You are on page 1of 8

b

SPE
=-m4M

~n

WIN=S

SPE 23440
Estimating the Stabilized Deliverability of a Gas Well Using the
Rawlins and Schellhardt Method: An Analytical Approach
J.L. Johnston, W.J. Lee, and T.A, Blasingame, Texas A&M U.
SPE Members

>opyright 1SS1, Society of Petroleum Engineers,


rhis paper was prepared for presentation

Inc.

at the SPE Eae.lern Reglonel Meeting heldin Lexinotm,Kentucky,October22-25.1991

rhls paper waa selected for presentnllon by en SPE Progrem Committee following review of information contained in an abslract submittad by the author(s). Contents of the paper.
M presented, heve not bwn reviewed by the society of Petroleum Engineere end are sublwt to correction
W theauthor(s).
Thematerial,
asPresented,
doesnotnece$sarlfY refle~~
my posltlon of the society of Petroleum Engineers, ile offlcera, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
M Petroleum Engineers. Permieaion to copy is reetrlctsd to an abstract of not more than 2M wwds. Illustrations may not bo copied The abstract ehould contain conspicuous acknowledgment
M where and by whom the paper Is presented. Write Librarian, SPE, P.O. Box 833S3S, Richardson, TX 7S0S3.3836 U.S,A. Telex, 73C389 SPEDAL.

ABSTRACT
.,
This paper introduces a direct method to use the results of
Houpeuztdeliverability analysis to derive the constants C and
n in the Rawlins and Schellhardt gas well deiiversbllity
equation. The motivation for this effort is the need to report the
resultsof Rawlins and ScheUhardtanalysisto regulatory agencies,
and the widespread use of their deliverabi lity equation by
engineers. We present a detailed procedure which shows how
these results can be applied to deliverability forecasting. This
paper includes an ?lustmtive example in which the new nwthod is
applied to field data from the literature. This example presents
comparisons between Houpeurt and Rawlins and Schellhsrdt
analysesand shows the correlation betweenthe two methods.
INTRODUCTION
The purpose of deliverability testing is to determine a gas wells
production capabilities underspecific reservoir conditions. A
common prmiuctivity indicator obtained from these tests is the
absolute open flow (AOF) potential, which is defined as the
maximum rate at which a well could flow against a theoretical
atmospheric backpressure at the sandface. Although in practice
the well cannot produce at this rate, the AOF is often used by
regulatory agencies for establis%tg field proration schedules and
setting maximum allowableproductionrates for individualwells.
A number of testing techniques have been developed to assess a
gas wellsdeliverability characteristics, Flow+fter-flowl tests am
conducted by producin$ the well at a series of different flow rates
and measuring the stabdized bottomholeflowin$ pressures, Each
flow rate is establishedin successionwithoutan intermediateshutin period. The primary limitation of these tests is the long time
required to reach stabilization in low permeability reservoirs.
Consequently, the isochrorta12and modified isochronal tests
were developed to shotten test times.
An isochronal testis conducted by alternativelyproducing the
well, then shutting it in md allowing it to build up to the average
reservoir ressure prior to the beginning of the next flow period.
The rnodiRed isochronal test is conducted similsriy, except the
Referencesm~Uustradons at end of paper

duration of the shut-in tiis often is not long enough to reach the
true average reservoir pressure in the well% drainage area.
Although isochronaland modifkd isochronaltests were develo
to circumvent the long flow times required in low permesblPity
reservoirs, these tests may still require a single, stabili=d fiOW
period at the end of the test in order to estimate the srrzfdfized
producingcapacityof the well.
The conventional deliverability test analysis technique was
proposed by Rawlins and Schellhardt.} They observed that a loglog plot of the difference between the squares of the average
reservoir pressure and the bottomhole flowing pressure against
gas flow rate can be representedby a straightline defined by
qg = c G* P;f)n . . .. . . .. . . .. . . . . , ..,.,,,,....,,, . . . . . . . . .. . . (1)
where C is defined as the stabdizai performancecoefficient,and n
is the reciprocal of the slope of the straight line, Extrapolation of
this line to the difference between the squares of ths average
reservoir pressure and the bottomhole flowing pressure equal to
atmosphericpressuredefines the AOF.
Eq, 1 was developed empirically from the observation of a
number of gas well tests. Extrapolation of Eq. 1 over large
variations in pmssum can result in incorrectestimatesof the AOF,
Subsequent theoretical developments by Houpeurd have shown
that a more accurateanalysisfor gas flow is possiblewith
,,, ., ... ......... .... ..... ... ...,, .,4,.. (2)
P2-p~f=aq8+bqi
where the flow coefficients,a and b, are defined by
a= L422x106~ Z T
k,ll g [151~g(%)-i+]

.b.,(3,

~ = L422x106 jit Et TD
kt h

.,, ,!,.,.,.,,, ,,, ,,, ,,, ,,, ,, ., .,... (4)

2 is a solution to the diffusivity equation for radial flow,


7A ;hough t$e Houpeurt equation has a theoretical basis and is
rigorously correct, the more famiiiw but em ifically b~ed
Rawlins and Schellhardt equation continues to L used, inkd
favored, b I the natural gas industry. Consequently, we have
combined i e two analysistechn4uea and havedevelopeda mm

Estimatingthe Stabtized Deliverabilityof a Gss WellUsing the


Rawtinsand SchellhardtMethod: An AnalyticalAppioach

~=[1
~g+%l

accurate version of the Rswlins-Schellhardtmethod which can be


used in deliverability forecasting. Our technique can be used to
estimate the stabihzed performance coefficient, C, without
requiring stabilizedflowing conditions and is especially useful for
analyzing isochronal and modified isochronal tests without a
stabilized flow period. This is a simple method which rquires
only data from a modified isochronal test to develop a
performance prediction for the well without a priori estimates of
reservoir properties.
I?ZFINITIONS AND THEORETICAL

P&
pb /404 go)

c(a~,~b
qf~

DEVELOPMENT

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

In terms of pseudopressure,the Rawlins and Schellhardtequation


becomes
qg = C ~P@) - PP@wf)l ,,, ., .,,.,,... %. . . . . .. . . .. . . . .. . . .. . .. (6)

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

In
. . . . . . . . .. . . . . ,...,,...,,,

= (a f~qg)

(15)

Eq. 15 is similar to a result derived by Poettmann and Kazemi.T


We show the applications and importance of this development in
the procedure in the next section of this paper, The flow rate
required in Eq, 15 is defined by solving Eq. 14 for the gas
flowrate, qg,
a(l -n)

.. ... ...... .... ........ .... . .. . ... ...... ... ... (16)

g= b(2n-1)
Implicit in our derivation is the assumption of radial flow of a
single-phase gas in a homogeneous, isotropic reservoir, For
naturally fractured reservoirs, our method, like conventional
deliverability analysis techniques, is valid only after the matrixfracture system has begun to behave like a single, homogeneous
unit. Similarly, our method is valid only after pseudoradial flow
is exhibited in hydraulically fractured wells, We also assume
wellbore storageeffects are negligible.

.. . .. . .. . . .. . .. (5)

Our method is a!su applicable, however, for deliverability


quations written with pnssure-squared as the dependent variable.

a+2bqg

To develop an expression for the performance coefficient, C, in


terms of the Houpeurt flow coefficients, we combine Eqs, 10 and
12 to obtain

-a+bg

qg a+2bqg

Our deliverability test analysis technique is derived by equating


derivatives of log (qg) with respect to log (AP ) from both the
Rawlins-Schellhardt and Houpeurt equations, 1 similar method
was used by Brigham,s Duong,s and Poettmann and KazemiTto
develop quations in terms of pressures-squared for estimating
reservoir properties from gas deliverability tests. Because of the
pressure-dependentgas properties, the pressure-squared forms of
the deliverability equations (Eqs, 1 and 2) often are inaccurate at
high pressures. Therefore, in the subsquent derivation, we use
the pseudopressure transformation introduced by A1-Hussainy,et
al,a
pp=2

SPE 23440

DELIVERABILITY TEST ANALYSIS


I Application
of our method assumes that the slo~,

l/n, of the
empirical deliverability plot remains constant wnh time. This
assumption implies that, if we can calculate values of a and b
(I@. 8 and 9, respectively) for given reservoir properties, we
also can calculate a flow rate with@. 16, We then substitute this
flow rate into Eq. 15 and calculate a stabilized C value, and
assuming a constantvalue for n, calculatethe AOF
AOF = C ~p(ji) - Pp@b)r . . . . . . . . . . .. . . . .. . . . .. . . . . . .. . .. (17)

Similarly,the Houpeurtquation is
Pp@) - PP(IW) = a qg + b q~ . .... .. .... ... ... ... .... .. ... ...(7)

where the flow coefficients,u and b, are defined by

To apply our new deliverability analysis technique to field data,


we present an analysis procedure below. We will then apply this
procedureto a field example,
6

b = ~x10

kg h

, .,, ., ...,..,, . .. . . . . . . . . .. . ,.! ...,,,0,,,,

General Analysis Procedure. We recommend the following


proceduro to analyze isochronal and modified isochronal tests
using our technique,
Although presented in terms of
pseudopressures, this procedure also is applicable with the
pressure-squaredvariables,

(9)

Taking the logarithm of both sidesof Eq, 6 yields


/og(q8) =/og(C)+ fl/og~~@ -PP@w/)].. ...0.. . .. ....(lO)
Reatmnging Eq. 10 and solving for n shows that n is the slope of
a log-log plot of qg vs. *P Alternatively, n can be expressed as
the derivativeof log (q8) with respect to log (A@:

1.

Plot 4P = Pp@) - Pp@wf) vs. qg on log-log graph paper for


the measured flow data.

2, For each flow time, construct the best tit line through the
data points, Typically, some of the earl data points will not
agree with the general trend of the Jata, so these points
should be ignored in all subsequentanalyses.

=m&3i$m=hm#kz-T(
ll)
Similarly, taking the logaxithmof both sides of E@7 yields

3. Determine the deliverability exponent, n, for each best-fit


line by least-squaresregression analysis using the following
equation:

OtiP@-pp@~j)] =10~aq8+ bql]... !...!..! . .. ..!. (12)


Differentiatinglog (@P) with respect to q, gives
I

N$(fOgq810g&p~-,

a+2bq

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

(13)

aqg+bq~

Substituting F@ 13 into Eq 11 yields an quation for n in terms


of the Houpeurt flow eoefftciem
L

!i

,
1OW.5(10,4P)J

=-$=(8)

4, For M isochronal tests, compute the arithmeti~ average


deliverabilityexportcn4ii,

SPZ23440

J,L. Johnston, W.J, Lee, and T.A. Blasingame

-. . .

E=

i= 1

2 . F

Ili

. .. . . .. .. .. . . .. . . .. . . . . . . .. . . .. . . . . . . . . . . .. . . .. . .. . . . . (19)
TF2580F,

5. Calculatethe theoreticalvalues of the Houpeurt coefficients,


a and b (Eqs. 8 and 9, respectively), using values of
permeability, skin, and the non-Darcy flow coefficient
obtained from buildup or drawdown tests on the well, or in
absense of these tests, from the rnddified isochronal test
analysis methcds presentedby Brar and Aziz.g

p, psia
14,7
225.0
450.0
675.0
.900.0
1125.0
1350.0
1575.0
1800,0
2025,0
2250.0
2475.0
2700.0
2925.0
3150.0
3375.0
3600,0
3825.0
4050.0
4275.0
45C0.O
4725.0

6. Calculate the .ratc (Eq, 16) at which the change in


pseudopressure, App, determined using the RawlinsSchellhardt equation equals the value determined using the
Houpeurt quation. For this calculation, use the average
deliverability exponent, E, value from Step 4 and u and b
from Step 5.
7. Calculate the stabilized C value using RI. 15 and the result
of Step 6 (Eq. 16).
8, Using E from Step 4 and C from Step 7, c~culate the AOF
with-%q,17.
ield Example. This field example is a {modified isochronal
st from Well 8, analyzed by Brar and Aziz9. The well and test
ata are given below in Tables 1-5, To compare our me:hod with
wand Aziz, we do not use the final, extended flow point in our
nalysis.
.,
he permeability and skin factor given below were obtained fmm
pressure drawdown test, but could have been obtained usin
nly the tndfied isochronal test data, as Brar and Aziz showJ .
crmeabllity and skin factor are not necessary if Houpeurt
nalyaisia performed, since the values of a and b can be obtained
ic@y from the test daw rather than from theory, Eqs. 3 and 4.
hese reservoir properties obtained ftom the drawdown test are
sed only for comparison in this example.

k~

ft

0.5

1,0
;:;

q,=31 .612
3726
4;274; 1
4,266.8
4,262,5
4,258.3
p~,

q,=

1.0
;:;

#= 0.0675

z~ = 1.69x

T= 718 R (258%)
~ = 4,372,6 psia
j = 0,023 Cp
Z = 0.87

I/w;o:
640

10~ psia-l

56.287

4;111:8
4,103.3
4,097.0
4,093.5

0,5

~52; md

PDI Psia2/cp
0.20038 X ld
0.34636X 107
0.13901 x l@
0,31254X lC$
0.55399 x lIYJ
0.86142 X l@
0,12323X 109
0.16635x 109
0,21515 X 109
0.26924 x 109
0.32820X 109
0.39163X 109
0.45909x 109
0.53018X 109
0.60451x 109
0.68171 x 109
0.76143 X 109
0.84338X 109
0.92725 X 109
0.10128 x 1010
0.10998X 101
0.11881 x 1010

Test ~
p~,

s.

A=

0.98716
0.97537
0.96476
0.95545
0.94756
0.94116
0.93634
0.93314
0.93156
0,93159
0.93319
0.93628
0.94078
0.94659
0.95361
0.%174
0.97087
0.98090
0.99175
1.00334
1.01557

cs, P&-1
0.68318X 101
0,45000x 102
0.22733X KV2
0,15274X 102
0.11512X 10-2
0.92246X 103
0,76726X 103
0.65398X 103
0.56694x 103
0.49749x 103
0.44052X 103
0.39282X 103
0.35228X 103
0.31744x 103
0.28725X 103
0.26092X 10.3
0,23783X 103
0.21750X 103
0,19953 x 103
0,18359X 103
0.16941X t&3
0,15675X 103

k?, CP
0oO14635
0.014739
0.014918
0.015145
0.015414
0.015720
0.016062
0.016436
0.016841
0.0!7273
0.017730
0.018208
0.018705
0,019217
0.019742
0.020277
0.020819
0.021366
0.021917
0.022469
0,023021
0.023572

3 .

able6 gives a comparison of our results the results we obtained


sin$ the extended flow point. We do not show the conventional
mchfied isochronal test analysis with the extended flow point
ere.
.
vom ~
Well &
h = 454 ft
rw = 0,2615

:-

099913

fi=O.65

psia
44,313
t.
4,206.5
4,199.3
4,192.0
4,190.4

q,=

psia
q, = 70,265*
4,3436
3,994:6
3,973.3
3,959.6
3,951.9

*(all ratesare in MMwf/D)


Ie 4 .

acres
P (Pw$s

CA= 30.8828
(assumewell is
centeredin a
4usre drainage
area)

XilI&lK
o IPD(PWJI
3.04,0
5.0
6.0

q,=31 .t%2
1.049 x 109
1,011 x 109
1.008 X 109
1,006 x 109
1,005 x 109
PpfPw+t

TimA.11
o hMPw.Jl
3.04.0
5,0

6.0
*(au rates are inMMsm)
255

q, = 56.287
1,038 X 109
9.488 X 1($
9.456
9,433

X 1($

x IN
9.419 x 108

psia2/cp
q,=44!313
1,042 x 109
9.848 X l@
9,821 X 108
9,793 x lp
9,787 X 1~

psia2/cp
q@= 70.265*
1.037 x 109
9.047 x 108
8.96q X lt$
8.916 u 108
8.888 x 1(N

SPE 23440
.

Estimadngthe StabilimdDeliverabilityof a Gas Well Using the


Rawlittsand SchellhardtMethti: An AnalyticalApproach

+.

. .

.-

. .. . . ...

3,794.0 psia
~ = 4,372.6 psia

p~=

P~P~} = ai~:;:~

@a21w

qg = 77.%5 Mh4scfm
PP(14.65) = 2003.8 psia2/cp

For t=3.Ohr

In this example, we will f~st illustrate the Brar and Aziz analysis
technique; then we will use values of u and b from that method to
illustrateour new procedure, the stabilized C method.
Brar ~

1.

~,ot*.

point

qg

qg 2

31.612
56.287
70,265

999
3,168
4,937

3
4

versus flow rate, q, on


Cartesi% coordinateq~aperfor the measured isochronalflow

1,202 x 106
1,585 X 106
1.883 X 106

XiIj@#t
.
4.0
5,0

6.0
..- ...

q.
1.2;2 x 106
1.297 X 106
1.360 x 106
1.392 X 106

.
4.0
5.0

1.642 x 106
1.682 x 106
1.707 x 106

6.0

4, 44,313
1.2;1 x 106
1,352 X 106
1.415 x 106
1.428 X 106

103)-(158,164~

ld psia2/(cph4M~f/D2)

For t=4,0 hr

f7R 70.265
1.8i3 X 106
1.995 x id
2.069 X 106
2.121 x 106

q)

point

qg

q82

1
3
4

31612
56:287
70.265
158.164

999
3;168
4,937
* 04

4.100 x
9.240 X
1.402 x
2.736 X

107
107
10fJ
1~

1.297 X
1.642 x
1,995 x
4,934 x

106
106
106
106

3 (2.736X l@) - (158.164)(4,934X ld)


52=

3(9.1Ch4X 103)-(158,164~

~= 1.760x 104psia2/(cpMMscf@)

The deliverability Dlot is shown below in I%?.1. We now


construct best fi~]ines through the modifl-d isochronal
data points for each time. l%e data points at the second
rate (44.313 MMscf/D)do not fit on the same straight line
as the other three points. These points at the second rate
are therefore ignored in the subsequentcalculations.

For t=5.O hr

I
t, hr

2.2 J

3(9,104

bl= 1.737x

~dq~,
psia2/(cp-~scf/D)
f?,

3 (2,595X I@). (158.164)(4.670X 106)

1-

flpd% psia21(cp-MMscfAX

2.4x10C

qx

3.800 X 107
8.920 X 107
1.323 X l@

Pp (PWs)
Pp (.PWjj

data. This plotting function is calculated below for each


rate.

2.

@
4P

point

qg

qg2

1
3
4

31612
56:287
70.265
158,164

999
3,168
4,937
* 04

6.0

3(2,831

::;
3.0

b3 =
b3=

&p

4,300 x
9.470 x
1.454 x
2.831 X

qx

107
107
lN
I@

1.360 x 106
1.682 X 106
2.069 X 106
5,11IX 106

l@-(158.164)(5.111

3(9.104

106)

103)-(158,M4~

1.782x 104psia2/(cp-MMscf~2)

For r=6,0 hr

o~
m

30

40

50

60

70

point

qg

%2

:
4

31612
56:287
70,265
158.164

999
3,168
4,937
$ 04

q, MMSWD

&
Am

4,400 x
9.610 X
1+490x
2,891 X

qR

107
107
I(P
lN

1,392 X
1,707 x
2,121 x
5,220 X

b4 = 3 (2.891 X 1($) - (158,164)(5.220 X l@


3,

3(9.104 x lo3)-(158.164p

Determine the slopesof the lines, b, for each time by leastsquares regression analysis on the fmh third and fourth
points using Eq. 20,

b4=

220

1,815x K)tpsia2/(cpMMscf/D2)

106
106
106
106

w%23440

J.L. Johnston, W.J. Lee, and T,A, Blasingame

Determine the average value for the slope by taking an


arithmeticavemgeof the b values calculated at each time.

4.

b=

For t=6.Ohr

1.774x 1($ psia%@bfMsc4D2)

&

point

qg

qg2

1
3
4

31612
56:287
70.265
158.164

999
3,168
4,937
9,104

Calculate the transient deliverability line intercepts, at, for


each isochronal line using 13q,21.

5.

~,4_(5.220

a(4 =

For t=3.Ohr

4P

qR

107
9.610 X 107
1.490 x 10
2.891 X 10
4A00 x

1,392 X I@
1.707 x 106
2,121 x 106
5,220 X 106

106)(9.104 X 103)-(158.164)(2.891X 108)


,
a,
3(9.104 x lo5)-(158.164f

7.829 x 105psia2/(cp-MMscf~)

Plot a graph of at versus log r. Draw the best-fit line


through the data. Note that all of the data lie on the same
strakht line, therefore all points should be used in the linear
re~-ssion tialysis,
8X105.

70.265
158.164

4,937
P 04

1,323 X 10
2.595 X l@

1.883 X 106
4.670 X 106
7.

all =(4.670x 106)(9.M34x 103)-(158.164)(2.595


3(9.104

108)

103)-(158.164)2

.!

all =6.41 1 x 105psi@(cp-MMscf/D)


For t=4,0 hr
point

CF

qg

qgz

@p

qfi

1.297 X 106
1.642 x 106
1.995 x 106

1
3

31612
56:287

999
3,168

4.100x 107
9.240 X 107

70.265

4,937

1.402 x 10

~ti = [4.934 X 106)(9.104X 103)- (158.164)(2.736X 108)


X

34567s9

10

The, hr
2 . Brar ~

3(9.104

,.

to D~

It It

Calculate slope, m, and intercept, c, of the kest-tit line of


the p!ot of at versus log t. The least squares equations to
calculate slope and intercept are given by Eqs. 22 and 23,
respectively.

7.

103)-(158.164)2

a~ = 7.166 x 105psia2/(cp-MMscfJD)
For t=5.Ohr
*

point

qg

%2

1
3
4

31612
56:287
70.265
158,164

999
3,168
4,937
I 04

4,300 x
9.470 x
1,454 x
2,831 X

flg

107
107
10
10

1,360 x 106
1.682 X 106
2.069 X 106
5,1 11 X106

U,3=(5.111 X 106)[9.104 X 103)-(158,164)(2.831


3(9.104

108)

103)-(158.164)2
log(t)
0.477

(log(t))z
0,228

2
3

0,602

0,362

0.699

0.489

0.778

0!605

2,556

1,684

point

a13= 7.640 x l@ psia2/(ep-MMsef/D)

I
2S7

~
6,411 X
7,166
7,640
7.829

a, log(t)
105

X @

x l@
X 105
2.905 X 1~

3.058 X
4.314 x
5.340 x
6,091 X
1.880 x

105 I
105
105 I
105 *
106

Estirnatirmthe StabilizedDeliveratilitvof a Gas Well Usirmthe

SPE 29440

~, ~ 4(1.880

I@) -(2.905X l@)(2.556)

4(1 ,684) - (2.556P


m= 4.674 x I@

1.

psia2/(cp-MMscf/D)/cycle

~, = (2.905x l@(l,684)-

P1ot ApP = PP (PWJ - PP (PJ Venus flOW rate? qg$ On W


log graph paper for the measured modified isochronal flow
data.

(1.880 x 106)(2,556)

4 (1.684) - (2.S56~
c= 4.276 x I@ psia21(cp-MMscfD)
8.

1.632 X 10%

Time, hr
3.0
4.0
5.0
6.0

kg = S.S2 md (kg = 4.23 md from Drawdown:


2.

Calculatethe skin factor from Eq. 25


$=1,151

~ -log
[(ml

(&d+3231

3.23]

1oa

10. Determine the stabilization time of the well using Eq, 26 for
a symmetrical, non-circular drainage area. The well is
centered in a square drainage area with a shape factor of
30.88.
3791 4ii E,A

t = 3,791(0.0675) (0.023)(0,000169)(640

t,
I /

(26)

6.0
5.0
4.0
3.0

43560)

(5.52)(30.88)

t~ = 162;7

hr

2
2

11, Determinethe stabilizeddeliverabilityline intercep~a, using


the stabilization time and the semilog slope, m, and
intercept, c, in Eq. 27,
a=mfog(t$)+c
a =4.674x

1($ log (162.7)+(4.276x

105)

q,

3 . Da~

J$,
[log

q log APP)J- J$l log qj j:,

.(1.461

a 10)+

~(1.461

For t=3.O hr

X tOF

+ 4(1.774

2(1.774

AOF = 20S,S MMacflD

X 104)[(I.049

104)

x 109)-

2,003.8]

(fog APp)j

(29

n =

(28)

2h

89

t4Mscf/D
.

$~0g4p~L$l(Ogop)JT

AOF =

We may now determine the deliverability exponent, n, for


each line by least-squares regression analysis on the last
three points, using
-. Eq. 29.

12, The pseudopressure at base pressure is 2003,8 psia2/cp.

. a + ~az + 4 ~[pP@)-PP(14.65)]

456

ManU&

3,

13, We now calculate the AOF ~otential using Eq. 28, the
average b value, and the stabihzed a value calculated from
Eq. 27.

10

(27)

u = 1,461x 106psia%p/(MMsef/D)

*OF=

hr

kg CA

6.270 X 107
6.330 X 107

The deliverability plot is shown below. Note that the fmt


data point does not follow the trend of the higher rate points,
and should be ignored.

(s = -S.2 from Drawdown)

t$ =

5.720 X 107
5390 x 107

(25)

*=,.*5,
[[~:$).,og[~)+
s = -s.0

107.

App, @a21cp
q,= 70,265
qe = 56.287
1.323 X 10~
8.920 X 107
1.402 X 108
9.240 X 107
1.454 X 108
9.470 x 107
1.490 X 108
9.610 X 107

(24)

m,h

x 107
x 107
4.400 x 107

6.0

1.632 X 1($(718)
g= (4.674 x 105)(454)

9.

3.800
4,100
4.300

4,0
5.0

Calculatethe formation permeabilityto gas using Eq. 24 and


the slope of the semilog straight line calculatedfrom Eq. 22,
kg=

App, psia2tcp
31.612
q,=44.313

q,=

s~ 23440

J,L, Johnston, W,J. Lee, and T.A, Blasingame

a(l -n)

3 (41.6872) - (5.2436)(23.8294)= ~ ~502


*=

3 (189.3466) - (23.8294P

(16)

%=m

1.467 X 106(1 - 0,54)

flg =

~r t=4.O hr

1.774 x 104 [2(0,54) -1]


qg =

6,

475.49 MMscf/D

Calculatethe stabilized C value using Eq. 15.


c=

n2 =

5.2436

log ApD
77973
7:9763
8.1626
23.~ 6 2

(log 4.)2

3 (41.8739) - (5.2436)(23.9362)=

7.

log qg log Apn

607979
63:6214
66.6280
191.0473

3 (191.0473) - (23.9362P

128383
. .
13,9617
15.0739
41,8739

75.9

Substitute the stabilized C value from Step 6 into Eq, 17 to


calculate the AOF of the well. Use the average value of n
calculated in Step 3.

05486

AOF = 211.4 MMscf/D


Table 6- Comparison of Results

n
3 (41.91 13) - (5.2436)(23.9573)=

c
a, sia2/(cpA scf/D)
b, psiaZ/(c MMsefJ )
AOF, MMscflD

0.5375

3 (191.3865) - (23.9573?

We now calculate the avecageslope,ii, to be


+ 0.5 423 + 0,5486 + 0.537 ~ = 0,54
ii = QJ502
4

Determinethe theoreticalvalue of the Houpeurtcoefficient a


using permeability and skin factor values calculated
previously using Brar and Aziz analysis (k= 5,52 md, s =
-5.0) in Eq. 30. We can usc the average value for the
coeffkicnt b obtained from Brar and Aziz analysis.
~= L422x106T
,h

[51o(%1-~+l
30)
n
c

aMp$sia/;p-

10,06(640 X 43560) -2.5,0

= 1,422x l@(718
(5.52)(454)
a=

~ 1151 o* (30,8828) (0,26~5)2

1[
1.467x l@ psia2/(cp-MMscfP)

b = 1,774x

5.

II

psiaz/(c MMscf~ ~
AOF, MMscf/D

Stabked C

Brar&A ZIZ

054
2.843x 103

.-.
.. .

1.467 X

1.461 X 106

106

1,774 x 104

1,774 x 104

211.4

205.5

Rawims&
0,54
2.426 X 10-3

1 ---

I
. ..

. ..
1.455 x 106

4
I

b,

104psia2/(cpMMscf/D2)*

Calculate the rate at which the change in pseudopressure


determined using the Rawlins and Schellhardt equation is
quai to the change in pseudopressuredetermined usin the
Houpeurt equation, We use the average slope of the
deliverability plot, and the theoretical a and b values in Eq,
16.

---

1,774 x 104

180.1

205,6

SUMMARY AND CONCLUSIONS


In summary, this paper has presented a method for gas weli
deliverability forecastingin which the results of the more rigorous
Hou ttrt analysis can be written in terms of the more famiiiar
Rawr ins and Schelihardt quation, includin the famiiiar C and n
parameters, Although rcfs, 5-7 identify ti!e correlating results
(E@, 15 and 16) for relating Houpatrt analysis and Rawiins and
Scheiihardt anaiysis, these references do not address the use of
these correlating rcsuits for deliverability forecasting, Nor do
these references illustrate the iinkage between Houpeurt analysis

(17)

AOF = 2.843 X 10-3 [1,049X 109- 2003.8~54

--

= 2,S43X 103

AOF = C [PPE) - PP (14.65)~

or t=6.O hr

4 =

(15)

[(1.467 x 106~475.49)t (1;74 x lo4k47s,49~]054

---log(jg
16465
1:7504
1.8467
n3 =

c=

rx r=5.O hr
2

g n

aqR+ bq82

3 (41.7932) - (5.2436)(23.88981= 05423


3 (190.3091) - (23,8898P

piiiti

From Step 4 of Brar and Aziz analysis of these data


269

..

and Rawlins and Schellhardt analysis with a detailed, illustrative


example+

In addition, wc have drawn the following conclusions regarding


this work:
1. An analytical relationship between the ,::ore rigorous
Houpeurt relation (Eq. 7) and the cortventional Rawlins
~d:;h~#mrdt relation (Eq, 6) has been developed (Eqs,
The relationship between the Houpeurt analysis and
Rawlins and Schellhardt analysis has been illustrated and
verified using a field case from ref. 9. The results for
delivers-bility forecasts, in this particular case, were
within 3 percent,

NOMENCLATURE
a

at

A
AOF
b

Cf
C8
c1
Cwl
c

cA
D
:g
m

n
P
Pa
Pb
Pp
Pp@l

Houpeurt flow coefficient, (psiaVcp)/(MMscf/D)for


calculations in terms of pseudopressure or
psiaz/(MMscf/D) for calculations in terms of
plessure-squared
transie%t deliverability coefficient, (psia2/cp)/
(MMscf/D) for calculations in terms of pseudopressure or psia2/(MMscf/D) for calculations in
terms of pressure-squared
drainage area of well, ftz
absoluteopen flow potential, MMsct7D
Houpeurt flow coefficient, (psiaz/cp)/(MMscf/D)z
for calculations in terms of pseudopressure or
psiaz/(M Mscf/D)z for calculations in terms of
pressure-squared
plot, (psia2/cp)/
intercept of a, or (al + bq) vs. log (t)
(MMscf/D)
for calculations in terms of
pseudopressureor psia2/(MMscf/D)for calculations
m terms of pressure-squared,
formationcompressibility,psirl
gas compressibility,psia-1
total systemcompressibility,psia-l
water compressibility,psiw1
Rawlins and Schellhardt stabilized performance
coefficient, (MMscf/D)/(psia2/cp)for calculations in
terms of pseudopressure or (MMscf/D)/psia2 for
calculationsin temnsof pressure-squared
well drainage area shapeconstantor factor
non-Darcyflow constant,D/MMscf
net formation thickness,ft
effectivepermeabilityto gas, md
slope of al or (al + @) vs. log (r) plot, (psia2/cp)/
(MMscf/D) pr cycle for calculations in terms of
pseudopressure or psia2/(MMscf/D) per cycle for
calculationsin terms of pressure-squared.
exponent of the Rawlins-Schellhardt quation and
reciprocal slope of the line on a log-logdeliverability
plot
pressure, psia
atmosphericpmssu~, psia
base pressure, psia
real gas pseudoprcssum,psiaz/cp
average reservoir pseudopressure,psia2/cp
flowing,sandfacepseudopressure,psia2/cp

average resr,mmirpressure, psia


bottomhole flowing pmssm at sandface,psia
gas flow rate, MMscflD
wellbore radius, ft
skin factor, dimensionless
temperature,oR
temperatum,F
real gas law deviation factor evaluatedat the average
mervoir pressure and temperatum,dimensionless
gas viscosity evaluated at average reservoir pressure
and tempenmuv,cp
porosity of reservoir rock, fraction

We have presented the Houpeurt analysis (in terms of


pseudopressure)for a case from the literature and verifiedthat the
Houpeurt/Rawlinsand Schellhardtcorrelation is valid and should
yield accurate performance predictions as well as satisfy
regulatory requirements which ray~ire anai.lsis results from the
Rawlinsand Schcllhardtxthod.

2.

---

REFERENCES
1.

Rawlins, E,L. and Schcllhardt, M,A.: Backprewre Data


on Natural Gas Wells and Their Application to Production
Praclices, Monograph Series, USBM (1935) 7.

2. Cullender, M.H.: The Isochronal Performance Method of


Determining the Flow Characteristics of Wells, Trans.
AIME (1955) 204, 137-142,
3, Katz, D.L,, Cornell, D., Kobayashi, R., Poettmann, F.H.,
Vary, J.A,, Elenbaas, J.R., and Weinaug, C,F,: Handbook
of Naturaf GUSEngineering, McGtaw-Hill Book Ce.,New
York, N.Y, (1959).
4. Houpeurt, A.: On the Flow of Gases in Porous Media,
~fyl$MLInstitut Fracdais du Petrole, XIV (1 1), (1959)
.,
5. Brigham,W.E.: EstimatingResavoir Pammeters Fmm the
Gas Backpressure Equation, SPERE (May 1988)649-650.
6. Duong, A.N.: Discussion of Estimating Reservoir
Parameters From the Gas Bac~ressure Equation} SPERE
(Nov. 1988) 1328.
7, Poettmann, F.H, and Kazemi, H,: Discussion of
Estimating Reservoir Parameters From the Gas
Backpressure Equation, SPERE (Nov. 1988) 1328-1330.
8. A1-Hussainy, R,, Ramey, H,J., Jr,, and Crawford, P,B,:
The Flow of Real Gases Through Porous Media, JPT
(May 1966),
9, Ilrar, G,S, and Aziz, K,: Analysis of Modified Isochronal
Tests to Predict the StabilizedDeliverabilityPotential of Gas
Wells Without Using Stabilized Flow Data, JPT (Feb.
1978) 297-304; Trans. AIME, 265,

You might also like