= 0.2 cs, s f O· • (36)

SKIN.I
Using the same criterion, Eq. 28 may be solved
approximately to yield

= o.4 c, s = o. • . (37)

It is recommended that Eq. 37 be used for skin

effects ranging from zero to +2 and Eq. 36 be used
to estimate the time for skin effects greater than +2.
The reason for this recommendation may be seen by
inspection of Fig. 1. The dimensionless times that
r:,I pressures depart from Eq. 32 increase as the skin
104 105 10 7
effect increases above zero. Because Eq. 36 is only
l)IMEHSION1.ESS TIME, fo
an approximate simplification of a very complex
FIG. 1 - P.,,n VS 'n FOR INFINITE RADIAL SYSTEM form, it would forecast smaller times than Eq. 37 if
WITH STORAGE AND SKIN EFFECT. the skin effect is between zero and +2.
 In summary, data obtained from the start of a test procedure could be used foe postt1ve skin effects
until dimensionless times specified by che and that only a tabulation for a skin effect of zero
appropriate Eq. 36 or 37 may be used co estimate is necessacy. Even if such is true, and it is a good
che wellbore storage coefficient, C. Well-test data approximation, this would noc be practical because
obtained for times between the appropriate Eqs. 36 tabulations for a tremendous range in dimensionless
or 37 and Eq. 19 may hopefully be interpreted for time '\llrould be required.
skin effect and formation flow capacity. Well-test Finally, the face thac che infinitesimal skin
data obtained for times after chat specified by Eq. effect is not a valid concept for shore times is even
19 can be analyzed in the normal fashion. The more: evident for the negative skin effect case
method of interpreting the intermediate time data shown on Fig. 1 than for the positive skin effect
remains to be shown, although we would expect that cases. Note the peculiar behavior for the zero
the modified Gladfelter el al. 17 method discussed storage constant case on Fig. 1 for s = -5.
by Rameyll would be appropriate.
One problem remains. Nothing yet has been said INTERPRETATION OF
SHORT-TIME WELL-TEST DATA
concerning results for negative skin effects. In the
heat-conduction analog to the subject fluid-flow le has already been shown that the initial pressure
problem, the skin effect is analogous co a convection data obcained from the start of a test until
heat transfer film coefficient. Heat transfer dimensionless times of the order of those from Eqs.
coefficients always have values equal to or greater 36 or 37 may be interpreted to obtain the wellbore
than zero. There is nothing in the statement of the storage constant only. The problem we consider here
fluid flow problem (Eqs. 1 through 5) that precludes is whether the intermediace shore-time pressure data
the possibility of a negative skin effect. The can be incerpreced for the formacion flow capacity
uansform solution, Eq. 8, and the line-source and the skin effect.
transform, Eq. IO, are both valid for a negative skin One cechnique chac has been used extensively in
effect. Furthermore, the skin effect term in all of the field of groundwater hydrology is "type-curve
the long- and shore-time approximations may be matching".18 Drawdown data would be plotted as
positive or negative. But the real inversion integrals P; - PwJ in any convenient units vs time in any
given by Eqs. 9 and 11 apply only for skin effects convenient units on log-log paper of the same size
equal to, or greater than, zero. cycle as Fig. 1. The curve would be moved over
Eqs. 8 and 10 can be inverted for negative skin Fig. 1 until the best match with one of the
effects, but this does not appear worthwhile. An precomputed curves was obtained. The storage
alternate procedure that makes more physical sense constant and skin effect could be read direccly from
is to postulate that a negative skin effect can be the curve matched, and the flow capacity could be
represented physically by an increased wellbore obtained from any pair of matching dimensionless
radius. We recognize that many factors could be the time and real time from the two graphs, or from any
actµal cause of the appearance of a negative skin pair by matching dimensionless pressure, PwD• and
effect. For example, one would expect that the pressure difference, P; - Pwf· This procedure has
result of acidizing a well would be to increase the the theoretical capability of also yielding the
specific permeability of che rock matrix adjacent to hydraulic diffusivity, k/¢ µ c, from matching times.
the well. This situation could be approximated by As in pressure buildup or drawdown, an accurate
Eq. 23. If the increase in permeability were large, value of rw is required.
then Eqs. 24 or 25 could be used to represent this In o.r;der to apply the preceding to pressure
physical situation. The proper procedure would be buildup, superposition may be used in the normal
to use either of the real inversion integrals given
by Eqs. 9 or 11 for a zero skin effect, but evaluated TABLE 9 -PwD(.s,C,tD) VS 'D FOR s=-S, CYLINDRICAL
for a dimensionless time, t0 ", and a wellbore storage SOURCE WELL, Pwo(-5, C, tD) FOR C OF
constant, C:", defined as foliows. tD 102 l0l 10 4 105
100 0.0697 0.0447 0.00896 0.00099
25 • (38)
t "
D = 'tn (r.jr1)2 = tD e 200 0.0992
0.1S57
0.071S
0.1263
0.0172
0.0394
0.00197
0.00487
500
1,000 0.2164 0.1872 0.0718 0.00963
c" = C (r.jr1 )2 = C e2s • · (39) 2,000
5,000
0.2977
0.4446
0.2697
0.4199
0.1267
0.2518
0.01896
0.0458
10,000 0.S913 0.S701 0.3990 0.0879
Evaluation of Eq. 9 for a skin effect, s, of -5 and 20,000 0.7722 0.75-48 0.5972 0.1655
a range of storage constants, C, is presented in S0,000 1.0646 1.0S23 0.9313 0.3622
Table 9 and also shown on Fig. I. These 100,000 1.3232 1.3145 1.2254 0.6219
200,000 1.6086 1.6028 1.S422 0.9926
dimensionless pressures are given as functions of S00,000 2.0170 2.0139 1.9806 1.6088
t 0 and C, since a skin effect of -5 listed as a 1,000,000 2.3420 2.3401 2.3201 2.0895
parameter v.ould be observed in field operations. le 2,000,000 2.6757 2.6747 2.6630 2.S324
should be emphasized that dimensionless pressures 5,000,000 3.1248 3.1243 3.1187 3.0598
10.000,000 3.4677 3.4615 3.4644 3.4323
were found using Eq. 9, s = 0, and t 0" and C:" 20,000,000 3.8124 3.8123 3.8107 3.7932
values which are e· 10 times the t 0 and C: tabulated. so.000,000 .C.2693 4.2693 4.268S 4.2608
At first glance, it would appear chat a similar 100,000,000 4.61S4 4.6154 4.6150 .C.6108

SEPTEMBER. 1970 2BT

 manner. That is, solution with reasonable accuracy. The storage
constant C is equivalent to the reciprocal of the
~[p.-p (t+l:it)] = Pwn(s,C,t+At) storage constant, C1 , used in Ref. 15. Of interest
q µ 1 W at this point is the range of validity of the
approximation given by Eq. 44. Fig. 2 presents
- Pwn<s,C,ll.t) . (40) p n<s, C, tn) as determined from the rigorous Eq. 9,
a:d the approximation given by Eq. 44. Results are
shown for s = O and 20 and for E = 1,000. As can be
~'!11 (pi-pwf) = Pwn (s,C,t) . (41) seen, the range of validity of the approximation is
limited. But the long-time approximation has the
same general shape as the rigorous solution,
where t is the flowing time before shut-in, and 11t providing an empirical basis for the method.
is the time since shut-in. Combination of Eqs. 40 Eq. 44 is interesting for another reason. It is
and 41 yields: dear that the time criterion given by Eq. 19 may be
substituted in the last bracket on the right in Eq.
:'!11 [P./t+l:it)-pwf] =pwD(s,C,t)
44 to yield:

~ : [pw< t+l:it) - Pw.r] = [ Pr> + s]

-Pwn< 6 ,c,t+At)+J>wn( s,C,f.t) · · C42>
• [ 0.9833] · · · (45)
If 11t is small compared to t, t+11t_is approximately
equal to t and the first two terms on the right in Eq.
42 are nearly equal. Thus, This shows that the time criterion of Eq. 19
provides the time at which the approximated PwD is
~ Pwn(s,c,At) within 1. 7 percent of (PD+s). A dimensionless time
of 60r is also shown on Fig. 2•
• • • • • (43) CONCLUSIONS
The log of (fiw(t+/J.t}-p.,,t' should be plotted vs the A thorough investigation of both the wellbore
log of l!i.t on a uansparent log-log coordinate of the storage and skin effect concepts was performed. It
same size as Fig. 1. If the initial buildup data was found that many studies have been made of
display a line of unit slope, q/C may be read from these effects, although the connection between
the pressure scale at a l!i.t of unity as discussed various studies has not been obvious. It is hoped
below Eq. 32. Eq. 6 may be used to compute the that the review presented in this paper will help
appropriate C. The plot may then be placed ov';_r clarify this important problem.
Fig. 1 using the family of curves for the proper C. In regard to wellbore storage, it has been shown
The curve should be positioned such that it matches that it is possible to forecast the duration of the
the family of precomputed curves on Fig. 1. Then initial flow period controlled by storage. During
the permeability, k, may be determined from Eq. 33 this time, it is possible to find the storage constant
using the ~rresponding values of PwD and from well-test data, but the fomiation flow capacity
[p 111(t+l!i.t)-p"'J; and corresponding values of to and and skin effect cannot be found. After the initial
l!i.t and Eq. 34 evaluated for t = 11t. An example
calculation is shown in Ref. 19.
Another possible interpretative scheme is to match
field data against some approximating equation such
as the long-time expression given by Eq. 18. This
is essentially the procedure recommended by
Russell 15 for short-time pressure buildup analysis.
Eq. 5 in Ref. 15 may be rearranged to the form:
-EQ.9
2Jrkb --- EOS.18 OR 44
=
qµ

• • • • (44)

which was pointed out as an acceptable approximation f04 ,o5

for the equivalent of Eq. 43, employing the long-time DIMENSIONLESS TIME, t 0
approximation given by Eq. 18. This procedure is
FIG. 2 - P..,D VS ID FOR LONG-TIME APPROXIMATION
valid as long as Eq. 44 represents the rigorous AND RIGOROUS SOLUTION.
 period controlled by storage, it is possible to perform PwD = dimensionless pressure drop within the
analysis of well•test data for flow capacity and wellbore
skin effect by type-curve matching techniques. This Po = dimensionless pressure drop on the formation
procedure extends well-test analysis into the side of skin effect
shore-time region and provides an additional useful P; = initial formation pressure
interpretative tool.
It has been recognized, in regard to skin effect, Pw = well pressure
that the steady-state skin effect concept becomes Pw/ -= flowing well pressure
physically invalid at very shore test times. If test q = surface flow rate
data are to be interpreted properly, it will be qs/ = sand face flow rate
necessary to generalize the skin effect concept to r = radial distance
include a damaged region of a finite storage capacity. r 1 = radius of the region of permeability k 1
Regarding negative skin effect, one possible
interpretation is that the wellbore radius has been r0 = dimensionless radius, rlrw
enlarged. Transient solutions have been presented rw = well bore radius, cm
for this case. It is also likely that negative skin s = skin factor, dimensionless
effects arise more commonly as a result of fracture t = flowing time, seconds
communication with the wellbore. This latter At = shut-in time, seconds
situation is not considered in this study.
t0 -= dimensionless time (see Eq. 34)
In regard to the combined effect of wellbore
storage and skin, it appears that the time required u = variable of integration
to reach the usual straight line is not usually Yo = Bessel function of second kind, order zero
affected significantly by a finite skin effect. The Y 1 = Bessel function of second kind, order one
duration is mainly controlled by the wellbore storage. a = reciprocal of C: (see Table I)
A systematic inspection of a number of significant y = Euler's constant, 0.57722
factors affecting shore-time transients involved in
µ. = viscosity, cp
well testing has been made. It is not implied,
however, that all significant factors have been a = infinity valuer defined by Hurst (see Table I)
studied. Nevertheless, a number of pertinent r = dimensionless time (see Table I)
findings have resulted. Wellbore storage should cf, = porosity, fraction of bulk volume
control the initial transients if the well is damaged
or has a zero skin. If a well exhibits a negative SUBSCRIPTS
skin, the effect of wellbore storage will not be D "' dimensionless quantity
prominent because the rate of pressure change in i = refers to initial reservoir condition
the wellbore may be quite low. It appears that proper sf= refers to conditions at sand face
interpretation of short-time well-test data can be
w = refers to conditions at wellbore radius
made under favorable circumstances.
wf = refers to flowing conditions at wellbore
NOMENCLATURE radius

c = total system isothermal compressibility, ACKNOWLEDGMENT

vol/vol-atm
This study was initiated while the authors were
C = wellbore storage constant, cc/atm (see Eq. 6) at Texas A&M U. Much of the initial computer work
C1 = equivalent to the reciprocal of C (used by involved in evaluation of real inversion integrals
Russell 15) was accomplished with financial support of the
C = dimensionless wellbore storage constant school. The authors also wish to acknowledge the
(see Eq. 6) encouragement of R. L. Whiting, head of the Dept.
d = differential operator of Petroleum Engineering at Texas A&M, and the
b = formation thickness contributions of R. D. Carter of Pao American
Petroleum Corp. and R. A. Wattenbarger of
JO = Bessel function of first kind, order zero Scientific Software Corp.
/ 1 = Bessel function of first kind, order one
k = permeability of the formation
REFERENCES
k 1 = permeability of the region immediately
adjacent to the well 1. van Everdingen, A. F. a.ad Hurst, W.: "The Applica-
tion of the Laplace Transformation to Flow Problems
Ko = modified Bessel function of second kind, Trans. AIME (1949) Vol. 186, 305-324.
order zero
2. van Everdi.agen, A. F.: "The Skin Effect and Its
K1 = modified Bessel function of first kind, order Influence on the Productive Capacity of a Well",
one Trans. AIME (1953) Vol. 198, 171-176.
L = Laplace transform of the quantity 3. Hurst, W,: "Establishment of the Skin Effect and Its
Impediment to Fluid-Flow into a Well Bore", Pet.
p = variable of Laplace transform Eng. (Oct., 1953) Vol. 25, B-6.

SEPTEMBER, 19TO
 4. Jaeger, J. C.1 "Radial Heat Flow in Circular 13. Hawkins, M. F., Jr.: "A Note on the Skin EfCect",
Cylinders with a General Boundary Condition, I", Trans. AIME (1956) Vol. 207, 356-357.
J. Roy. Soc., N.S, Wales (1940) Vol. 74, 342. 14. Loucks, T. L. and Guerro, E. T.: "Pressure Drop in
S. Jaee:er, J.C.: "Radial Heat Flow in Circular Cylinders a Composite Reservoir", Soc. Pet. Eng. J. (Sept.,
With a General Boundary Condition, ll", J. Roy. Soc., 1961) 170-176.
N.S. Wales (1941) Vol. 75, 130. 1S. Russell, D. G.: "Extensions or Pressure Buildup
6. Carslaw, H. S. and Jaeger, J. C.: Conduction of Analysis Methods", J. Pet. Tech. (Dec., 1966) 1624-
Heat in Solids, 1st ed., Oxford at the Clarendon 1636.
Press (1947) 16 and 284; 2nd ed. (1959) 22 and 342. 16. Beck, A., Jaeger, J. C. and Newstead, G.: "The
7. Lesem, L. B., Greytok, F., Marotta, F. and McKetta, Measurement or the Thermal Conductivities of Rocks
J. J., Jr.: "A Method or Calculatine: the Distribution by Observations in Boreholes", Aust. J. Phys.
or Temperature in Flowine: Gas Wells", Trans. AIME (1956) Vol 9, 286.
(1957) Vol 210, 169-176. 17. Gladfelter, R. E., Tracy, G. W. and Wilsey, L. E.:
8. Hurst, W.: "The Simplification or the Material Balance "Selecting Wells Which Will Respond to Production-
Formulas by the Laplace Transformation", Trans. Stimulation Treatment", Drill. and Prod. Prac., API
AIME (1958) Vol. 213, 292-303. (1955) 117.
9. Jaeger, J. C.: "Conduction of Heat in an Infinite 18. Hantush, M.: Advances in Hydroscience, Ven Te
Region Bounded Internally by a Circular Cylinder of Chow, ed., Academic Press, New York (1964) Vol. 1.
a Perfect Conductor", Aust. J. Phys. (1956) VoL 9, 19. Ramey, H. J., Jr.: "Short-Time Well Test Data
No. 2, 167. Interpretation in the Presence of Skin Effect and
10. Blackwell, J. H.: "A Transient-Flow Method for Wellbore Storage", J. Pet. Tech. Uan., 1970) 97-104.
Determination of Thermal Constants of Insulating 20. McConnl.ck, J. M. and Salvador!, M. G.: Numerical
Materials in Bulk", J. Appl. Phys. (1954) VoL 25, Methods in FORTRAN, Prentice-Hall, Inc., Englewood
No. 2, 137. Cliffs. N. J. (1964) 54.
11. Ramey, H. J., Jr.: "Non-Darcy Flow and Wellbore 2L Davidi P. J. and Rabinowitz, P.: Numerical Integration,
Storap Effects in Pressure Buildup and Drawdown Blaisdell Publishing Co, 1 Waltham, Mass. (1967) 166.
of Gas Wells", J, Pet. Tech. (Feb., 1965) 223-233.
22. Wattenbarger, Robert A. and Ramey, H. J., Jr.: "An
12. Chatas, A. T.: "A Practical Treatment or Nonsteady- Investigation or Wellbore Storage and Skin Effect in
Flow Problems in Reservoir Systems", Pet. Eng. Unsteady Liquid Flow: IL Finite Difference Treat-
(May, June and Aug., 1953) Vol. 25, ment", Soc. Pet. Eng. J. (Sept., 1970) 291-297.
***
 An Investigation of Wellbore Storage and Skin Effect in
Unsteady Liquid Flow: Il. Finite Difference Treatment
ROBERTA.WATTENBARGER• MOBIL RESEARCH & DEVELOPMENT CORP.
DALLAS, TEX.
H. J. RAMEY, JR. STANFOROU.
MEMBERS AIME STANFORD, CALIF.

ABSTRACT 'Wellbore storage is caused by having a moving

An investigation of the effect of wellbore storage liquid level in a wcllborc, or by simply having a
and skin elfect on transient flow was conducted volume of compressible fluid stored in the wellborc. 3
using a finite-difference solution to the basic When surface flow races change abruptly, wellborc
partial dilferential equation. The concept of skin storage causes a rime lag in formation flow rates
eflect was generalized to include a damaged and a corresponding damped pressure response.
annulw region adjacent to the well bore ( a composite A recent study4 was made to determine the
reservoir). The numerical solutions were compared combined effects of infinitesimally thin skin and
with analytical solutions for cases with the usual wellborc storage. Analytical methods were used
steady-state skin elfect. It was found that the along wich numerical integration of a Laplace
solutions for a finite-capacity skin effect compared transformation inversion integral. Tabular and
closely with analytical solutions at short times graphical results were presented for various cases.
(wellbore storage controlled) and at long times It was recognized during the study that this
after the usual straight line was reached. For representation of skin was oversimplified; that
intermediate times, presence of a finite-capacity skin effect should be thought of as a result of
skin elfect caused significant departures from the formation damage or improvement to a finite region
infinitesimal skin solutions. Two straight lines adjacent to the wellborc.
occurred on the drawdown plot for cases of large It was suggested that a skin effect could arise
radius of damage. The first had a slope physically in a number of ways. One simple example
characteristic of the flow capacity of the damaged would be to assume that an annular volume
region; the second straight line had a slope adjacent to the wellbore is reduced uniformly to a
characteristic of the flow capacity of the lower permeability than the original value. This
undamaged region. Results are presented both in would be similar to the composite reservoir
tabular form and as log-log plots of dimensionless problem. Perhaps a better example would be to
tressures vs dimensionless times. The log-log assume that the permeability increases continuously
plot may be used in a type-curve matching from a low value at the wcllbore to a constant
procedure to analyze short-time (before normal value in the undamaged reservoir. In either case,
straight line) well-test data. the damaged region would have a finite storage
capacity and would lead to transient behavior
INTRODUCTION within the skin region.
A negative skin effect could arise from an
Skin effect was defined by van Everdingen 1 and increase in permeability within an annular region
Hurst2 as being an impediment to flow that is adjacent to the wellbore. This might physically
caused by an infinitesimally thin damaged region result from acidizing. Bue it is believed that cases
around the wellbore. The additional pressure drop of more practical importance arc those in which
through this skin is proportional to the wellbore negative skin effects arc caused by hydraulic
flow rate and behaves as though flow through the fracturing. A high-permeability fracture
skin were steady-state. communicating with the wcllborc gives the
appearance of a negative skin effect.
Orislnal inanuacrlpt received In Society of Petroleuin Enalneers For the purposes of this study, it was decided
orfice Jan. 16, 196!1. Revised inanu•crlpt of paper SPE 2467 to represent a skin effect, either positive or
received Sept. ISi. 19159, @Copyrisht 1970 American Institute
of Mlnlnc, Metalluralcal, and Petroleum Enclneera, Inc. negative, as an annular region adjacent co the
lR,1f•rencu slven at end or paper. wellbore with either decreased or increased
•PreHnlly with Scientific Software Corp •• Enalewood, Colo.
permeability. This then is the composite reservoir
Thi ■ paper will be printed In Tranaacllona voluine 249,
•ihlch wlll cover 1970. problem wherein a permeability 1 1 exists from the

SEPTEMBER, 1970 291

 well radius to a radius of the damaged region, 'l· Outer boundary condition,
For composite reservoirs. there are an infinite
number of pairs of values, , 1 and k 1, which
correspond to a value of skin effect, s. The main
p = 0 · • · .••.. (7)
purpose of this study was to investigate the 20
behavior of a well during inicial transient flow in And the initial condition is,
the presence of a finite-capacity skin effect with
the presence of wellbore storage. The goal ~·as to
provide information which might be useful in
interpretation of short-time well-test data (either
buildup or drawdowo).
The dimensionless variables have usual definitions,
The initial-value problem is a special case of a
composite reservoir problem wherein porosity,
viscosity and compressibility are the same for = r/r , · • • • · • • • • · • (9)
both damaged and undamaged formation regions, w
but with the permeability changing at the boundary
of the two regions;· the complication of wellbore
storage is also added. The skin effect, s, docs not = _k_t_, . . . . . . . . . . . (10)
appear explicitly in the formulation of the problem, 2
qµcrw
bur may be inferred from the steady-state pressure
drop through the damaged annular region adjacent
to the wellbore.
2nkh
= qµ
The initial-value problem may be stated as
follows. For the "damaged" or skin region,
•• (11)

2TTkh
= (pi-Pr,t) ;
qµ

(1) . . . • . . • • (12)

For the undamaged formation,

2TTkh {p.-p ) . . . (13)
qµ. I W

C = C
2nhqx:r 2 '
. . . . • . . . . (14)
w
rlD<rD ... m (2) where C represents the fluid storage capacity in
the wellbore, cc/attn.
Inner boundacy condition, The equivalent steady-state skin effect, s, can
be expressed as a function of k, k 1 , , 1 and rw.
Thar is,
= 1 k r
C s = ( - 1) In (-1)
~ r
w
(3)
= c.L - 1) ln rlD • • • . • (15)
kl
PwD(to) = plD(i,to)
Thus, it is possible to select appropriate values
for a JJ t (4)
of rlD and (k/ k 1) to provide a specific value of
0
the skin effect. The relationship between the skin
Interface conditions between skin region and
effect and the pressure drop attributable to the
formation, skin is

(5)
s =2nkh
- - 6pskin · • • (16)
qµ

Strictly speaking, the above problem may be

k solved to provide the pressure at any time and
radial location within either region. But since rhe
 main objective was information for well-test TABLE 1 - VALUES OF •10 AND (lei/le] USED FOR VARIOUS
analysis, only pressures at the well were sought. SKIN EFFECTS
All results reported in the study are for Skin-Zone
dimensionless pressures at the well which are Radius-, rm (lei/le) for Skin Effec:ts, s, of

denoted as PwD (to). -5 +5 +20

Although the subject problem may be handled 10 0.3153 0, 1032
100 0,4794 0, 1872
analytically with ease, the analytical result would 1,000 3,6209 0,5801 0,2567
require numerical integration to provide final
tabulations of the dimensionless variables.
Fortunately, a finite-difference computer program
was available that could readily handle the
problem. The program was a one-dimensional
radial model prepared to solve real gas flow with
damage and wellbore storage. The details of this
TA.81.I! 2- - l',,,,o ' °' YS 'a P'OR •• O

_!.ci,_ _£_!,_ _IQ_ ~ ~ ~ ~

program have been described previously. 5,6
0.77935 0.090124 0.009t7U 0.000ff71t 0.00009Hl2 0.000009'122
Solutions for liquid flow were obtained by employing 1.00.. 0.l72J4 0.019<-59 O.OOINlC 0.000ltNl 0.-19'64
1,:,,,3 0.3114U 0.04"4 o.oomn 0.00041tf7 0.0000499 IQ
constant-fluid physical properties. The U411 Urntt 0.095.51 0.009':112 0.0009'771 0.00009'111
1.9513 LIQ.. 0.11.5" O.0IH21 0.001"'° 0.OOOlflO
finite-difference solution represented a finite-radius 2.3116 1.1307 0.43514 0.0,,21, 0.OOffl<O 0.0004t9GC
2.7164 2.lllt 0.7H41 0.097.oll 0.009'57t 0.000"7!1
reservoir. Bue the reservoir radius selected was 3.05'0 111'4 L:1659 O.lt1A6 o.o ...... 0.001""
UCH :u:1:211 :u226 0.4.5722 0.OA9A67 0.0049161
large enough so that pressure at the outer boundary l.lSl.t 1112' 3.2"1 D.U.575 0.09124l 0.00996,IJ
.. 1... 4.1767 3.tUt 1.sm 0.1'299 0.019907
was not affected for producing times of tv x 10s. USd
5.0001
.....,3
Utsl
U40
....A6. 2.1641
4.0120
0.A6UI 0.0495'2
0.""7 0.09U54
A check of che outer boundary pressure was made 5.3"5 S.3<:11 S.3114 4.tl,O U20t 0.ltUC
a.as:14
for each run. In addition, all solutions were found
to agree closely with the infinite-reservoir
...,,,
J.I0:13
,.1sn
u52,
S.1021
,.1 ..5

'·"'"
um
S.1'13
6.1439
, ...,2'
,.u12
S.'617
,.012:1
f.."6lS
,.nu
1190•
USlt
S.Nlt
6.7714
0.tll22
l..6165
1-WI
s.1..,
1.,..,1
1.2993 7.2993 7.29U 7.2tlt 7.2169
analytical solution for the longest producing times
,..... ,_,,.,
7.6<5I 1.IUI 7.&•12 7MA2 ,.1:11:1
,.,02, l.'I02.C I.IQ22 1.1007 a.au, 7.17..
run. Since solutions to the diffusivity equation • .... 4 0.4414 I.U16 ....ltt l.:l.504

were obtained in terms of pressure, the usual

1.1',0
t.2$17
t.SfU
t.2:$17
t.ffll
'·""
t.2Sl7
f . .5911
1.7942
,.uu
f •.5911
1.7'02
l.2ffl
,.m2
1.7472
t.222'
t.!112
assumption of small pressure gradients throughout
the flow system was made.
Given a value of equivalent skin effect,
corresponding values of riv and (ki/k) were used
TAILE 3 -11n (tn) VS ,,,FOR•• ... S. "U> 1.000
in the model. Values of riv = 1, IO, 100 and 1,000
were used to give a range of conditions. The case ..£..!... _to_

o.--
__!.R_ ~ ~ _l!,!2L 100 000
I 0.Jl47• 0.010373 0.OC,7524 0.000fl,U 0.0000ff7tt
of riv = l is equivalent to the van Everdingen-Hurst
0.-121
2 0•• 1:177 0.IU47 0.019266 0.0019193 0.000,n~ 0.OOOOlffU
5 o.52~1 0,215ll 0.0A670, 0.OO .. S7S 0.0004'177
infinitesimally thin damaged .region. When no I w i0 1 0.61-'22
o.706I,
0.C31S9' o.oetS71 0.00911tJ 0.0009f71I o.ooonn12
2• 10 1 0•.51973 0.16676 0.0ltstl 0.0019'%7 0.OOOltffl
wellbore storage is present, the solution to this s~. 10• 0.130.U a.mu 0.3'°'9 0.040lt 0.004'7l7 0.000ffl'2
Ill. JOI a.u.. 1 o.ttut O.J.591t 0.09:1311 0.009tlAJ 0.cootf7$4
case can be obtained by adding the skin effect to 2ll.
,,.,
10 1
101
1.019' 1.00.. 0.79511 0.17733 o.Olt7ll 0.001"40
l.1AS 1.1392 1.0,11 a.. . . 0.04609 0.00"'779
the s = 0 solution. When wellbore storage is Ix
2..-
10J
10>
1.2412
1.:1367
1.2J77
I.Jl4
1.2002 uau 0.095162 0.009':MI
1.3163 0.ff432 0.11302 0.01'711
present, the solutions to this case are equivalent ... 102 I.A621 LA620
LSSll(I
1.,1543
1,5540
1.327• 0.41:ZU 0.041t26
h-10' 1.$514 1.$021 0.70IS7 0.096:IIM
to those solutions given by Agarwal et al. 4 h 10'
,~ 10'
um um 1.6.516 1.6277 1.0970 O.IN41
l.7I03 L7102 L779J 1.76'7 1.5772 0.427:IA
A negative skin effect in a composite reservoir Ix 10 1
2x.,.
1.1111
2..0052
I.UIQ U,05
2.00 ..
U751
2.0012
1.1023 0.7J2<S
2.0052 l.'61' UIS3
.implies t},at the region around the wellbore has Sx 101
,,.,. 10•
2.22511 2.2357 2.ms
2. ... ,
2.2222 2.20t2 1.1477
2."'669 2."669 2.A6S2 2. ...3 2.2541
improved permeability, k 1 > k. It should be noted 2,r 10•
J,r 10•
2.7<02 2.7<02 2.7401
2.1471
2.7391 17219
11 .. ,,
2.6154
J.1472 J.1472 11"'6 :LOUI
that for a given radius of permeability improvement,
r 1v, there is an upper limit to the magnitude of
Ix IO'
2x 10'
J. 10'
Ix 10-
3.4754
11125
1.lUI
1.6011
2.054
11125
L'JUI
uoaa
147Sl
J.1125
.&.26,&J
uoaa
1D50
:1.112:1
4.2640
...-1
2.47»
2.1106
4.2133
.__3
......'°.,..,
1 .....
J.79:14

negative skin effect. This limit is reached as the

permeability around the wellbore approaches
infinity. From Eq. 15, it can be seen that the
TAIU.I ~ - ~ (rg) YS •aP'OR ••+ ,- 'ao• I
limiting negative skin effect (as k 1 ➔ co) is
..!it_ ~ _10_ _ ~ --2:!2L ~ 100.000

s = - ln r 10 (17) I
2
5
,.....
S.77115
1.3'05
0.-41
0.1"03
O.DIOI
0.009972•
O.0IH21
0.04fftS
O.OOOHll2
0.0019'61
0.00....,
o,_.,.11
0.0001fl63
D.00049907
o.ooooottau
O.OOOOIHS.
~
Ix l0 1 UJ23 O.t2I07 0.099000 0.00997:it o.ooottlll 0.-
2,i. 10J um L7112 O. lff50 0.01":ll 0.001ff6I 0.00019'62
J:,,..10' 7.2721 :U239 o.-3 o.om2, 0.00 ....2 0.000-
Table 1 shows the values of (k.i/k), which Ix 102 7.7076 5.l657 0.'2990 O.0fflOI 0.0099750 0.000HIIZ
2 .... 102 a.oar 7.0361 1.7A63 0.lt6t2 0.019'37 O.OOltffl
correspond to the various values of r1v and s in
the study. The first two entries for s = -5 are
5x 102
Ix lo-'
2v IOJ
.......
1.,001 t.Z5S6
L7Cf6
2.6121
S.7702
o.•l26
0.t:Jll7J
O.CMffd
0.""21/
0.004fflJ.
0.009'760

blank because this negative skin effect is not

s~toJ
IY 10'
'·"" ,.....
UA60
t.9920
,.1399
, ••260
7.IOII
t,3341
t.111•
1.7767
3.116'
6.1196
O.lt721
0.4521
O.t4.S72
O.0IH'1
o.o-imo
0.0tf37t
2x 10' IO.Jll IG.l:12 IG.211 LSOOO 1.1005 O.l'75.S
obtainable for r 10 = 10 or 100, according to Eq. 15. , . 10'
, ., 10•
t0.794. IG.792 IQ.m 1G.:lt6
11,002
1t2St 0. ...11
O.tSlll
11.1'I ILl<O 11.130 6.4242
2• I01 11.47 11.47 11.41 lt.•2J ,.1,n 1.120,
S• JOI n.,.... 11.,.... 11.t•I II.tit 11.'4I .a.Oltl
RESULTS Iv lo"
2.,, 10•
12.2'1 12.2'° 12.21, 12.271 12.131 Ul:14
12.636 12.636 12.636 12.630 12.SU t.7270
S w 10 1 IJ.09• 11ot, 11093 11091 11066 12.A67
Finite-difference solutions were generated for J,. IO'P
2x 107
13.4"0
13.716
12. ...
1:1.716
12..Qt
12.716
IJ.01
l2.7U
l:L...
12.779
13.256
13.701
a total of 60 cases. Tables 2 through 11 give the
results of these computer runs. Each table gives
s. 10'
h,ol ,.....
, .. 2,3 14.24'
1-.,a,
l'-243
14.5119
14.243
1.....
14.:IAO
U.5111
14.213
ILSU

SEPTE!11BER, 1910 293

 ' '°' 'o'O• ·- +S, •ao• IO the results for given values of s and TlD· In each

__
TAAL.Is -"D vs
table the various wellbore storage cases are listed
...!Ir..- ..!:!..
.__
10 IGO
~ ~ ~
ranging from C = 0 to C = 100,000.
, , '
I
2
s
lx10'
.__
2.0007
2.,...
2.7M7
D.O,.ff3
O.IUl6
o•.am•
0.12450
0.0Dff2l7
D.Olttl•
D.IUl:ut
O-Of7111
0.00099772
D,DDlt9'9
D.-
0-0Mfflt
o.-17
0.00Cllfff3

O.OOOfN02
o.-.:n
D.OOC,llt9'5
D.DDOOfflll
o.oooo,,e21
Only the solutions at selected values of tD are
shown. Smaller time steps were taken between the
2k I0 1 Uf03 D.192:» D.D'ltfDD D.OOltf.ll O.DOOlfff•
Sx: ta• s.aa 2.fll7 o.aoo, 0.0- 0.00.... , o--. printed results to reduce the time truncation error.
..,,72
hio"
2.1oz '·""
7.6356
LJnl
'-31»
1.00'5
O.tO.. I
1-'U2
:un,
0.-211
0.1ff2S
o ... u2
a.DOff721
O.Oltt30
0.0'9730
0.0009tl12
O.OOlfff 1
0.00...tJ A total of 816 time steps were taken for each case.
'" 10"
htol 1-1171 IMl7 U.001 o.n». D.Off1'3 O.DOH7.16
The cases of infinitesimally thin skin (r10 = 1)
, ... 27

,....,
hlOS t.1017 7.6"7 1.7691 t.lt71t O.Oltf40
UI.,
S• lOI
hlO'
t.62.16
t.t7U
t.300
t.1627
3.IO.O
,.10:1&
D.•500
, ...531
0-0'97'7
D-Otf27S
were compared to the "analytical" solutions of
10.m ID,:llO l0.271 1.a12 L7fft 0. lt7SA
Ref. 4. Solutions in Ref. 4 were obtained by the
' " 10-'
hlO" 10.n, o....70
,.10•
2• 101
10-7t3
ILMI
IL•7
ILl.00
IUN
ID.ffl
ILl2'
11.•1
ICl.292
11,002
IU2l
,.m•
1'2"5

t.144
O.H134
1.12113 numerical integration of an inversion ll!tegral. The
J,c 10• ILt.. IU43 11.t'I ILtlt 11.uo 4D1'1
I ,c 101 12.2'0 12.290 12.:IH 12.271 12.131 UU2 agreement between the analytical solutions and
2x 101 12.,» 11'3' 1:LI6S t.7261
Sx 101
IU»
11D,. 12.0,• 12.0,3
l:UlO
12.0fl 110U 12..., finite-difference solutions was excellent. The
I• 107 12.- 12.440 12.ot 11•:11 11Q 12.2.16
maximum differences between the two were about
2X 'I07
S• 107
hclO'
13.714
IUG
1. . .,
13.714
1.. ,.,
IUH
13.1N
14243
1Ult .....,
13.715
142G
1177t
14.2.00
14.SII
13.701
1... 21,
1U7•

TABLI!, - 1'o (r0 J VS •a POR •• +5. 'ID• 100 TAaL!t-,otto> VS 'o POR ·••20. "&D- 10
_£.:.!_ __jL_ __1_00_ _ .......!,!!!_ 10,DOO 100,DOD
,jz_ _!lz.._ 3::.!..... -'°- 100 _!,,!!!,_ ....3!!L 100,DOD
D,ODtt161 D.0009f7S7 0.0000ffl1' 0.DOooottt23 I 3.11., D-Ot74SS 0.DOttSAI D.DDOtt791 D.-tllll 0,0000llffl23
1 L227I O.Of3StS
D,UIMI o.01m1 D.001t9"5 O.DD01t9'3 O.DODO l"6S 2 4 IOSA D. lt2:II D.DltltD 0.001tfS7 0.0001"'• D.DOOOlfffS
2 1-'117
J 2.2321 o••,.,. o.o ..ou o.~25 O,D00'9t03 0.0000.tf 11 s s.,2., D...7IO o.o..su D.-171 O.DDOfflDI O.DDD04tf11
l ,t 101 7.7232 O.tOS.00 0,091123

..,...
O.ooott7ff 0,DOOOf9121 O.DD"722 0.DDOf9112 D.DDDOtte22
lx. lQI
2x 10 1
2.112'
2.31SS
0.77t73
l,ll67S
O.Of71N
D.1tl3S
O.DDHSSS
O.Olft7t ,.oo,,,.16 D.OOOltfU 2,c 10 1 t.S762 1.721t D.1t651 D.D1tt33 0.00ltff4 D.DDOltff.l
Sx I0 1 0.00..170
.__,
O,D00'9t07 Sx I0 1 12.212 3.15,111 o.«s:i. D.0'9771 0.00fflt7 D.OOOffllO

. ,..
2.S1l3 D."600 D.0..502

,__
D.Ott"56 D.DOtHU Ix IOI M.034 6.7171 O.t5-M2 O.Oft372 D.0Dtt777 O.DDOfflUI
h 102 4'2SS 1US3 D.17:IIA
2• 10 2 1Ln4
......
D,001HAD 10.tlS Ll»A D.19117 D.Olff50 D.001m3
2x102 S-42'1 .. ,us 1,Sttt D.lt502 D.01tf17
Sx ,oz 22.BA l7.7SS L330t o.•112 D.04tl27 D.oo.-.tCM
Sx ,:iZ '-5601 6.243 3.2300 o.am 0,0'9U7 D,OOna7
Ix 101
h
2x 'IOI
ID" 7.:DSI 7, 11IO
,...., 0.,112•
1.72113
0.0flffS
O.lffst
o.DOH7ll6
O.Dltf34 2>< IOI
22."72
2..0.00 %1.7:»
7.1271
13.D:13
o.n211
'·°""'"
D.1"7D
D.0Dft7H
D.OlffSS

,,..,..
Sx. ,:ia
7,tf3S
Lt3'2
t..lD7,
7-tOSI
l.t013
, ... 9.
1.ms
t,3"I
USOI
S.U2t
o.a200
O.t2t1S
o.oma
D.0"207
Sx 101
lxlD" ,._,..,
24.57, 24.514
24.tl7
20.U2
24.0:13
4.at3
Ll342
0 •..,:162
O.t772l
o.o .....
O.OfHOt

==~ 1.UIOI 1.7157 D.1'721 2xlo" 21.301 25.217 21.112 11Sl7 1.tl,S O. ltlll
1D.!60 l0.152 IO.D7S
ID,257 2.lt17 .... 20 Sx ID" 2s.7M 2s.7st 21.106 21.saa LSl.16 D....OOS
ID.733 I0.730 I0.7N
Ix 101 11,112 ILIII 11.100 10.tst ,.21.. 0.,502' 1 X 10'
2x101
216.113
216."60
216.110
216.4.ll
216.015
:16 ....
2S-IU
216.2166
1-2362
111..
o.ms,
I.tiff
2>< JOO 11.03 1L03 11...7 1L.I07 t.Ottt 1.1112
11,9:16 11.913 11.a• LOUI J,c 10• 21U17 21U1' 2un 26.157 22.275 LS172
SX 'IOI ILt:11 1Lt31 27.2164 27.2164 27.3'1 21.ns 216.lU l.lCl52
1,c 10• 12311 12.211 12.ll7 12.27S 12.127 '-'771 '" 10'
2x 27.610 27-'10 27- 27.S,S 27..001 14.0U
2x »• 12.US 12.,u 12.,,. 12-'21 12..164 t.7210
.Sx 101 21.0.7 21.0.7 21-067 2a.o.i 21.00• 21.tn
Sx IOI 13.DtJ 12.0f2 13-0tJ 13.0IO 1:1.0U ,2. ...
11'2.S 12.2.16 hcl07 2LA13 21.413 21.03 21.,10 21.3U 27,170
lx 11)7 1U3t IU2t 12.m 13.dl 2x 107 21.7Sf 21.7ff 21.7S, 41.7.ll 21.7... 41.534
2x 107
.sx 107
13.716
IUG
13.716
14243
12.714
.....,
IL20 ,..2.,
12.715 1.1.77t
IU40
14.SII
13.701
14213
1U7~
Sx I07
1" JO'
2'.21'
2t•.163
2'.216
2'..563
2'.216
2'.SU
2t.21'
2t•.163
2t.210
2'.560
2'.ISI
2'.532
Ix Ill" 14.51t ILSlt ILS19

TABLE 7 - ~.ot•z,) VS •.o POR ••+J. raa• 1.000 TAILI 10 -IIJJ: C•oJ VS •o POR ••+XI. ,.ID. 100

_!.i,._ C-o __!L-_j!;!__~ 10,000 100000

°f•O
~ 10 100 ~ ~ 100,DDO
I 1.0f2S o.oau D.DDHOU 0.000"7 .. 0.-IS D,DOOODffl23
D.DD1t9C O.D0Dlft62 0.00001tf'5 1 2.1'37 D.096302 D.DOH ..1 D.ooot9716 0-0000ffllt O.IIOODOftl23
2 Lc.!H D.17tSS O,D1t70
0.004N14 O.DD0'9t02 D.ooocuttll 2 2.1709 D.lltlt 0.0ltlSA D.DD1t9SA 0.0001tf63 D.DDD01"6S
s LNG D.•IS71 D-OaHS
D.DD009tl211 s L102A 0.4S500 D.11.,4:11 D.DD..... D.000""°' O.OODOfflll
1x10• :&.m• D.7U11 D.096150 O.DDHSII o.DDOft7'2
lx.101 S.:»71 D.16132 D.Ofl37t 0.0Dt9676 D-ODDHIOC D.OD0099122
2x 10 1 2."IS L30• D,M22 O.Olffl? O.DDltfSS O.DOD1"'6•
2• 10• ,s10s 1-'14S D,l,S17 O.Dlttlt D.001t960 O.DDOltfU
Sx 10 1 1"10 2.3174 0,"5S07 D.04'Ul D.00..... O.OOO,t907
J-, 10 I Ul7I :LOO o.m10 D.o.-.701 O.DD4tttD 0--
I>< ID'
J,c 10•
Sx 'I0 2
Ix 101
. ,...,
42213
.uo7,
uuo
UlOS
S.»71
'·°'"
O.IW2
i.w•
103"
Ut72
D.0912"5
O.lf.C21
0.011•
D.tom
0,00,9663
0.019'IO
D,Mff27
0.-11
O.DD09tl07
0.001"S,
0.00 ....3
0.0099721
l x 101
2x 101
sx 1a1
10.:IIIS
11.m
1L341
S.'514
I.IHI
12... ,
o.,3297
l.71U
4.DIU
D.Oft127
D.lt727
a.ans
D.oott7U
D.OJttG
D.04t7H
D,DDOf911'
O.DOl9H2
D.DD.mGO

,.....
'-l7'2 1.:x 101 1'.1.S, IS.1.53 UU4 D.tst.00 l.ottm D,DDtt712
2>< to>
,,. to>
,.nao '.7DII S.U23 urn 0.1- O.Oltf.l
2• JOI 17.tl2 17,"53 II.US 1.162' D.1tt2• 1.01nso
7.5.162 7.52'3 7.:nos O.Ot73 D.0"712
Sx 101 20.•10 20.110 16.NO U12l 0.Af1'2 0.04tl3'
2".,.
1><111"
S><IO-'
LIS23
1.1•
1.1:IID
1.7403
t.5321
7,tt3t
IM71
,.so12
S.37'1
7,2HJ
f.D704
D.'2t03
1.7S17
3.7500
D.OftDt7
D,1ff91
o.••"
1x 10"
2x 10 4
22.2"
21'39
22.12•
Zl.113
20.620
n.2a
7.72SS
12.797
D.t71.f6
1.tODJ
0-0tfUI
0.1... ,
t..53SA
,.,275 ,.o,u o.,a, Sic 10 4 21.00S 213'3 Js.260 211.611 LOD7 O.ff352
"'ID" 10.133 10.131 10.115 1 ,c 10•
,....,
111.. .- 1,5293 L7tl9 25.tll Z.t77 '25.f,O 24.672 L1511 ~-'7711
10.729 ID.741 10.720
.,,.105
' " 10'
Ix 101
11.SOt
12.D.ll
11.sot
12.DSA
II.S06
12.DSA
11... ,
13.039
1D.167
IU22 ,....,
,.,.2 2x 10•
Sx 10 1
1,c 10•
216 •.001
216.ltS
216 •.000
216.atA
216.31S
27.lSO
:».ISO
».a21
27,223
13.720
22.112
216.130
1.t174
•.5311
L2t"5
2x101 12.5216 12.5216 12.52' 12.SII l2.43S 9.5412
2x 10 1
27.2.53 27.2.53
10' 12.:161 27-'M 27-'M 27.603 27.590 27.390 IL02t
Sx
Ix 107
2x 107
13.DSA
11a1
11054
11a1
13.053
IU20
11051
1Ult
12.023
13..IOA 13.22J
,~ 10•
1x 10 7
21.D6S
:a.'12
:a.ou
21-412
21.DU
.... ,2 ,._..,,
.I.Ost 21.002
21.312
22.tll
27.IU

,.....
13.777 13.ff7 l3.ff7 1111, l2.7U 12.6H
J,c 107 14:nt 1.. 2:16 142119 2x 10 7 ;l.7S, :a.7s, 41.7~ 21.7S1 21.1 .. :a.,31
14.Z:Jt 14.:nt IL:nt Ix I0 7 2t.210
hclD' U.!11 1.C..517 lA..572 2'.2M 2'.216 2t.2M 2'.21' 2t.lSI
14511 1LSII 1x10• 2t..S'3 2'.563 2t..S'3 2'.562 2'.560 2'.531

TAil.i i _,,,, ''o' vs 'o POR •• +20. ,.ID. 1 TABLE 11-,:, C,_o) VS t0 POlt ••+20. •io• 1.000
...!e_ ..!:!... 10 IOD
~ ~ 100,DOD
....'.R,_ C•O ID 100 l,DOD 10,00D IDD,DDD
I 20.,.. D.09'551 0.009979' O.ocot911t
D.DD1fff3
D.DD009tllS
D.ODDlfffl
D.DOODOff7U
O.DODOltta
1
2
1.7117 0.0,S.534 0.DOH37t 0.DOOff771 D.DDOOt911' 0.-23
2 20.Ht 0.1NS D.01t9U 2.27D7 D.117°' D.DIN32 D.001t9SI D.DD01tt63 0.DDDOltffS
s 2U1' O.ff3'6 D-04NSD D.ODfflO• O.DOO<ttOS D.0000'9170 s 2.J.!13 o .... JJ D,0'9337 D.DOfflS3 D.0004tt0S D.00004ttll

. .,.
1 • I0 1 21-'0A o.t1..1 D.OftSU O.OOH797 0.DOOfflOI D.DDDOH7l6 Ix 10 1 L270A D.&422 D-09'0.1 D . -. . D.OOOfflOS 0,oooot9122
2,c 10 1 21.tM uoa D.1N7D D.DltfSS D.DOlfff2 O.OOOltfU 2,c 10 1 S-3044 1.542' D.1.. 15 D.OltfOI D.001ffst D.0001,,..
J,c JOI 22.3-16 0,'93» o.oms, O.DO<ttO• D.DD0491t9 Sx IO' ,.1211 :t.170 0,0117 o.o...., O.DO•'tUS D.O<IG-
J ,c: 101 22.UI 1,0074 O.t7.161 O.OftSt3 O.D09t7tl 0,DOOfN07 Ix 10' 1.DSIO S.0234 0.,1120 O.D91HI 0,DOft7J6 D.DDOffl1'
2K 10 2 23.021 11163 l.tOH D.19174 0.DlttSS D.001fff2 2x 101 9.l211S 7.21'1 l.7lSI D.1H77 O.D1H3S D.DDltfU
._as, SX )OZ
Sx JOI
hclO'
22.0• 20.273 D.49360
0.97669 ._.,_.
D.0....16 D.004990•
D.DOH7tf hC"lOS
11-0.16
12.:114
I0.041
11..1311
3.7tl0 0 ..... 2 O.Mt7U DJI0.4tlff
2x IOI
Sx IOI
%I.Ill
2LM3
2U20
211211
ZLDOD
24..165
1.0199
114111
20."'
1.t127
•.•m
O.IN7t
D.'9:115
O.Oltt.16
D.DfflSA
h!O'
, . 10•
12.7211
IJ.ff3
13.431
IS.».
f,3307
t •.1614
11617
'·'•"
1.12SS
41312
O.OfflDI
0.1t713
0.4t3S
O.Offt771
o.o,,, ..
0.0'9112
I>< 10'
,,. .,.
2><.,.
2LHf
21.311
Js.761
24.Ut
Js.291
2S-7U
2.. , ..
Js.122
2'-711
'·""
13-'lO
21.621
o.m'7
Lt1'•
4517'
D.Dfffl•
0.1'tU2
D...Q
lxlD"
2>< lo"
5><10"
1'.a:11
11.IM
It.NJ
M.ffl
11.1-0
If.ta
lf.000
.,.,,.
17.771
7,1761
IL340
17-031
O.tUlO
1.1713
L:12161
0.Dtf .. S
D.19134
D.fflll
Ix IDS 2',115
,.,..,,
:16.112 ».oar :ZS.171 L2lff D.t7162 l,c I01 21.313 21.JOS 21.227 211.100 7.72U O.t71S2
2.x IOI
Sx IO•
:16... 1
2'.917 36,tl7
2'."7
••,12
2'.2169
».a57
11154
22.27t
l.t211D
'-m•
2>< to'
Sx 101
22-"D
2•~
22-'SS 22-'lS
2Ult
22.141 IUS7
.,.,
...1.,,..
,-.111
2)1.10•
27.:aM
27.IIO
27.:aM
21.,10
37,2'1
27.601
27.2:IS
27..Sff
2'.11S
27.401
1.lOSS
14.0SS
he 10'
2>< JO'
2S.7U
216.tol
2.1.437
25.7M
:16.IOI
2s.7SS
216.'°4
2L231
25.664
26.167
"·"'
2LO"
216.33'
1.0925
1U33
hlO'I 21.067 41.0.7 2&0.7 .I.DAI ll.llCM 22.t22 27.1ft 27.Nt 37.161 27-"G 27.7'1 22.:111
I• 107
2x I07
,~ ,07
Ix 10'
21.413
:a.75'
2t.2M
2U13
:a.75'
2'.21'
2U13
;I.1st
.2t.21'
2',.163
21.41D
3,7SI
2',21'
2t.563
:a. • 3
21.7,S
21.210
2t•.16D
27.171
21..534
2t.lS1
2t.sn
'tx
" 107
100
2,c 107
Sx 107
h IOI
:a.3211
:a.121
2'.202
41.220
21.721
2t.l02
;1.3311
21.721
2'.2112
21.3216
41.71'
2'.2111
21.ffl
...,os
2t.1ts
......
26 ....
2',134
2t•.163 2'.SU 2'.5.16 2'.S.16 2'.556 2t.SSS 2'..553 2'.524
 10 the radius of the damaged region upon short-time
well-test data. Results for a skin effect of +5 were
similar.

DISCUSSION
As shown in Fig. 2, by comparison of results
for different r10 values, representation of a skin
effect as an annular region of altered permeability
can lead to significant changes in the early
ANAl.'TflCAl. - - pressure-time history, ll;ith or without wellborc
,uuT& Dl"IIIINCI • •
storage. (All lines on Fig. 2 for r1D = 1 represent
10• 101
infinitesimally thin skin cases, or previous
'• analytical cases.) The lines for the finite-radius,
FIG. 1 - PD (ID) VS ID FOR ANALYTICAL AND
damaged regions fall successively below the
FINITE-DIFFERENCE SOLUTIONS (s= 0). infinitesimally thin skin case, but finally join the
infinitesimally thin skin case at times which
0.5 percent, with the analytical solutions being increase as the damage radius squared, as would
higher than the finite-difference solutions. This be expected. If there is significant wellbore
maximum difference occurred for the larger values storage, a separation is noticed within the
of C at the transition between the wcllborc transition region from storage control to outer
storage-controlled period and the period in which formation control, as damaged-zone radius increases.
wcllbore storage was not important. On a log-log The separation diminishes with increasing wellborc
plot of p0 (t0 ) vs to, this is the region of greatest storage constant and is essentially negligible for
curvature. Ac other points on the curve the solutions storage constant of 100,000. This results because
are almost identical. the finite-storage constant cases must join the
Fig. 1 shows a comparison of the analytical zero storage case. Thus previous criteria for the
solutions from Ref. 4 (solid lines) and duration of the wellbore storage effect4 requires
finite-difference solutions (points). The differences some modification. If the damaged-zone radius is
between the two arc not noticeable on a graph. All great, transients caused by the large volume of the
of the curves in Fig. 1 are for no skin effect, s = 0. damaged region may last longer than those caused
Fig. 2 presents results for skin effects of -5, 0 by wellbore storage. In extreme cases, this could
and +20. The skin effect of +5 was not plotted to result in an early period caused by wellbore storage,
make Fig. 2 more readable. Dimensionless pressures followed by two straight lines on a conventional
at the well are plotted with skin effect, semilog well-test plot - the first having a slope
dimensionless radius of the damaged zone, and the indicative of the permeability of the damaged region;
dimensionless wellbore storage constant as the second having the correct slope indicative of
parameters. The results for the skin effect of -5 the undamaged formation. This is shown on Fig. 3.
are only for a dimensionless "damaged" zone Fig. 3 presents results in a conventional semilog
radius of 1,000. le was impossible to achieve a plot of the pressure-time daca for a skin effect of
negative skin effect as large as -5 with r1D values +5, a storage constant of 1,000, and two different
less than 150. Thus, the data for a skin effect of damaged-zone radii: r 10 of 1 and 1,000. Pressure
+20 on Fig. 2 arc best for studying the effect of data co a time of about 2.5 x 10 3 a.re almost

,oz...------r-----.------r-----~----,-----,------,-

FIG, 2 - SHORT-TIME SOLUTIONS WITH WELLBORE STORAGE AND FORMATION DAMAGE.

St:PTt:!IUH:R, 1 •JTII
 completely dominated by the wellbore storage CONCLUSION
effect. This can be seen better on a log-log plot;
the data form near perfect straight lines with a unit The infinitesimally thin skin concept of
slope on a log-log plot. .At times of about 2 x 10s, van Everdingen and Hurst is applicable if wellbore
both cases reach a straight line. For the storage is significant for damaged-zone
infinitesimally thin skin case, r 10 = 1, the slope dimensionless radii from I to about 100. If the
is the correct value of 1.151 and is indicative of damaged-zone radius is as large as l,000rw, two
the formation permeability. For the large-radius straight lines will be evident on a semi-logarithmic
damaged region case, the slope is 1.984, which is plot. If the damaged-zone radius is small, it is not
indi~ative of the damaged zone permeability (see possible to detect the value of the radius from
Table l). .At a time of about 4 x 106, the well-test data.
large-radius damaged region case finally reaches Uthe damaged-zone radius is greater than l00rw,
the proper sttaight line • .As can be seen on Fig. 3, short-time well-test data should be inte.rpreted by
the proper straight line for the r10 = 1,000 case means of solutions for a skin region of finite
could be easily misinte.rpreted. It is far easier to storage capacity. One way to accomplish this end
inte.rpret this case with a log-log plot and type-curve is a type-curve matching procedure employing
matching procedures, 4 and misinterpretation is far plots such as Fig. 2.
less likely. Type-curve inte.rpretation procedures If wellbore storage is not significant, a
are discussed in detail in Ref. 7. damaged- or skin-region radius of l0rw or greater
Several comments regarding well-test may distort the early pressure data significantly.
inte.rpretation may be made utilizing Fig. 2. First, In this event, two straight lines may appear upon
storage constants for oil well tests often are of the semi-logarithmic plots.
order of 1,000 or greater. Inspection of Fig. 2
indicates that the radius of the damaged region NOMENCLATURE
would have to be greater than l00rw to cause a
significant effect on the pressure-time history. c = total system compressibility, atm-1
C = wellbore storage capacity, cc/atm
Thus~ previous interpretation methods (such as
L dimensionless wellbore storage con-
described in Refs. 4 and 7) should be valid. In
stant, Eq. 14
other words, the infinitesimal skin concept is valid
h = formation thickness, cm
under these conditions. Second, storage constants
k = undamaged formation permeability,
for gas well tests may be much less than 1,000.
dardes
Inspection of Fig. 2 indicates that the radius of
k1 = permeability of "damaged'' region,
the damaged region may play an important role in
darcies
pressure-time data for these conditions. Thus Fig. 2
p = pressure, atm
represents an additional type-curve which may be
particularly useful for interpretation of well-test Pi = initial pressure, atm
data when storage constants are small. In our PID = dimensionless pressure in "damaged"
opinion, type-curve matching procedures are region, Eq. 11
extremely useful and will find increasing application P2D = dimensionless pressure in "undamaged"
in well-test analysis. Finally, it is apparent chat reservoir, Eq. 12
Fig. 2 should also be useful in forecasting results PwD (to) = dimensionless pressure in wellbore
of production from a composite reservoir by proper q = surface flow rate, cc/second
inte.rpretation of the skin-effect parameter. r = radial distance, cm
r1 = radius of "damaged" region, cm
rw = wellbore radius, cm
'D = dimensionless radius, Eq. 9
r1D = dimensionless radius of "damaged"
region: r 1/r w
IZ s = dimensionless skin effect
t = time, seconds
10
t 0 = dimensionless time, Eq. 10

-·•
~
8
CONOITIOIIJ
tf, = porosity, fraction
p. = viscosity, cp

• ••I .ACKNOWLEDGMENT
4
'Z"II; /ta: a.11
Ta ..NSITIOII ••o•
- STOUH·~~ The computer time required for this study was
TD-STUtCMT•l.1111• J provided by Stanford U. Financial support '\llr·as
received from Stanford and the National Science
Foundation. The authors gratefully acknowledge
10•
'• this assistance.
FIG. 3 - EXAMPLE OF B'VILDUP DISTORTION WITH
WELLBORE STORAGE AND FORMATION DAMAGE.
 REFERENCES ln Unsteady Liquid Flow~ I. Analytical Treatment",
Soc. Pet. Eng. J. (Sept., 1970) 279-290.
1. van Everdingen, A. F.: "The Sktn Effect and Its In- S. Wattenbarger, Robert A. and Ramey, H. J., Jr. "Gas
fluence on the Productive Capacity or a Well", Well Testing with Turbulence, Damage, and Wellbore
Trans., AIME (1953) Vol. 198, 171-176. Storage", J. Pet, Tech. (Aug., 1968) 877-887.
2. Hurst, W.: "Establishment of the Skin Effect and Its 6. Wattenbarger, Robert A.: "Effects of Turbulence, Well-
Impediment to Fluld-Flow into a Well Bore", Pet. Eng. bore Damage, We!lbore Storage, and Vertical Fractures
(Oct., 1953) Vol. 25, B-6. on Gas Well Testing", Ph D thesis, StanCord U., Stan-
3, Ramey, H. J., Jr.: "Non-Darcy Flow and Wellbore ford, Calif, ( 1967).
Storage ECCects ln Pressure Build-Up and Drawdown or 7. Ramey, H. J., Jr.: "Short-Time Well-Test Data Inter-
Gas Wells", J. Pet. Tech. (Feb., 1965) 223-233. pretation in Presence of Skin Ef!ect and Wetlbore
4. Agarwal, R. G., A1-Hussainy, R. and Ramey, H.J., Jr.: Storage", J. Pet. Tech. (Jan., 1970) 97-104.
"An Investigation of Wellbore Storage and Skin ECCect
***

SEPTEMBER, 1910 297