Chapter 9

Decline-Curve Analysis for Gas Wells

9.1 Introduction

This chapter discusses decline-curve methods for estimating ulti reser

mate gas recoveries and predicting performance from the analysis
of long-term gas-production data either from individual wells or
from entire fields. We begin with conventional analysis techniques
first presented by Arps. I These conventional techniques include
equations for exponential, harmonic, and hyperbolic decline. Next,
we introduce production decline type curves and illustrate their ap
plication to the analysis of gas-production data and the estimation
of formation properties. We also show how decline type curves can
help predict well or field performance. All analysis techniques are
illustrated with examples.

9.2 Introduction to Decline-Curve Analysis

The basis of decline-curve analysis is to match past production per

formance histories or trends (i.e., actual production rate/time data)
with a "model." Assuming that future production continues to fol
low the past trend, we can use these models to estimate original
gas in place and to predict ultimate gas reserves at some future
in place and reserves, however, are based on the total gas volume techniques. For example, infield development wells could reduce
in place, part of which may be unrecoverable with the existing the current drainage area and subsequent ulti mate gas reserves of
wells because of unidentified reservoir discontinuities or existing wells, or proration schedules set by gas regulatory
heterogeneities. Again, we emphasize that the basis of decline- agencies may require some wells to be shut in peri
curve analysis for estimating gas in place and reserves at some odically.
future abandonment condition is the assumption that future
production performance can be modeled with past history. Any 9.3 Conventional Analysis Techniques
changes in field development strategies or production operation
Early attempts at decline-curve analysis sought to find plotting tech
practices could change the fu ture performance of a well and
niques or functions that would linearize the production history. Be
significantly affect reserve estimates from decline-curve
cause linear functions are simple to manipulate mathematically or
DECLINE-CURVE ANALYSIS FOR GAS WELLS 215

Rate, q Log
Rate,q

Harmonic

Tune, I

Fig.9.1-Decllne-curve shapes for a Cartesian plot of rate changes. If, for example, water is entering the well's drainage
vs. time. area, the char acter of the well's decline may change suddenly,
abruptly, and nega tively.
3. The equation assumes that the well analyzed has constant per
trend is linear, extrapolate the trend graphically or mathematically
meability and skin factor. If permeability decreases as pore pres
to some future point.
sure decreases, or if skin factor changes because of changing damage
Eq. 9.1 is based on four important and widely violated as·
or deliberate stimulation, the character of the well's decline changes.
sumptions.
4. It must be applied only to boundary-dominated (stabilized) flow
I. The equation assumes that the well analyzed is produced at
data if we want to predict future performance of even limited du
constant BHP. If the BHP changes, the character of the well's
ration. If the data "fit" with a decline curve are transient, there
decline changes.
is simply no basis for predicting long-term performance. Until all
2. It assumes that the well analyzed is producing from an un
the boundaries of the drainage area (or reservoir) have influenced
changing drainage area (i.e., fixed size) with no-flow boundaries.
the well's decline characteristics, predictions of the long-term
If the size of the drainage area changes (e.g., from relative
decline rate are not unique and, except by sheer accident, are in
changes in reservoir rates), the character of the well's decline
 Fig.9.2-Decline-curve shapes for a semllog plot of rate vs.
time.

Figs. 9.1 through 9.4 show typical responses for exponential,

hyperbolic, and harmonic declines. Because of their characteristic
shapes, these plots can be used as a diagnostic tool to determine
the type of decline before any calculations are made. We elaborate
on the different curve shapes in the following sections. In addi
tion, we illustrate the analysis of gas-well production data using
conventional techniques.

9.3.1 Exponential Decline. Exponential decline, sometimes

called constant-percentage decline, is characterized by a decrease
in pro duction rate per unit of time that is proportional to the
production rate. The exponential decline equation can be derived
from Eq. 9.1. When b=O, Eq. 9.1 takes the special form (which
must be derived with a limiting process as b->0)

qi
q(t)=-- =qie-D;r_ ............................ (9.2)
eD;t

Taking the natural logarithm (In) of both sides of Eq. 9.2 gives

In[q(t)] =ln(qi)+In(e-D;t ), ........................ (9.3)

which, after rearranging, gives

216 GAS RESERVOIR ENGINEERING

Because the natural logarithm is related to the logarithm to the Substituting the rate from Eq. 9.12 into Eq. 9.14, we obtain the
base 10 (log) by ln(x)=2.303 log(x), we can rewrite Eq. 9.4 in rate/cumulative production relationship for harmonic decline,
terms of the log function as
q;
D;t Gp(t)=2.303-[log q;-Iog q(t)], ................ (9.15)
log[q(t)]=log(q;)---......................... (9.5) D;
2.303

or, in terms of production rate,

The form of Eq. 9.5 suggests that a plot of log gas flow rate,
q(t), vs. twill be a straight line with a slope -D;f2.303 and an
intercept log(q;). Fig. 9.2 shows the linear relationship on semi
log q(t)=Iog q;-()ap(t). ................ (9.16)
log coordinates. If the production data exhibit linear behavior on 2.303q;
this semilog plot, we can use Eq. 9.5 to calculateD; from the slope
and q; from the intercept. After calculating the initial decline rate The form of Eq. 9.16 suggests that a plot of log q(t) vs. Gp(t)
and the initial gas flow rate, we can use Eq. 9.2 to extrapolate the
will be linear with a slope of -(D;f2.303q;) and an intercept of
production trend into the future to some economic limit. From this Iog(q; ). This is a much simpler method of calculating the decline
extrapolation, we can estimate gas reserves and the time at which rate for harmonic decline than the rate/time plot because we can
the economic limit will be reached. make a direct plot without prior knowledge of D;.
The curve of rate vs. cumulative production for exponential
decline will be linear on a Cartesian graph, as the following deri 9.3.3 Hyperbolic Decline. When O<b< 1, the decline is hyper
vation indicates. If we integrate Eq. 9.2 from initial time to time bolic, and the rate behavior is described by
t, we obtain

q;
1 1 q(t)= .............................. (9.1)
Q(t)= j q(t)dt= j q;e-D;Idt . ...................... (1+bD; t) ltb

(9.6)
Taking the logarithm of both sides of Eq. 9.1 and rearranging
0 0
yields
The cumulative gas production is
1

Gp(t)=(- e-D;t) .......................... (9.7) I

Iog[q(t )] =log(q; )--log (I +bD; t). ................ (9.17)
Dl o
b
Rearranging yields
The form of Eq. 9. 17 suggests that, if rate/time data can be
1 q; of plot.
GP(t)=--(q;e-Di 1)+-. ...................... (9.8)
D; D; 9.3.2 Harmonic Decline. When b =I, the decline is said to be har

Combining Eqs. 9.2 and 9.8, we can write the cumulative pro
duction relation in terms of rate,

1 q;
Gp(t)=- -q(t)+-. ........................... (9.9)
D; D;

Rearranging and solving for production rate, q(t ), gives

q(t)= -D;Gp(t)+q; . ............................

(9.10) Eq. 9.10 suggests that a plot of q(t) vs. Gp(t) will yield a

straight
line of slope -D; and intercept q;. Fig. 9.3 illustrates this type
 216 GAS RESERVOIR ENGINEERING
modeled with the hyperbolic equation, then a log-log plot of
q(t) vs. (1 +bD;t) will exhibit a straight line with slope of 1/b
and an intercept of Iog(q; ). To analyze hyperbolic decline data,
however, requires that we have prior estimates of b and D; or
that we use an iterative process to estimate the values of b and
D; that result in a straight line.
•
The cumulative production/time relationship is obtained by in
tegrating Eq.
9.1:

Gp(t)= j q(t)dt= j I q; dt, ............... (9.18)

o o (l+bD;t)llb

or, after integrating and rearranging,

q;
Gp(t) = [(1 +bD; t)O-bll( -b) -1]........... (9.19)
D;(b-1)
DECLINE-CURVE ANALYSIS FOR GAS WELLS 217

Example 9.1-Estimating Future Performance of a Gas Well

TABLE 9.1-GAS WELL PRODUCTION HISTORY,EXAMPLE 9.1
With Conventional Decline-Curve Analysis Techniques. Use the
gas-production flow rate and cumulative production history in Ta
Producing Cumulative Gas Flow
Production Rate ble 9.1 to predict rate behavior 15 years into the future. Assuming
Time
(days) (MMscf) (Mscf/D) that the economic limit for this well is 30 Mscf/D, estimate the ul
timate recovery and total productive life of the well using Arps' 1
30 13.4589 413.3
392.8 conventional analysis techniques.
60 25.3066
90 36.3221 375.9 Solution.
120 47.815 371.3 l. Because we do not know which decline equation best models
150 60.7706 377.5 the past production performance, we construct the rate/time and
180 71.1327 367.8 rate/cumulative production plots in Figs. 9.5 through 9.8 to ex
210 80.6358 356.8
amine the production characteristics.
240 90.3544 349.0
2. Examination of the curves in Figs. 9.5 through 9.8 does not
270 105.643 361.7
show conclusively that the decline is exponential, harmonic, or
300 113.646 349.1
330 122.878 341.9 hyperbolic. Note that not only are the semilog rate/time and the

360 137.776 350.1 Cartesian rate/cumulative production plots linear (indicative of ex

390 142.799 333.6 ponential decline), but also the semilog rate/cumulative production
420 147.511 291.4 plot is linear (indicative of harmonic decline). Consequently. we
450 168.504 338.2 will analyze this example using both exponential and harmonic
480 175.674 329.1 decline methods.
510 183.737 322.5
Exponential Decline Analysis.
540 198.204 327.1
l. The slope of the semilog rate/time plot (determined with a least
570 199.765 310.9
215.121 316.6 squares fit of the data) is
600
660 230.559 305.6
m= -0.0001317.
720 248.155 298.7
780 264.898 291.6 The initial decline rate is
840 287.17 290.8
900 296.938 278.0 D;
960 327.427 284.8 m=- = -0.0001317,
1,020 341.435 276.9 2.303
1,170 376.068 259.9
1,200 379.859 254.8 where D;=0.0003033 day-! =0.1107 year I

1,320 416.501 249.4

2. The intercept of the semilog rate/time plot (determined with
1,410 426.793 236.1
a least-squares fit of the data) is
1,500 458.434 237.4
1,620 482.743 230.1 log q(t)=2.58,
1,710 508.14 224.2
1,800 531.01 219.4 which is equivalent to the log of the initial rate, q;. where q; =
1,980 554.58 199.9 J02.58 =380.2 Mscf/D.
2,070 575.818 202.1 3. We can now substitute q; and D; into Eq. 9.2 to get a partie- •
2,190 601.082 196.4
ular decline equation for this well:
2,280 626.139 189.8
2,310 635.765 190.5 q(t) =380.2e(-0.0003033r),
2,400 648.646 183.8
2,580 678.628 176.4 with time in days, or
2,700 702.659 170.3
q(t) =380.2e<-O.t107t),
2,880 722.806 143.7
2,910 735.055 156.1 with time in years. Note that time is counted from t=O, so to ex
3,000 742.635 154.6 trapolate for the next 15 years, we must start at t= 16 years. The
3,400 791.57 139.6
calculated future rate performance in Table 9.2 is calculated with
3,600 835.583 138.1
the decline equation from Step 3.
4,000 881.494 123.4
4. Recall that we assumed an economic limit of 30 Mscf/D for
 218 GAS RESERVOIR ENGINEERING

g
()

"'
100

£
"'
Q o: ;

+.. . . . . .,... .. . . ..+..-............... +---·-·-"''""'"!- .......,.,... ---1..-·---..--..

,............
0•- ---+--+ -+--+--+--
1 t - +
10-1----+----t---T----t----+--'- 0 200 1200
600 1000
0 1000 2000 3000 4000

Time (t), days Cumulative Production (Q), MMscf

Fig.9.8-Semilog plot of rate vs.cumulative production, Ex

Fig. 9.7-Semllog plot of rate vs. time, Example 9.1. ample 9.1.

q(t) = 380.2e< -0.00030331), 2. The slope (determined with a least-squares fit of the data) of
the semilog rate/cumulative production plot is -0.0005478. For
30=380.2e< -0.00030331),
cumulative production (in MMscf), the slope of the line on a semi
or, solving for time, log rate/cumulative production plot equals

1=8,373 days=22.9 years. l,OOOD;

-- = -0.0005478.
5. We can use the equation derived for the rate/cumulative pro 2.303q;
duction behavior for exponential decline (Eq. 9.10) to calculate
the ultimate recovery when the economic limit is reached. Substituting for q; gives

q(1)= -D;Gp(t)+q1 • l,OOOD;

---- =0.0005478.
We know that q;=380.2 Mscf/D and D;=0.0003033 day-1: 2.303(407.4)
therefore,
D 1=0.000514 day -I =0.1876 year-!.
q(1) = -0.0003033Gp(t) + 380.2.
3. We can now substitute q; and D; into Eqs. 9.11 and 9.12 to
With the assumption that the remaining production history of this
obtain specific rate/time and rate/cumulative production decline
well can be modeled with exponential decline, the ultimate recov
equations, respectively, for this well. The rate/time relationship is
ery from this well at an economic limit of 30 Mscf/D is
407.4
380.2-30 q(t)
Gp(t) Mscf I +0.0005141
0.0003033
L First, we calculate q; from the intercept of the semilog
=I,155,000 Mscf= 1,155 MMscf.

Harmonic Decline Analysis.

 218 GAS RESERVOIR ENGINEERING

with time in days, or

407.4
q(t)= ,
I +0.18761
DECLINE-CURVE ANALYSIS FOR GAS WELLS 219

10 (407.4/30) 1
t=-----

0.000514
! 10'
=24,475 days=67 years.
.s

1 6. At an economic limit of 30 Mscf/D and with the assumption

that future production can be modeled with harmonic decline, the

i
c
ultimate recovery from this well is

-a
c=
Gp(t)=4,764,264-(1,825,374)log q(t)
•
E =4,764,264-(1 ,825,374)log(30)
a
=2,067,965 Mscf=2,068 MMscf.

9.4 Decline Type Curves

Dimensionless Decline Time
Type curves (see Chap. 6) are plots of theoretical solutions to flow

4
equations and can be generated for virtually any kind of reservoir
Fig. 9.9-The Fetkovich rate/time and cumulative produc
model for which a general solution describing the flow behavior
tion/time decline type curve.
is available. For type curves to be applied correctly, the engineer
must completely understand the assumptions underlying the solu
0.0005140 ) tion. Furthermore, those assumptions must accurately model the
log q(t) =log(407.4)- ( Gp(t), well or reservoir conditions being analyzed. Decline type curves
2.303 x407.4
have been developed so that actual production data can be matched
without special graph paper or the trial-and-error procedures re
or Gp(t)=4,764,264-(1,825,374)log q(t).
quired for the conventional decline-curve methods in Sec. 9.3. Type
4. We can extrapolate future performance (corresponding rates curve methods use log-log graph paper to match preplotted theo
and times) for 15 more years using the rate/time equation in Step retical solutions with actual production data. Further, type-curve
3. Note that time is counted from t=O. Therefore, to extrapolate analyses allow us to estimate not only original gas in place and gas

for the next 15 years, we must start at t=16 years. Table 9.3 sum reserves at some abandonment conditions, but also the flowing char
acteristics of individual wells.
marizes the predicted future performance.
This section presents the theoretical basis (including assumptions)
5. Recall that we assumed an economic limit for this well of 30
and practical applications of two type curves-the Fetkovich4 and
Mscf/D. We can substitute that rate into the specific rate/time rela
the Carter5 type curves-that are particularly useful in gas-well
tionship developed in Step 3 for this well to find the total produc
decline-curve analysis. We recommend these type curves for manual
tive life of the well.
or graphical decline analysis because they are based on theoretical
considerations, unlike Arps' I empirical decline-curve analysis·
407.4
30= -- techniques. Note that no decline-curve methods presented in this
1 +0.000514t chapter include non-Darcy effects.

TRANSENT+DEPLETION
220 GAS RESERVOIR ENGINEERING

9.4.1 Fetkovich Decline Type Curve. The Fetkovich4 decline q(t)=gas flow rate, Mscf/D; t=time, days; k=permeability to gas.
type curves are based on analytical solutions to the flow equations md; h=thickness of productive zone, ft; Pp(P;)=real-gas pseu
for production at constant BHP from a well centered in a circular dopressure function evaluated at initial reservoir pressure,
reservoir or drainage area with no-flow boundaries. Although these psia2Jcp; Pp(Pwtl =real-gas pseudopressure function evaluated at
type curves were developed for a homogeneous-acting reservoir, bottomhole flowing pressure (BHFP), psia2fcp; Psc=pressure at
they can be used for analyzing long-term gas-production data from standard conditions, psia; Tsc =temperature at standard conditions,
hydraulically fractured wells during the pseudoradial flow period oR; T=formation temperature. oR; <;!>=porosity, fraction; p. =gas
8
and once the outer reservoir boundaries affect the pressure response. viscosity, cp; and c r =total system compressibility, psia I.

Fig. 9.9 is an example of the Fetkovich decline type curves for Like the type-curve analysis procedures in Chap. 6, application
both rate/time and cumulative production/time analyses. of the Fetkovich type curves requires that we match the shape of
The type curves in Fig. 9.9 include both transient or infinite- the field data with a type curve. From this match, we can estimate
acting and boundary-dominated flow regimes. Both the transient gas reserves and formation properties. The following procedure is
rate/time and cumulative production/time type curves are recommended for decline-curve analysis with the Fetkovich type
characterized by a correlating parameter defined as the ratio of curve.
the outer drainage radius to the apparent wellbore radius, relrwa• Procedure for Gas-Well Decline-Curve Analysis With the Fet
while the pseudo steady-state flow regimes are characterized by the kovich Type Curve.
Arps decline con stant, b. Again, b=O corresponds to exponential l. Plot q(t) and Gp(t) vs. ton log-log paper (3-in. log cycles)
decline behavior, while b= I represents harmonic decline. or tracing paper with the same size logarithmic cycles as the Fet
Values in the range kovich type curve.
0 < b < I suggest hyperbolic decline characteristics. 2. Match the cumulative production data to the best-fitting type
Fig. 9.10 shows a more complete example of the rate/time type curve. Note that the cumulative production data plot often is much
curve. Again, two flow periods are represented. The curves at small smoother than the rate plot and therefore is easier to match to de
values of dimensionless times, representing the transient or infinite termine the Arps decline constant.
acting rate response, were generated with the analytical solution Because decline type-curve analysis is based on boundary
to the radial diffusivity equation. All the transient curves converge dominated flow conditions, there is no basis for choosing the proper
at a dimensionless time of about 0.3, indicating the approximate b values for future boundary-dominated production if only tran
beginning of boundary-dominated flow. The boundary-dominated sient data are available. In addition, because of the similarity of
flow responses which were generated with Arps' empirical decline curve shapes, unique type-curve matches are difficult to obtain with
equation, are characterized by b. Fetkovich4 and Fetkovich et al. 6 transient data only. If it is apparent that boundary-dominated data
showed, however, that the Arps equation for exponential decline are present and can be matched on a curve for a particular value
(i.e., b=O) is a late-time solution for the constant-pressure case. of b, we can extrapolate into the future accurately.

Like most type curves developed for pressure and flow-rate data 3. Record values of the correlating parameters for transient and

from gas wells, the Fetkovich type curves are plotted in terms of boundary-dominated flow (i.e., relrwa and b, respectively) from

dimensionless variables. Specifically, the Fetkovich rate/time type the match of the cumulative production data. Next, force a fit of

curves are plots of dimensionless rate, the rate/time data with a type curve having the same values of
rJr wa and b. Note that the cumulative production/time and
1
50,300q(t)psc T[ln(relrwa)- h] rate/time curves are not matched simultaneously but individually;
qDd= , ............. (9.22) i.e., we move and rematch the field data plot overlying the type
Tsckh[pp(P; )-pp(Pwf)]
curves.
vs. dimensionless time, 4. Select a rate match point [q(t), qDd1MP on the rate/time curve
and calculate formation permeability using the definition of dimen
0.00633ktl¢p. c ,r£a sionless flow rate given by Eq. 9.22:
tDd = _;;8:__ . .. ........... (9.23)
l!z[(relrwa)2 liz]
J. q(t)] 50,300pscT[ln(relrwa)- \12]
k L .........(9.30)
Similarly, the cumulative production/time type curves are plots
qDd MP Tsch[pp(P;)-pp(Pwt)]
of dimensionless cumulative production defined by
)]
637.8PscTGp(t)
Qvd= ....... (9.24)
2 2
220 GAS RESERVOIR ENGINEERING
5. Calculate the initial surface gas flow rate, q;. at t=O from the rate match point:
5 8
DECLINE-CURVE ANALYSIS FOR GAS WELLS 221

TABLE 9.4-RESERVOIR PROPERTIES, EXAMPLE 9.2

Net pay, ft Wellbore 32.0

radius, ft Initial 0.365
pressure, psia 3,500
Pseudopressure evaluated at initial pressure,
2
psia /cp 8.322x 108 ...
u

BHFP, psia 500

2
Pseudopressure evaluated at BHP, psia /cp 7
2.106x 10
Reservoir temperature, °F
180
Wet gas gravity (air= 1.0) §
0.689 ...0
Water saturation, fraction
0.34
1
Water compressibility, psia -
3.6x1o- 6 §
Formation compressibility, psia-
1
4x10- 6 ..u

Porosity, fraction 0.12

-;::-
Gas viscosity evaluated at initial pressure, cp 0.02095 '"&
Gas FVF at initial pressure, RB/Mscf 0.8174
Total compressibility at initial
1 4
pressure, psia- 1.5741x10-

gebraically by substituting the chosen b value and calculated D 1

and q1 values into the Arps general decline equation,

q(t) ...............................(9.1)
Fig.9.11-Gas production rate and cumulative production vs.
(l+bD;t)ltb
time, Example 9.2.

The production data should not be extrapolated past the economic curve. It is apparent that boundary-dominated data are present be
limit for the well in question. The productive life of the well can
be estimated from the rate/time extrapolation at this point or can
be calculated with the general decline equation developed with Eq.
9.1.

Example 9.2-Decline-Curve Analysis With the Fetkovich Type

Curves. Use the Fetkovich type curve, the gas flow rate and cu
mulative gas production data from Example 9.1, and the well and

reservoir properties in Tables 9.4 and 9.1 to estimate k and s. If

possible, predict the gas-flow-rate behavior 15 years into the fu
ture and, assuming that the economic limit for this well is 30
Mscf/D, estimate the productive life of the well and the ultimate
gas recovery.
Solution.
l. Plot gas flow rate, q(t), and cumulative production, Gp(t),
vs. t on log-log paper or tracing paper with the same size logarith
mic cycles as the Fetkovich type curve (3-in. log cycles). This plot,
shown in Fig. 9.11, is not to scale.
2. Match the cumulative production data to the best-fitting type
 DECLINE-CURVE ANALYSIS FOR GAS WELLS 221

q; =[q(t)lqDd]MP

"'1,000/2.8

=357.1 Mscf/D.

Note that this rate is lower than the rate at 30 days. q 1 repre
sents a hypothetical initial rate that would have occurred had
the well been in boundary-dominated flow at t=O. However,
as the type-curve match shows, the early data are in transient
flow, and boundary effects have not yet been felt.
Consequently, the calcu lated rate at time t=O is lower than
the actual measured rate.
6. While the data are in the same matched position, select a time •
match point, (t,tDd)MP· For this example, we chose t= 100
days and tDd=0.034.
From the time match point, the initial decline rate is estimated
t
o
b
e

D;=(tDdlt)Mp

=0.034/100

=0.00034 day-!=0.1241 year-l.

7. The reservoir PV, VP (ft3), in the drainage area of the well

222 Future Time GAS
TimeRESERVOIR ENGINEERING
q
(years) (years) (Mscf/0)
---
The drainage area is
1 9.5-FUTURE RATE
TABLE 16 PERFORMANCE
82.9
A FROM THE2 FETKOVICH TYPE 77.4 9.2
17 CURVE, EXAMPLE
3 18 72.5
=11"(1,087)2
4 19 67.9
=3,712,000 ft2 =85 acres. 5 20 63.8
6 21 60.0
8. We know the drainage area, so we can estimate s using the 7 22 56.5
transient matching parameter, relrwa·
(
8
0.365)(800) l 23 53.3

l
9 24 50.3
=In 10
.J3,712,000/7r 25 47.6
11 26 45.0
12
1.3. 27 42.7
13 28 40.5
9
. 14 29 38.5
15 30 36.6
W
e

c
a
n

n
o
w

e
x
t
r
a
p
o
l
a
t
e

t
h
e

r
a
t
e

c
u
r
v
e

i
n
t
o

t
h
e

f
u
t
 u is correct only during
r
e transient flow (and possibly
. 9.4.2 Carter Decline Type after boundary effects have
Curve. The Fetkovich type been felt if the pressure
S
u curve was developed to drawdown is small). The
b model the flow of a slightly accuracy of the Fetkovich
-
compressible liquid and type curves for analyzing
stituting the chosen b value and calculated Di and q; values into
consequently assumes that gas wells with large
Arps' general decline equation, we have
the liquid viscosity- pressure drawdowns can be
q compressibility prod uct is improved, however. if we
seems to verity this
; constant over the entire define the dimen sionless
observation. The results
productive life of a well. rate and cumulative
( we obtained from con
I Although valid for modeling production variables in
ventional analysis in
liquid flow during both terms of the real-gas
+ Example 9.1, assuming an
b transient and boundary pseudopressure function.
exponential decline,
D dominated flow regimes, •
; proved to be pessimistic
t for gas flow this assumption
predictions; however,
) t CarterS offered
when we assumed -
= improved accuracy
1 by plotting
l
a harmonic -
i - - functions that in
b decline, we 1-
obtained very -
clude the changes in
I
optimistic 0.0
=357.14[1 results. gas properties with
493
+(0.000136)t
6 pressure. The Carter
]
-2.5Mscf/D. type curve was

Note that time is in days developed specifically

and is counted from time 1 for gas-well decline-

t=O, so if we wish to 0.0 curve analy sis and

/ 493
extrapolate for the next 15 6 improves the
years, we must start att= 16 ( accuracy by

years (5,840 days). For time in considering the

2
years, the general decline variation of the
equation is product P.g( p)cg( p)
.
with average
q(l) =357.14[1 + (0.0496)1]
-2.5Mscf/D. 5 reservoir pressure.
Carter cor
Table 9.5 summarizes the )
related rate/time
future performance estimated behavior during
with the general decline boundary-dominated
flow with
equation developed
in years.
for time
J =
a parameter, c
3 A., defined eval
Recall that we assumed an 4
as the ratio u
economic limit for this well of p. 8 8
of 30
(p;)cg( Pi)
Mscf/D. We can substitute that to ji
rate into the rate/time decline 8
rela ated at the average reservoir
I 0. We can integrate the pressure, p, and calculated
tionship for this well to find
general decline equation to as
obtain a rela
the total productive life of
tionsh.ip between cumulative p.(p;)cg(P;) [pp(P;)-
the well. q(1)=357.14[1 production and time measured pp(Pwf)]
in years. A.= ...
+(0.04936)t] -2.5. (9.37)

Solving for time 1 yields

DECLINE-CURVE ANALYSIS FOR GAS WELLS 223

. .t..:lU

Fig. 9.12-Th
"• L'IIW curve.

!f•LOol5
The correla
is f.., which is
11•L0t
11•LON pa rameter du
is a function
R is
given in Fig.

7
1
(
(
1 R
A=. 1 2
-
2
-
A= .75 ----- 1 1
)
A= .55----- (1
= (
qo = 1424 qT (1/Bt) 0.
:
:
)
akh[m{pt)-m(pw)] 25 .
to = 2.6,3.4 X ]Q -4 X 24 kt
.
rpp.jCgi rw2 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
 (9.40)

where R=relrwa,
..................
...............
(9.41)

re=.J7ij;, ......
................
..............
(9.26)
224 GAS RESERVOIR ENGINEERING

TABLE 9.6-RESERVOIR AND WELL PROPERTIES DATA,

EXAMPLE 9.3

Net 3
Wellbore 2
.
ft 0
psia 0
.
Pseudopressure
3
evaluated
initial 6
5
3
BHFP, ,
5
Pseudopressure
0
evaluated
0
BHP,
Reservoir
8.3
temperature,
22
X
Wet
10
(air=
8

saturation,5
0
fraction Water
0
compressibility,
2.
psia
10
6x
Formation
10
compressibility,
7
psia
1
Porosity, 8
Gas 0
evaluated
0
initial .
cp 6
Gas 8
pressure,9
RB/Mscf 0
Total .
3
compressibility
initial 4
3.
6
x
1
o
-s
4
x
1
0
-

0
.
1
2
0
.
0
2
0
9
5
0
.
8
1
 7 4 10-4 85
1

R Fig. 9.14-Variation of the

parameter 11 with R for the 5. Calculate permeability

Carter 5 from the rate match point

type curve. using the defi nition of
dimensionless flow rate. For
radial flow, estimate a from
We recommend the
Fig. 9.13 and calculate
following procedure for using
permeability using Eq. 9.47.
the Carter type curve.
Procedure for Gas-Well
Decline-Curve Analysis
Using the Car
k=[ q(t) l 1 ,

t 424TTJ[ln(R)-0.75]
e
r
................ ( .4 )
9 7
T
y qD MPah[pp(Pi)-
p pp(Pwjl]
e
For linear flow, aB 1 = 2 and
C
u
r Lph[pp(P:)1
:p(Pwj)]
(9.48) . ................
v
e k=[ : )
.
I. Compute the
pseudosteady-state-flow
correlating parameter, A:

9 37
A= !L(Pi)cg(Pi) [pp(Pi)-pp(Pwjl] . .............. ( .
)
2 [(p/z)i -(plz)"f]

6. If the drainage area or

2. Plot the gas flow rate in
drainage area radius is known,
Mscf/D vs. tin days on
we can calculate the skin
tracing paper with 3-in. log
factor. Combining Eqs. 9.26
cycles (i.e., the same size
and 9.41, we obtain
log-log paper as the Car ter
type curves).
l
3. Match the
data on the Carter type curve
production
s=ln l .J A ;;
rw R =In

with the appropriate value of

A calculated in Step 1. (r-";;.R )...............
Maintaining the data in the
matched position, select
rate, [q(t),qD]MP• and time,
(t,tD)MP• match points. In
.........
(9.49)_
addition, choose a value for 1J
from the early-time match. 7. Estimate the recoverable
4. With the value of 1J from gas, G, (in Mscf), when the
Step 3, read the
corresponding value aver age reservoir pressure
has reached the constant
BHFP. Use the rate and time
m e ...............
P . ..... ... (9.50)
a c
t t For linear flow that occurs
in hydraulically fractured 8. Extrapolate the future
c e wells be performance to the economic
h d fore the onset of pseudoradial limit of the well using the
flow, determine a 1 and aB 1
type curve chosen for the
from
p i Eqs. 9.45 and 9.46: match.
o n
i
n
S
t
t
s
e
p
s
e
3
l
.

o
f
R G
fr
o
m
,=
F
i
g
. l
9
.
1
4 :
.
F
o
r
r L
a
d
ia
l p
fl
o

c
w
,
c
al
c
u
la
te
B
:)
1
fr
o M
m

E
q
.
9
.
4
4
:

IIB
 (1,424)(640)(1.004)[ln(l00) -0.75]
DECLINE-CURVE ANALYSIS FOR GAS WELLS

rate behavior 15 years into the future for the well given in Exam
ples 9.1 and 9.2. Assuming again that the economic limit for the
well is 30 Mscf/D, estimate the productive life of the well. In ad
dition, compare the results from the Carter type curve with those
obtained from Examples 9.1 and 9.2. Table 9.6 gives the well pro
duction data. Note that the drainage area, A, is 85 acres.
Solution.
1. First, compute t-. with Eq. 9.37:

2[( plz)i -( plz)wf]

0.02095(2.3059 X IQ-4)(8.322X 108 -2.106X IQ7)

2[(3500/0.89152)-(500/0.95248)]

=0.58.

As an approximation, use the curve for t-.=0.55.

2. Plot q(t) in Mscf/D vs. tin days on tracing paper with 3-in.
log cycles. Note that the plot in Fig. 9.15 is not to scale.
3. Match the Carter type curve with t-.=0.55 calculated from Step
1. Maintaining the data in the matched position, we select the fol
lowing rate and time match points and a value for 71 from the early
match:

q(t)=l,OOO Mscf/D and q0 =2.6.

t=IO,OOO days and t0 =4.2.

and 11= 1.004.

4. Read the corresponding value of R from Fig. 9.14 at the value

of 71 from Step 3.

R=IOO at 71=1.004.

This is an acceptable value of R because we know that we have

radial flow and R should be >30. Therefore,

l!B 1 =TI[ln(R)-0.75]

= L004[ln(l00)-0.75]

=3.87.

5. Calculate k from the rate match point. From Fig. 9.13 for a
cylindrical reservoir, a= I. Therefore,

k=[q(t)] 1,424T71[ln(R)-0.75]

% MP ah[pp{pi)-pp(Pwj)]
 225 reser voirs, which take long times to stabilize, may not represent
true

8. Graphically, continue along the curve for t-.=0.55 and boundary-dominated flow.
11 =1.004 and extrapolate future performance (i.e., corresponding
rates and times). Table 9.7 summarizes the estimated future per
formance.
The productive life of the well is estimated from the type curve
match to be 31.5 years (11 ,500 days) at an economic limit of 30
Mscf/D.

Table 9.8 summarizes the results of the analyses from Exam

ples 9.1 through 9.3. Note that the results from the Fetkovich and
Carter decline type-curve analyses agree; however, in theory the
results from the Carter type-curve analysis should be more
accurate because this method incorporates the effect of changing
gas prop erties. Note also that neither the exponential nor
harmonic analy ses agree with the decline type-curve results,
which suggests that the decline behavior of this well is
hyperbolic.

9.4.3 Limitations of Decline Type Curves. Decline-curve methods

provide a method for estimating original gas in place and ultimate
recoveries at some abandonment condition from a well or an en
tire field. In addition, decline curves can be used to estimate future
production and productive life. However, decline-curve analysis
techniques have several important limitations.
Recall that the Fetkovich type curves were generated with the
assumption that the well is produced at a constant BHP. Further,
the type curves assume that k and s remain constant with time.
Any changes in field development strategies or production operation
prac tices, however, could change the production trends of a well
and significantly affect reserve estimates from decline-curve
techniques. For example, proration schedules may require that
some wells be shut in periodically or that production be curtailed,
thus changing the BHFP. In addition, if the well is stimulated
either from acidiz ing or hydraulic fracturing then s changes.
Consequently, Fetkovich4 and Fetkovich et al. 6 recommend
incremental analy sis of the changed production data relative to
the established pro duction trend.
As mentioned, the basis of decline type-curve analysis is the as
sumption that boundaries affect the rate response-i.e., boundary
dominated flowing conditions have been reached. The boundaries
may be no-flow reservoir boundaries, sealing faults, or interfer
ence effects from adjacent producing wells. If true boundary
dominated flow is not established, then there is no theoretical ba
sis for decline-curve methods and predictions of future production
may be inaccurate. Much of the late-time production data from
wells located off-center of the drainage areas or in low-permeability
226 GAS RESERVOIR ENGINEERING

TABLE 9.8-COMPARISON OF RESULTS FROM TABLE 9.10-WELL AND RESERVOIR PROPERTIES

EXAMPLES 9.1 THROUGH 9.3 AND PRODUCTION DATA FOR EXERCISE 9.2 AND 9.3

N ,
Exponential
Harmonic e
Fetkovich Carter
t
Type p
Pr
od Decline s
uc
p i
tiv
e Decline a
a
y Ps
Curve eu
, do
pre
Curve
ss
ure
f ev
t alu
ate
d
at
W initi
e al
pre
l ss
ure
l
,
b
o
r
e

r
a
d
i
u
s
,

f
t

I
n
i
t
i
a
l

p
r
e
s
s
u
r
e
7 Limit, years 2 a l
psia /cp
6 22.9 9 6 14 p
0 67.0 34.3 x 10 r
31.5 BHFP, psia e
37.2 15.3
Ultimate Pseudopressure s
evaluated at s
Recovery, 2 82.5 63.3 u
BHP, psia /cp
MMscf 1,155 6 r
2,068 1,362 6.9052 x 10 38.8 e
60.4
< 1,585 Reservoir ,
temperature, °F
57.5 34.5
Wet gas gravity p
Gas Flow 7 15 s
(air= 1.0)
Rate i
(Mscf/0 Water a
33.3 13.7
saturation,
fraction
79.5 61.5 -
3
F .
u 6 1
56.9 36.9
( x
D
51.5 32.0 r
1 a
0 8 i
- n
6
29.8 a
g
1
e
76.7 TAB
a
x LE
r
1 9.9-
53.6 e
0 PR
a
OD
,
- 49.0 UCT
a
s ION
9 c
Porosity, DAT
r
I 64.7 fraction A
e
101.8 26.7 FOR
83.4 0.095 z factor EXE
s
84.0 at initial RCI Ec
Producing
74.0 Cumulative SE Gas Flow
pressure on
2 57.9 Time Production 9.1 Rate o
97.3 77.9 1.0653 z factor mi
50.6
(days) (MMscf) (Mscf/D)
78.0 at BHFP c
3 51.8 0.97691 lim
182.5
46.0 1,932.1 7,714.9
93.1 72.9 Gas viscosity it,
70.5 evaluated at 365.0 10 3,211.0 6,486.6 M
initial pressure, 547.5 4,311.0 5,635.4 sc
4 46.4 f/
cp
89.3 68.4 730.0 23.9 5,298.2 5,208.5 D
0.0299
68.5
912.5 6,237.3 5,088.5
Gas FVF at 71.6
5 41.5
initial pressure, 1,095.0 7,116.3 4,569
85.7 64.2
RB/Mscf 1,277.5 4,069
62.0
0.61962 47.9 43.5 7,902.5
1,460.0 II 8,609.2 3,691.1
G
a 1,642.5 9,290.0 3,767.7
s 1,825.0 21.4 9,937.4 3,343.6
c 2,007.5 10,520.9 3,059.4
o 69.3
2,190.0 11,079.7 3,064.8
m
p 2,372.5 11,601.3 2,665.6
45.3 42.5
r 12,071.7 2,493.7
2,555.0
e 12
s 2,737.5 12,529.7 2,525.2
s 2,920.0 19.1 12,969.6 2,301.7
i
b 3,102.5 13,373.7 2,130.1
i 3,285.0 67.2 2,112.6
13,760.8
l
i 3,467.5 14,121.0 1,842.7
t 43.0 41.0
3,650.0 14.457.6 1,847.1
y l3
3,832.5 14,780.8 1,698
a 4,015.0 17.1 15,082.7 1,611.1
t
4,197.5 15,367.8 1,515
i 4,380.0 65.2 15,638.1 1,448.2
n
i 4,562.5 15,896.2 1,381.2
t 40.8 37.0
4,745.0 16,141.2 1.305.1
i
9
.
DECLINE-CURVE ANALYSIS FOR GAS WELLS 227

TABLE 9.11-PRODUCTION DATA FOR EXERCISE 9.4 TABLE 9.12-PRODUCTION DATA FOR EXERCISE 9.5

Incremental Incremental
Time Production Time Production
(months) (MMscf) (months) (MMscf)

1 137.792 25 38.195
2 156.079 26 37.856
3 132.681 27 30.186
4 136.731 28 31.671
5 120.615 29 28.547
6 115.589 30 24.549
7 103.547 31 24.018
8 101.154 32 20.904
9 90.616 33 19.660
10 78.505 34 15.847
11 74.353 35 21.624
12 68.654 36 20.655
13 68.500 37 18.012
14 62.803 38 17.638
15 57.499 39 16.076
16 56.763 40 18.732
17 57.599 41 18.244
18 56.193 42 15.752 r
t
19 50.461 43 14.640
i
20 49.463 44 16.198 e
21 48.055 45 18.392 s
?
22 43.747 46 14.758
23 43.482 47 13.082
W
24 38.089 48 13.692 h
a
n a
d t
a
r
e e
s
t
i t
m h
a e
t
i d
n r
g a
w
f b
o a
r c
m k
a s
t ?
i 2.
o W
n hat
ass
p um
r pti
o on
p s
e are
ma
 de in the development and 6. You are asked to 5. Estimate the future decline type curves.
applica provide a production
forecast for a new well. production for the well Assume economic limit
t
with the produc tion data rate is 1% of peak rate.
i You are given reservoir
o in Table 9.12.9 Use
properties, an estimate of
n
conventional decline- Nomenclature
reservoir size and shape, and
curve analysis and both
o gas properties, and the BHP A = drainage area,
f Fetkovich and Carter L2, acres
is specified and con stant
decline type curves. b = Arps decline-
t with time. Describe the tools
Assume eco nomic limit curve constant
h
(equation, plots, etc.) that
e rate is 1% of peak rate.
you could use to make this
6. Use conventional
F prediction. What are the
e decline-curve analysis to
t limitations of this predict rates 15 years
k a
o p into the future for a well
p
v r whose historical rate data
o
i a
c is sum marized in Table
c h
h ? 9.13.
7. Use the data in
t
y Table 9.14 to estimate
p permeability and skin
e
factor and to predict

c rates 15 years into the

u future. Use the Fet
r
v kovich type curve for this
e analysis, and compare
?
your results with those
3. Describe the analysis of based on the conventional
gas-well production data with decline-curve analysis of
decline curve techniques. Exercise 6.
What are the limitations of 8. Use the data in
analyzing gas performance Table 9.14 ro estimate
data with liquid relations permeability and skin
(e.g., exponential decline factor and to predict rates
semilog plot, Fetkovich type 15 years into the future.
curve)? Use the Carter type curve
4. How can we solve the for this analysis.
gas flow problem using Compare your results

solutions based on liquid with those based on the

flow? What steps are conventional decline-

curve analysis of
involved in this process?
Exercise 6 and tfie
5. Suppose that the rate
Fetkovich type-curve
and pressure for a well vary
method of Exercise 7.
continuous ly and
9. Estimate the future
significantly during
gas production for the
production. Suggest a method
well whose pro duction
of analyz ing and
data are summarized in
interpreting these data with Table 9.15.10 Use
conventional decline-curve conventional decline-
analysis and type-curve curve analysis and both
methods. Fetkovich and Carter
228 GAS RESERVOIR ENGINEERING

TABLE 9.14-WELL AND RESERVOIR PROPERTIES TABLE 9.15-PRODUCTION DATA FOR EXERCISE 9.9
AND PRODUCTION DATA FOR EXERCISE 9.7 AND 9.8

N , 53.0
e ps 0.3280
t ia 2,700
Ps
eu
p do
a pr
es
y
su
, re
ev
al
I ua
te
t
d
at
ini
W tia
e l
l pr
es
l su
b re,
o
r
e

r
a
d
i
u
s
,

I
t

I
n
i
t
i
a
l

p
r
e
s
s
u
r
e
 Gas Flow Cumulative
Time 2
psia /cp Rate Production
5.8120x 10'
10 8
(days) (Mscf/D) (MMscf)
3
9.3600 1.7553x 10

5
2.3828x10
3 1.1796x10
BHFP, psia 8 1.8322 X 103
4 . 1.7509 105
1.1152 X 10 X
1
9
3.1570x 10 9
7
3 0
2.3398 X 10
x
4
6.3596x 10 1
0
1

Pseudopressure 1 1. 2.2411
5
evaluated at BHP, . 54 X 10
psia 2 /cp 2.8887 x 1 26 2.7016
7 5
10 9 X x 10
0 10 2.9970
Reservoir 3
temperature, °F 2 X 10
5
x 1.
Wet gas gravity (air= 58
1.0)
1 06
X
0 10
2
3
1
. 1.
4 55
8 57
5
1 x
10
X 3

1
0
2

1
.
6
7
3
5

1
0
2
2
2.0637x 10
3
1.6765x 10
5
3.6276x10
Water saturation, 1.
fraction 2 3.9935
. 38 5
x 10
Water 3 60
compressibility, psia 0
X
-1 10
2 3
6
x
1
0

Porosity, fraction
Formation
compressibility, psia 0.081 z factor at
1
- initial pressure
6
4x 10- 0.8399 z factor at
 2
BHFP 2.6095 X 10 3.1834 5 8x 10 2
4
0.9350 X 10- .5
Gas viscosity 1.3642x 103 40 9.1560 X 102
evaluated at initial 25
pressure, cp 4.4155x 10
5 5
7.7758x 10
0.0196 2
2.8752x 10 6.3506x 10
2

Gas FVF at initial

pressure, RB/Mscf 3
1.2979 X 10 2
0.8929 9.2700x 10
G
a 4.7691 X 105 8.5209x 10
5

s 3.1507 X 10 2 2
6.8914x10

c 1.2073x 103 7.1210 X 102

o
m 5
p 5.1142x 10 8.9641 X 105
2
r 3.4902x 10
e
s 3
1.1948x 10
s
i
5
b 5.5220x 10
i 3.7568x 10
2

l
i 3
t 1.2296 X 10
y
5
5.8452x 10
a 4.0699x 10
2

t
3
1.1566x10
i
n
i 5
6.2187x 10
t 2
4.2866x 10
i
a
l 3
1.0575x10

5
6.4586x 10
2
4.6685x 10

1.0405 X 103

6.8592x 10 5
p a it, Mscf/D
r g
e e
s a
s r
u e
r a
e ,
, a
c
p r
s e
i s
a E
c
- o
n
1 o
m
D i
r c
a li
i m
n
 6
7 8. 9.2682 1.4272x 10
Cumula 5
tive 34 x 10 1 3. 1.4665
730.0 6
. 53 X 10
00
Gas 428 10
Flow x x
10 10
2
2
3
7.5919
2
10408 1.7999x 10
912.5
Time 448.7 X 6
1.5154 X 10
1,095.0 521.6 390 3
6.9290x 10
2 1.8923x10
Product
1,460.0 657.9 357 1.5509x 10
6
ion Rate
1,825.0 782.9 328
5 2.1109 10
3
9.5027x 10 X
(days)
2,190.0 897.6 7.8924x 10302
2
(MMscf) 2
2,555.0 1,004.0
(Mscf/D 279 3.4600x10
2
3,102.0 1,148.4 7.7570x 10249
54.8 6
3,650.0 1,277.5 223 1.6334x 10
5 3
816 9.7234x 10201 2.1527x 10
4,197.5 1,393.5
2
72.8
4,745.0 1,498.0 8.6526x 10181
2
2.5400x 10
5,475.0
621 1,622.2 159
6.8210x10 2
6
91.3 1.6460x 10
1.0277 X 10
6
2.3312 X 103
597
2
109.5 9.6914 X 10
2
2.4990x 10
2
7.0220x 10
6
146.0 1.6910 X 10
3
1.0996x 10
6 2.4654 X 10
3
1.0497x 10
182.5 2.7660x 10
2

2
6.7930x 10 6
1.7263 X 10
3
1.1553x 10
6 2.4982 X 10
3
1.0565 X 10 2
2.2090x 10
2
6.1400x 10 6
1.7345 X 10
3
6 2.5832x 10
1.1597x 10
3
1.1445x 10
2.3910x 10 2
2
5.8080x 10
1.7540x 106
3
6 2.6923x 10
1.2123x 10
3
1.2761 X 10 2
1.9640 X 10
2
5.1940 X 10 6
1.7778x 10
3
6 2.8189x 10
1.2847x 10
3
1.3673 X 10 2
1.8470 X 10
4.6970x 102
6
1.8019 X 10
3
1.3298 X 106 2.9301 X 10

365.0 1 5. 1.3745 2
. 75 x10
6 2.1030 X 10
40
6
x1 1.8238x 10
2 3
0 3.1042x10
2
1.5599 x 10
547.5 3
X 10 2
1.8910x10

4.1080
 1.8586x

Bl = first- Pp(p,. ) pseudopressure

order evaluated at constant
coefficient BHFP, psia2/cp
1
derived from
series
DECLINE-CURVE ANALYSIS FOR GAS WELLS 229

1/ = transient-flow-period correlation parameter for 2. Campbell, J.M.: Petroleum Reservoir Property Evaluation, Campbell

Carter type curve Petroleum Series, Norman (1973).

A = boundary-dominated-flow-period correlation 3. Thompson, R.S. and Wright, J.D.: Oil Property Evaluation, Thompson

parameter for Carter type curve Wright Assocs., Golden, CO (1984).

J.Lg = gas viscosity, miLt, cp 4. Fetkovich, M.J.: "Decline Curve Analysis Using Type Curves," JPT
(June !980) 1065-77.
J.Lg( Pi) = gas viscosity evaluated at original reservoir pressure,
5. Carter, R.D. · "Type Curves for Finite Radial and Linear Gas-Flow
m/Lt, cp
Systems: Constant-Terminal Pressure Case," SPEI (Oct. 1985) 719-28.
o = fraction of 211" radius defining approximate
6. Fetkovich, M.J. et al.: ''Decline-Curve Analysis Using Type Curves
equivalent reservoir ring shape
Case Histories," SPEFE (Dec. 1987) 637-56; Trans., AIME, 283.
q, = porosity, fraction
7. Al-Hussainy, R., Ramey, H.J. Jr., and Crawford, P.B.: "The Flow
of Real Gases Through Porous Media," JPT (May 1966) 624-36;
Subscripts
Trans., AIME, 237.
i initial 8. Fetkovich, M.J. and Vienot, M.H.: "Shape Factor, CA. Expressed as
MP = match point Skin, s01," JPT(Feb. 1985) 321-22.
"1= bottomhole flowing 9. Smith, R.V.: Practical Natural Gas Engineering, PennWell Publish
ing Co., Tulsa, OK (1983).
References 10. Fraim, M.L. and Wattenbarger, R.A.: "Gas Reservoir Decline-Curve

I. Arps, J.J.: "Analysis of Decline Curves," Trans., AIME (1945) 160, Analysis Using Type Curves With Real Gas Pseudopressure and Nor

228-47. malized Time," SPEFE (Dec. 1987) 671-82.