SPE 78695

Reserve Analysis for Tight Gas

Stuart A. Cox, SPE, John V. Gilbert, SPE, Robert P. Sutton, SPE, Ronald P. Stoltz, SPE, Marathon Oil Company
Copyright 2002, Society of Petroleum Engineers Inc.
This paper was prepared for presentation at the SPE Eastern Regional Meeting held in
Lexington, Kentucky, U.S.A., 2325 October 2002.
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.

Abstract
Tight gas reservoirs present unique challenges to the reservoir
engineer. Applying classical reservoir engineering techniques
to these reservoirs is problematic due to the length of time
required to reach pseudo-steady state flow and/or establish a
constant drainage area. This leads to an inability to accurately
estimate the recoverable reserves in a timely and consistent
manner. This paper focuses on reserve estimation techniques
for tight gas reservoirs.
Both decline curves and material balance methods were
found to have serious drawbacks when applied to tight gas
reservoirs that had not established a constant drainage area.
Gas production analysis (GPA) utilizing a combination of the
decline curves and material balance in conjunction with
classical pressure transient analysis was found to be more
accurate in determining reserves.

The paper provides a critical review and explanation of

common techniques used to estimate tight gas reserves
including rate-time decline curve analysis, material balance
and the combination of type curve and decline curve analysis
commonly referred to as gas production analysis. Discussion
topics are supplemented with conceptual simulation work and
field examples. The conceptual simulation study cases are
prepared as a control set to test the accuracy of each analytical
method. Well performance will be evaluated during both
early time transient and late time boundary dominated flow
(BDF) to determine the accuracy of each technique as
compared to a known recoverable reserve and/or original gasin-place. Four field examples are included comparing the
results of applicable techniques through time.
Background Theory
The following is a brief summary of the three reserve
estimation techniques detailing their assumptions, basic theory
and implementation.
Rate-Time Decline Curve Analysis
Decline curve analysis is an extensively used methods of
estimating future performance and ultimate recoverable
reserves. The data required to perform this analysis is readily
available. This method is based on the empirical equations of
Arps1. Table 1 provides a brief summary of useful decline
curve equations2.
Table 1
Decline Curve Equations

Introduction
For the purposes of this paper, gas reservoirs with a
permeability less than 0.1 md are considered tight gas
reservoirs. When tight gas reservoirs are produced, a
significant amount of time is required before the pressure
transient is affected by all the boundaries of the reservoir,
which can often take years. This length of time is the
principle reason why the determination of reserves for tight
gas reservoirs is problematic. Tight gas reservoirs appear to be
more heterogeneous than systems with higher permeability.
Relatively small changes in permeability can result in
uneconomical flow rates from a given well while the same
percentage variations in higher permeability systems would
result in little or no change in well performance.

Exponential

q(t ) = q(i ) e Dit

Hyperbolic
(for 0<b<1)

q( t ) =

Harmonic
(for b=1)

q( t ) =

q(i )

(1+ bDi t )

1/ b

q( i )
(1 + Di t )

The decline curve

assumption that past
mathematically and
However, care must

G p (t ) =

q( i )
1
q( t ) +
Di
Di

q(bi ) (1 b )
q
G p (t ) =
q((i1) b )
D (b 1) (t )
i

G p (t ) = 2.303

q( i )
Di

[log q

(i )

log q(t )

analysis technique is based on the

performance trends can be characterized
used to predict future performance.
be taken to ensure that the technique is

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

not improperly applied. Decline curve analysis is based on the

fundamental assumption that past operating conditions will
remain unchanged. The following is a list of inherent
assumptions implied when performing rate-time decline
curve analysis:

V f = c f ( Pi P)

During transient flow, the effective drainage area of the well is

increasing and thus the decline curve analysis techniques will
not accurately predict future well performance. Decline curve
analysis should only be attempted after the well has a defined
drainage area and/or has reached a pseudo-steady state
flowing condition. The pseudo-steady state flow period has
been defined as the point in time when the pressure change
with time, dp/dt, is constant at all points in the reservoir for
constant rate production3. Theory states that the time to
pseudo-steady state can be estimated from equation 1.

t pss =

ct A
(t DA ) pss .... (1)
0.006328k

ct = S w c w + (1 S w )c g + c f .. (2)
In order to properly apply equation 1, the fluid properties and
system compressibility must be determined at the average
reservoir pressure. If equation 1 is evaluated at initial
conditions, the time to pseudo-steady state will be
underestimated. However, the assumption of initial conditions
has been found to provide a good first order approximation of
the time required to reach BDF and thus establish a constant
drainage area.
Material Balance
Material balance can be used to understand a reservoirs past
performance and predict future performance.
These
calculations are based on the premise that the void created in
the reservoir through the production of reservoir fluids is
immediately and completely filled by the expansion of the
remaining fluids and rock. Equation 3 represents the general
material balance equation for gas reservoirs including water
and rock compressibility and aquifer influx2.
GBgi = (G GP ) Bg + VW + V f + (We W p ) Bw ..(3)

= c

( PI P )

wi

GB

1 Swi

gi

...(4)

GBgi ......(5)

1 Swi

For normally pressured reservoirs with no water influx

equation 3 reduces to a volumetric depletion form given by
equation 6. The simulation and field examples in this paper
are normally pressured and are depletion drive systems.

The well is produced at or near capacity.

The wells drainage remains constant.
The well is produced at a constant bottom hole pressure.

In most cases, tight gas wells are producing at capacity and

approach a constant bottom hole pressure, if produced at a
constant line pressure. However, it can be quite challenging to
determine when a tight gas well has defined its drainage area.

SPE 78695

p pi G p
= 1
z z i
G

..(6)

The following is a list of inherent assumptions implied when

performing material balance calculations:

Reservoir hydrocarbon fluids are in phase equilibrium at

all times, and equilibrium is achieved instantaneously
after any pressure change.
Accurate fluid properties and production data are
available.
The reservoir can be represented by a single weightedaverage pressure at any time.
Fluid saturations are uniform throughout the reservoir at
any time.

In the case of tight gas, it is difficult to accurately estimate the

average reservoir pressure through time. If the pressures
obtained during shut-in do not reflect the average reservoir
pressure the resulting analysis will be inaccurate. The shut-in
time required to obtain an accurate estimate of reservoir
pressure is excessive for tight gas reservoirs due to practical
economic considerations.
In an effort to obtain a reasonable estimate of average
reservoir pressure, engineers have employed pressure build-up
analysis techniques to estimate an average reservoir pressure.
The straight line portion of the horner plot can be extrapolated
to a false reservoir pressure referred to as P*. This pressure
can the be corrected for boundary conditions to provide an
accurate estimate of the average pressure. However, in order
to obtain a meaningful estimate of P* from the pressure
transient analysis, the well must be shut-in long enough to
reach pseudo radial flow3. The resulting estimate of P* must
be corrected to an average reservoir pressure before it can be
used for material balance calculations. The shut-in time
required to reach pseudo radial flow can be estimated by using
equation 7 for a hydraulically fractured well2.

t Dxf =

0.006328kt
3 ..(7)
c t x 2f

SPE 78695

RESERVE ANALYSIS FOR TIGHT GAS

Gas Production Analysis

Use of well production data to characterize reservoir
parameters using type-curve techniques was originally
presented by Palacio and Blasingame in 19934. This paper
showed the techniques required to convert gas well production
with both varying rate and pressure into equivalent constant
rate liquid data. Constant rate type-curve solutions are well
established from traditional well test analysis allowing
variable gas production rates to be analyzed in an equivalent
manner. The linkage allowing this transformation, called
pseudo equivalent time, ta, was developed by Palacio and
Blasingame4. This was an extension of the earlier work of
Fraim and Wattenberger5.
ta =

(c )

(c )
q (t )dt
=

q(t ) ( p )c ( p )
q(t )
g i

g i

z i Gi
[m( p )] .(8)
2 pi

The average reservoir pressure must be determined to

calculate the correct pseudo equivalent time. From material
balance the average reservoir pressure can be calculated if the
original gas in place is known. Therefore gas production
analysis is inherently iterative in nature, coupling material
balance, decline curves and traditional pressure transient
techniques into a more comprehensive tool. Because of this
combination of techniques gas production analysis sidesteps
many of the limitations of these individual techniques if used
in isolation.
GPA doesnt require that a well be produced at constant
bottom hole pressure or that it be produced at capacity.
However, the technique does require that flowing bottom hole
pressure be measured or more usually calculated from tubing
head pressure. GPA uses pressure transient techniques which
are not limited to the BDF region. The type curve analysis of
the transient flow period can be used to infer reservoir
parameters commonly only available after a lengthy
buildup test.
Various plots are available to determine the original gas in
place (OGIP) using gas production analysis. Agarwal used a
non-dimensionalized variation of the Rate/Cumulative
production plot by graphing reciprocal dimensionless wellbore
pressure, 1/PwD, against dimensionless cumulative production,
(QDA)6,7. This plot has the benefit of tending to a value of QDA
= 1/2 during BDF, regardless of permeability. It decreases
steeply during the transient period then once BDF is reached
decreases monotonically towards the anchor point at 1/2.
Dimensionless cumulative production is defined as shown6.

QDA =

4.5Tz i Gi m( p )
t DA
=
. (9)
p wD
hApi m( p )

p wD =

khm( p )
. (10)
1422Tq (t )

t DA =

0.006328kt
... (11)
(ct )i A

This equation will only equal 1/2 once the average pressure
in the system equals the bottom hole flowing pressure.
The assumptions of gas production analysis are the same as
those of traditional pressure transient analysis. The accuracy
and frequency with which production and pressure data is
gathered is the major limitation on the successful application
of production analysis. Insufficient rate and pressure data
though the transient period limits the accuracy of the analysis.
To correctly generate the time function requires that the
volumetric gas material balance equation be valid. This will
not be the case for strong water drives, changing fluid
saturations and non-sealing boundaries.
Simulation Cases
Two simulation cases were constructed and used as a control
set to test the accuracy of each analytical method. Each case
represents a single well system producing from a single layer.
The parameters that remained constant for both cases are
summarized in Table 2. Production forecasts from each model
were based on a constant flowing tubing pressure of 100 psia.
Reserve estimates were made after ninety days, one hundred
eighty days, one year, three years and after the well had
established BDF from decline curve techniques and the gas
production analysis technique. A minimum rate of 30 Mcf/D
or a maximum productive life of thirty years limited the
recoveries projected by each method.
In order to evaluate the material balance technique each model
was produced at a constant tubing pressure of 100 psia for
three years. Each model was shut-in once a year for 24 hours,
7 days and finally for a sufficient time to reach the average
reservoir pressure. The following is a summary of the results
for each simulation case.
Simulation Case 1
Simulation case 1 represents a single well producing from the
center of a 640-acre square reservoir. The model was
initialized with a gas in place of 16,415 MMscf. The ultimate
recovery, based on an abandonment rate of 30 Mcf/D,
predicted by the simulator was 13,859 MMcf after 212 years
of production. Producing time to reach BDF was estimated
from equation 1 for this system and found to be 1,807 days
assuming rock and fluid properties at initial conditions.
Table 3 summarizes the results of the rate-time decline curve
analysis and the production analysis techniques for this
system. The simulator projects a thirty-year recovery of 6,667
MMcf. The exponential and harmonic declines underestimate
recoverable reserves, while the hyperbolic decline
overestimates the recovery projected by the simulator. The

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

b factor for the hyperbolic decline was found to be greater

than 1 for each case. In each case, the wells production
performance appears to have established a reasonable decline.
However, the resulting forecasts could not accurately predict
future recoverable reserves. This is due to the fact that the
system has not established a constant drainage area. A final
decline curve analysis was performed on the data set after the
well had produced beyond the time required to reach BDF.
The resulting forecast can be fit with an annual exponential
decline rate of 4.2%. The resulting forecast projects a
recoverable reserve of 6,757 MMcf, which is approximately
equal to the simulation results. Figures 1 through 4 present
graphically the results of the decline curve analysis.
Table 2
Model Parameters
Formation Top, ft
Initial Reservoir Pressure, psi
Average Porosity, %
Net Pay, ft

10,000
5,000
10.0
40

Average Water Saturation, %

42.8

Gas Specific Gravity

0.65

Water Specific Gravity

1.05

Reservoir Temperature, F

225

Water Compressibility, 1/psi

3E-6

Rock Compressibility, 1/psi

3E-6

Effective Gas Perm. Md

0.05

Fracture Half Length, ft

200

Fracture Conductivity, md-ft

500

Gas production analysis results are consistently higher than

results obtained from exponential decline techniques. The
results of the 180-day case are presented as Figures 5 and 6.
Figure 5 is the type curve match of the daily production data.
The type curve match indicates that this well has not reached
BDF. The resulting gas in place shown on Figure 6 represents
a minimum drained volume, as the effective drained area in
the system will continue to increase until BDF is established.
Once the well has reached boundary dominated flow, the
GPA technique will result in a gas in place estimate consistent
with the simulation results from which a recoverable volume
can then be estimated. Figures 7 and 8 show the final gas
production analysis results.
The values included in Table 3 for the GPA analysis represent
the effective drained volume for the first three cases and a
thirty-year projected recovery for the last two estimates.

SPE 78695

Table 3
Time
Exponential Hyperbolic Harmonic
GPA
Days
MMscf
MMscf
MMscf
MMscf
90
596
7,016
1,773
1,120
180
1,109
7,325
2,267
3,020
365
2,119
7,576
3,598
6,090
1,095
5,569
7,775
5,061
6,158
After BDF
6,757
Na
Na
6,694
Simulation 30 year recovery 6,667 MMscf
The material balance calculations are summarized graphically
on Figure 29. These results highlight the importance of
ensuring that an accurate average pressure is obtained before
attempting to perform material balance calculations. The P/z
analysis was performed by placing a best-fit line through the
three data points, ignoring the initial pressure. The results for
the 24 hour shut-in case and the 7 day case both project an inplace volume of 6.0 BCF. The shut-in time required for the
well bottom hole pressure to equal the average reservoir
pressure ranged from 19 to 21 years.
Simulation Case 2
Simulation case 2 represents a single well producing from the
center of a 600 feet wide channel draining a 160 acre
reservoir. The model was initialized with an initial gas in
place of 4,112 MMcf. The ultimate recovery, based on an
abandonment rate of 30 Mscf/D, predicted by the simulator
was 2,807 MMcf after 81 years of production. Producing
time to reach BDF estimated from equation 1 was 5,467 days,
assuming rock and fluid properties at initial conditions.
Table 4 summarizes the results of the rate-time decline curve
analysis and the production analysis techniques for this
system. The model projects a thirty-year recovery of 1,892
MMscf. The exponential and harmonic declines underestimate
recoverable reserves, while the hyperbolic decline approaches
the actual recoverable reserves if the forecast is limited to a
thirty-year time period. The b factors for the hyperbolic
decline ranged from 1 and 2 for this case. In each case, the
wells production performance appears to have established a
reasonable decline. However, the resulting forecasts could not
accurately predict the wells recoverable reserves. A final
decline curve analysis was performed on the data set after the
well had produced beyond the time required to reach BDF.
The resulting forecast can be fit with an annual exponential
decline rate of 3.3% and projects a recoverable reserve of
2,367 MMscf, which is 25% above the simulation results.
Figures 9 through 12 present graphically the results of the
decline curve analysis for selected cases.
GPA results are consistently higher than the results obtained
from an exponential decline technique. The results of the 180day case are presented as Figures 13 and 14. Figure 13 is the
type curve match of the daily production data. The type curve
match indicates that this well is in linear flow and has not

SPE 78695

RESERVE ANALYSIS FOR TIGHT GAS

established a constant drainage area. The resulting gas-inplace shown on Figure 14 represents a minimum drained
volume. The effective drained area in the system will
continue to increase until boundary dominated conditions have
been established. Once the well has established a constant
drainage area, the gas production analysis technique will result
in a gas in place estimate consistent with the simulation
results, from which a recoverable volume can then be
estimated. Figures 15 and 16 show the final GPA results.
The values included in Table 4 for the GPA analysis represent
the effective drained volume for the first four cases and a
thirty-year projected recovery based on the final analysis.
Table 4
Time
Exponential Hyperbolic Harmonic
GPA
Days
MMscf
MMscf
MMscf
MMscf
90
130
1,540
862
480
180
448
1,720
942
680
365
630
1,794
1,073
950
1,095
1,132
1,867
1,509
1,650
After BDF
2,367
Na
Na
1,877
Simulation 30 year recovery 1,892 MMscf
The material balance calculations are summarized graphically
on Figure 30. These results highlight the importance of
ensuring that an accurate average pressure be obtained before
attempting to perform material balance calculations. The P/z
analysis was performed by placing a best-fit line through the
three data points, ignoring the initial reservoir pressure. The
results for the 24-hour case and the 7 day shut-in cases both
project an in-place volume of 1.2 BCF (29% of the actual gas
in-place). The shut-in time required for the wells bottom hole
pressure to equal the average reservoir pressure ranged from
115 to 138 years.
Field case 1
This well is producing from a sandstone reservoir at a depth of
7,700 feet with an average porosity of 8%, water saturation of
40% and net pay of 94 feet. The initial reservoir pressure was
4,060 psia. The well was fracture stimulated upon initial
completion and had an initial production rate of approximately
5.0 MMscf/D. Transient analysis indicates an effective gas
permeability of 0.03 md and an effective fracture half-length
of 160 feet. This well has a cumulative production of 360
MMscf after approximately one year of production.
Table 5 summarizes the results of the rate-time decline curve
analysis and the production analysis techniques for this well.
Ultimate recoveries from decline curve analysis techniques
were limited by an economic limit of 30 Mscf/D or thirtyyears of production. The b factors for the hyperbolic decline
ranged between 1.89 and 0.64. Figure 17 presents graphically
the results of the 322-day decline curve analysis.

Time
Days
90
180
322

Table 5
Exponential Hyperbolic
MMscf
MMscf
282
2,077
400
Na
400
516

Harmonic
MMscf
785
664
685

GPA
MMscf
400
580
650

GPA results are consistently higher than the results obtained

from an exponential decline projection. The results of the
322-day case are presented as Figure 18 and 19. Figure 18 is
the type curve match of the daily production data. The type
curve match indicates that this well is producing from a
bounded system approximately 400 feet wide and has not
established BDF. Therefore the drainage volumes estimated
for this well will continue to increase with time. Based on the
modeling results, it may be appropriate to consider the entire
drained volume as the ultimate recovery for this well at
this time.
Field case 2
This well is producing from a fractured limestone reservoir at
a depth of 18,000 feet with an average porosity of 10%, water
saturation of 25% and 150 feet of net pay.
The initial
reservoir pressure was 8,650 psia. The well was fracture
stimulated upon initial completion and had an initial
production rate of approximately 13.0 MMscf/D. Transient
analysis indicates an effective gas permeability of 0.04 md and
an effective fracture half-length of 390 feet. This well
has a cumulative production of 4.2 BCF after 3.7 years
of production.
Table 6 summarizes the results of the rate-time decline curve
analysis and the production analysis techniques for this well.
Ultimate recoveries from the decline curve analysis techniques
were limited by an economic limit of 30 Mscf/D or thirty
years of production. The b factors for the hyperbolic decline
ranged between 1.56 and 1.38. Figure 20 presents graphically
the results of the 1,374-day decline curve analysis.
GPA results are consistently higher than results from the
exponential decline technique. The results of the 1,374-day
case are presented as Figures 21 and 22. Figure 21 is the type
curve match of the daily production data. The type curve
match indicates that this well has reached boundary dominated
flow. The final gas production estimate shown in Table 6 is
based on a forecast of production to an abandonment rate of
30 Mscf/D after 32 years of production.

Time
Days
90
180
365
1,374

Table 6
Exponential Hyperbolic
MMscf
MMscf
2,701
Na
2,845
Na
4,304
12,315
5,401
8,216

Harmonic
MMscf
8,363
7,208
4,085
7,595

GPA
MMscf
3,670
4,630
6,060
7,270

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Several static pressures have been obtained from this well.

The shut-in periods ranged from 24 hours to 27 days. The
resulting material balance plot is included as Figure 31.
Figure 31 demonstrates that the static pressures measured the
average pressure.
The results of the material balance
calculations range from 2 to 7 BCF depending on the points
chosen for the analysis. The line shown is a least squares fit
of all the available data.
Field case 3
This well is producing from a sandstone reservoir at a depth of
14,400 feet with an average porosity of 8%, water saturation
of 40% and net pay of 130 feet. The initial reservoir pressure
was 11,200 psia. The well was fracture stimulated upon initial
completion and had an initial production rate of approximately
9.0 MMscf/D. Transient analysis indicates an effective gas
permeability of 0.03 md and an effective fracture half-length
of 55 feet. This well has a cumulative production of 2.55 BCF
after 3.8 years of production.
Table 7 summarizes results of the rate-time decline curve
analysis and the gas production analysis technique for this
well. Ultimate recoveries from the decline curve analysis
techniques were limited by an economic limit of 30 Mscf/D or
thirty years of production. The b factor for the hyperbolic
decline was found to be 1.69 after three years of production.
Figure 23 presents graphically the results of the 1,399-day
decline curve analysis.

Time
Days
90
180
365
1,399

Table 7
Exponential Hyperbolic Harmonic
MMscf
MMscf
MMscf
1,242
Na
4,115
1,680
Na
4,154
2,446
Na
5,596
4,354
7,837
6,601

GPA
MMscf
2,500
4,260
5,510
6,081

GPA results are consistently higher than results obtained from

the exponential decline technique. The results of the 1,399day case are presented as Figures 24 and 25. Figure 24 is the
type curve match of the daily production data. The type curve
match indicates that this well has reached boundary dominated
flow. The final gas production estimate is based on a forecast
of production to an abandonment rate of 30 Mscf/D after 53
years of production.
Two static pressures were obtained from this well as a result
of pressure build up tests. The shut-in periods ranged from
one to eight days. The results of the second pressure build up
are included as Figure 32. A review of the pressure transient
analysis indicates that the well was not shut-in long enough to
reach pseudo radial flow. Therefore a P* could not be
obtained from the tests. For comparison purposes however, a
material balance P/z analysis was performed on the available
shut-in pressures. The results of the material balance

SPE 78695

calculations range from 1.5 to 5 BCF depending on the points

chosen for the analysis. The line shown is a least squares fit
of all the available data.
Field case 4
This well is producing from a sandstone reservoir at a depth of
10,500 feet with an average porosity of 9.5%, water saturation
of 35% and net pay of 81 feet. The initial pressure was 4,000
psia. The well was fracture stimulated upon initial completion
and had an initial production rate of approximately 1.5
MMscf/D. Transient analysis indicates an effective gas
permeability of 0.08 md and an effective fracture half-length
of 55 feet. This well has a cumulative production of 649
MMscf after 1.4 years of production.
Table 8 summarizes the results of the rate time decline curve
analysis and the gas production analysis technique for this
well. Ultimate recoveries from the decline curve analysis
techniques were limited by an economic limit of 30 Mscf/D or
thirty-years of production. The b factors for the hyperbolic
decline were found to be greater than 2.0 for all three cases.
Figure 26 presents the results of the 490-day decline curve
analysis graphically.

Time
Days
90
180
490

Table 8
Exponential Hyperbolic
MMscf
MMscf
556
4,871
932
3,599
1,150
4,261

Harmonic
MMscf
1,612
1,629
2,447

GPA
MMscf
870
1,230
2,200

GPA results are consistently higher than results obtained from

the exponential decline technique. The results of the 490-day
case are presented as Figures 27 and 28. Figure 27 is the type
curve match of the daily production data. The type curve
match indicates that the well has not reached boundary
dominated flow conditions; therefore the drainage volumes
estimated for this well will continue to increase with time.
Based on the modeling results, it may be appropriate to
consider the entire drained volume as the ultimate recovery for
this well at this time.
Conclusions
1.

2.

3.

Simulation cases establish a benchmark which

demonstrates that the time to reach boundary dominated
flow should be considered when performing reserve
estimates.
Decline curve analysis after the well has reached
boundary dominated flow may provide a reasonable
estimate of recoverable reserves. Prior to establishing a
constant drainage area this method is inaccurate. The
inaccuracies are more severe for linear systems even after
boundary dominated flow had been reached.
Material balance techniques should not be performed on
tight gas reservoirs due to the inability to accurately

SPE 78695

4.

5.

RESERVE ANALYSIS FOR TIGHT GAS

estimate average reservoir pressure, which is a basic

requirement of this technique.
Gas production analysis combines decline curve analysis,
material balance and pressure transient analysis
techniques into a comprehensive method which can be
used to accurately predict reserves during boundary
dominated flow. Prior to boundary dominated flow this
technique was found to be more technically correct.
An accurate estimate of recoverable reserves for tight gas
reservoirs can only be obtained after a well has reached
boundary dominated flow.

Acknowledgments
We thank the management of Marathon Oil Company for
permission to print this article. Thanks to Emery Petrof III
for constructing the simulation cases. Acknowledgment is due
to various colleagues for providing production data for the
field cases.
Nomenclature
A = drainage area, ft
b = Arps decline curve constant
Bg = gas formation volume factor, RB/Mscf

Vf = change in formation reservoir volume, res bbl

Vw = change in water volume, res bbl
We = cumulative water influx volume , STB
Wp = cumulative water influx volume , STB
xf = fracture half length, feet
z i = gas compressibility factor at pi
z = gas compressibility factor at p
= porosity, fraction
= viscosity, cp
= 3.14159
Subscripts
i = initial
pss = pseudo steady state
References
1.
2.
3.
4.

c f = formation compressibility, psi-1

cg = gas compressibility, psi-1

cw = water compressibility, psi-1

ct = total compressibility, psi-1
Di = initial decline rate, day-1
G = original gas in place, Mscf
G p = cumulative gas produced, Mscf
h = reservoir thickness, ft
k = effective permeability to gas, md
m( p) = real gas pseudo pressure, psi2/cp
m( p ) = m( pi ) m( p ) , psi2/cp
m( p) = m ( pi ) m ( pwf ) , psi2/cp
pwf = bottomhole producing pressure, psia

p = average reservoir pressure, psia

pi = initial reservoir pressure, psia
p wD = dimensionless wellbore pressure
p* = false reservoir pressure, psi
q(t ) = flow rate, Mscf/D
Q(t ) = cumulative production, Mscf
Q DA = dimensionless cumulative production based on
area (A)
S w = water saturation, fraction
t = time, days
T = reservoir temperature, R
tDxf = dimensionless time based on fracture half length
t a = pseudoequivalent time, days
tpss = time to pseudo steady state, days
t DA = dimensionless time based on area (A)

5.

6.

7.

Arps, J.J. :Analysis of Decline Curves, Trans.,

AIME(1945) 160,
228-47.
Lee, W.J. and Wattenbarger, R.A.: Gas Reservoir
Engineering, SPE Textbook Series Volume 5 (1996).
Earlougher, R.C. Jr.:Advances in Well Test Analysis,
Monograph Series, SPE, Richardson, TX (1977) 5.
Palacio, J.C. and Blasingame, T.A.: Decline-Curve
Analysis Using Type Curves Analysis of Gas Well
Performance Data, paper SPE 25909 presented at the 1993
Rocky Mountain Regional Meeting/Low Permeability
Reservoirs Symposium and Exhibition, Denver,
26-28 April.
Fraim, M.L. and Wattenbarger, R.A.: Gas Reservoir
Decline Curve Analysis Using Type Curves with Real
Gas Pseudopressure and Normalized Time, SPEFE (Dec
1987) 671-82.
Agarwal, R.G., Gardner, D.C., Kleinsteiber, S.W. and
Fussell, D.D.: Analyzing Well Production Data Using
Combined-Type-Curve and Decline-Curve Analysis
Concepts, SPEREE (October 1999) 478.
Gardner, D.C, Hager, C.J. and Agarwal, R.G:
Incorporating Rate-Time Superposition Into Decline Type
Curve Analysis, paper SPE 62475 presented at the 2000
Rocky Mountain Regional Meeting/Low Permeability
Reservoirs Symposium, Denver, 12-15 March.

Gas Rate (Mscf/D)

Gas Rate (Mscf/D)

Jan-03

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Harmonic b=1.0, Di=0.85, Eur=2,267 MMscf

Hyperbolic b=5.14 Di=1.56 EUR=7,325 MMscf

Exponential, Di =0.53 EUR=1,109 MMscf

Simulation Data

180 Days of Production

Jan-03

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Harmonic b=1.0, Di=0.15, Eur=5,061 MMscf

Hyperbolic b=5.84 Di=2.04 EUR=7,775 MMscf

Exponential, Di =0.06 EUR=5,569 MMscf

Simulation Data

1095 Days of Production

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Jan-03

Jan-04

Jan-05

Jan-06

Jan-07

Time to BDF

Jan-08

Jan-09

Exponential, Di =0.042 EUR=6,757 MMscf

Simulation Data

Decline Estimate after Reaching BDF

Fig. 4 Simulation Case 1 Post BDF decline curve

100
Jan-02

1000

10000

Jan-03

Harmonic b=1.0, Di=0.32, Eur=3,598 MMscf

Hyperbolic b=5.49 Di=1.65 EUR=7,576 MMscf

Exponential, Di =0.23 EUR=2,119 MMscf

Simulationl Data

365 Days of Production

Fig. 2 Simulation Case 1 365 days decline curve

100
Jan-02

1000

10000

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Fig. 3 Simulation Case 1 1,095 days decline curve

100
Jan-02

1000

10000

Fig. 1 Simulation Case 1 180 days decline curve

100
Jan-02

1000

10000

Gas Rate (Mscf/D)

Gas Rate (Mscf/D)

8
SPE 78695

0.001
tDA

0.01

0.1

10

0.001
tDA

0.01

0.1

10

100

100

0.04

0.06

QDA

0.08

0.1

0.12

0.14

Simulation
Analytical

0.16

0.02

0.04

0.06

QDA

0.08

0.1

0.12

0.14

Simulation
Analytical

Gas in Place = 16.23 BCF - Equivalent Area 638 acres - Post BDF

0.16

Fig. 6 Simulation Case 1 Rate Cumulative Decline Curve

0.02

Gas in Place = 3.02 BCF - Equivalent Area 119 acres - 180 Days

Fig. 8 Simulation Case 1 Rate Cumulative Decline Curve

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

RESERVE ANALYSIS FOR TIGHT GAS

Fig. 7 Simulation Case 1 GPA Type-curve match

0.0001

Simulation pwD
Simulation pwD'
Analytical pwD
Analytical pwD'

Infinite Conductivity Fracture in 1 to 1 Rectangular Boundary Post BDF

0.01
0.000001 0.00001

0.1

10

100

Fig. 5 Simulation Case 1 GPA Type-curve match

0.0001

Simulation PwD
Simulation PwD'
Analytical PwD
Analytical PwD'

Infinite Conductivity Fracture in 1 to 1 Rectangular Boundary at 180 Days

0.01
0.000001 0.00001

0.1

10

100

1/PwD

1/PwD

PwD or PwD'

PwD or PwD'

SPE 78695
9

Gas Rate (Mscf/D)

Gas Rate (Mscf/D)

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Jan-03

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Harmonic b=1.0, Di=0.36, EUR=1,509 MMscf

Hyperbolic b=1.90, Di=3.47, EUR=1,867 MMscf

Exponential, Di =0.17, EUR=1,132 MMscf

Simulation Data

1095 Days of Production

Jan-03

Jan-04

Jan-05

Jan-06

Jan-07

Jan-08

Jan-09

Harmonic b=1.0, Di=1.94, EUR=1,073 MMscf

Hyperbolic b=1.82, Di=3.11, EUR=1,794 MMscf

Exponential, Di =0.59, EUR=630 MMscf

Simulation Data

365 Days of Production

Jan-04

Jan-06

Jan-08

Jan-10

Jan-12

Jan-14

Jan-16

Jan-18

Time to Pseudo-Steady State

Exponential, Di =0.033 EUR=2,367 MMscf

Simulation Data

Decline Estimate after Reaching Pseudo-Steady State

Fig. 12 Simulation Case 2 Post BDF decline curve

10
Jan-02

100

1000

10000

Fig. 10 Simulation Case 2 365 days decline curve

100
Jan-02

1000

10000

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Fig. 11 Simulation Case 2 1,095 days decline curve

100
Jan-02

1000

10000

Jan-03

Harmonic b=1.0, Di=2.39, EUR=942 MMscf

Hyperbolic b=1.75, Di=3.03, EUR=1,720 MMscf

Exponential, Di =1.18, EUR=448 MMscf

Simulation Data

180 Days of Production

Fig. 9 Simulation Case 2 180 days decline curve

100
Jan-02

1000

10000

Gas Rate (Mscf/D)

Gas Rate (Mscf/D)

10
SPE 78695

0.001
TDA

0.01

0.1

10

0.001
tDA

0.01

0.1

10

Fig. 15 Simulation Case 2 GPA Type-curve match

0.0001

Simulation PwD
Simulation PwD'
Analytical PwD
Analytical PwD'

Finite Conductivity Fracture in 18.6 to 1 Rectangular Boundary Post BDF

0.01
0.000001 0.00001

0.1

10

100

0.0001

Simulation PwD
Simulation PwD'
Analytical PwD
Analytical PwD'

Fig. 13 Simulation Case 2 GPA Type-curve match

0.01
0.000001 0.00001

0.1

10

100

Infinite Conductivity Fracture in 3.1 to 1 Rectangular Boundary at 180 Days

100

100

1/PwD

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.02

0.04

0.06

QDA

0.08

0.1

0.12

0.14

0.16

Simulation
Analytical

0.02

0.04

0.06

QDA

0.08

0.1

0.12

Gas in Place = 4.06 BCF - Equivalent Area 160 acres - Post BDF

0.14

Simulation
Analytical

0.16

Fig. 14 Simulation Case 2 Rate Cumulative Decline Curve

Gas in Place = 0.68 BCF - Equivalent Area 27 acres - 180 Days

RESERVE ANALYSIS FOR TIGHT GAS

Fig. 16 Simulation Case 2 Rate Cumulative Decline Curve

1/PwD

PwD or PwD'

PwD or PwD'

SPE 78695
11

Gas Rate (Mscf/D)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Jan-03

Jan-04

Jan-05

Jan-06

0.02

0.04

0.06
QDA

0.08

0.1

0.12

Gas in Place = 0.65 BCF - Equivalent Area 12 acres - 322 Days

0.14

0.16

Jan-07

Simulation
Analytical

Fig. 17 Field Case 1 Decline Curve Analysis

Jan-02

Harmonic b=1, Di=1.76, EUR=685 MMscf

Hyperbolic b=0.64, Di=1.56, EUR=516 MMscf

Exponential Di =1.19, EUR=400 MMscf

Field Data

322 Days of Production

0.001

tDA

0.01

0.1

10

Fig. 18 Field Case 1 GPA Type-curve match

0.0001

Actual PwD
Actual PwD'
Analytical PwD
Analytical PwD'

Uniform Flux Fracture in 3 to 1 Rectangular Boundary at 322 Days

0.01
0.000001 0.00001

0.1

10

100

100

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Fig. 19 Field Case 1 Rate Cumulative Decline Curve

10
Jan-01

100

1000

10000

1/PwD

12
SPE 78695

PwD or PwD'

Gas Rate (Mscf/D)

1/PwD

0.0

1.0

2.0

3.0

4.0

Jan-00

Jan-01

Jan-02

Jan-03

Jan-04

Jan-05

Jan-06

0.02

0.04

0.06
QDA

0.08

0.1

0.12

Gas in Place = 7.65 BCF - Equivalent Area 48 acres - 1374 Days

0.14

Simulation
Analytical

Fig. 20 Field Case 2 Decline Curve Analysis

Jan-99

Harmonic b=1, Di=0.34, EUR= 7,595 MMscf

0.16

Jan-07

Hyperbolic b=1.38, Di=0.37, EUR=8,216 MMscf

Exponential Di =.327, EUR=5,401 MMscf

Field Data

1374 Days of Production

Fig. 22 Field Case 2 Rate Cumulative Decline Curve

100
Jan-98

1000

10000

100000

0.01
0.0001

0.1

10

100

1000

0.01

tDA

0.1

10

Fig. 21 Field Case 2 GPA Type-curve match

0.001

Actual PwD
Actual PwD'
Analytical PwD
Analytical PwD'

Infinite Conductivity Fracture in 3 to 1 Rectangular Boundary 1374 Days

100

SPE 78695
RESERVE ANALYSIS FOR TIGHT GAS
13

PwD or PwD'

Gas Rate (Mscf/D)

0.0

1.0

Jan-99

Jan-00

Jan-01

Jan-02

Jan-03

Jan-04

Jan-05

0.02

0.04

0.06
QDA

0.08

0.1

0.12

Gas in Place = 6.64 BCF - Equivalent Area 64 acres - 1399 Days

Jan-06

0.14

0.16

Jan-07

Simulation
Analytical

Fig. 23 Field Case 3 Decline Curve Analysis

Jan-98

Harmonic b=1, Di=0.32, EUR= 6,601 MMscf

Hyperbolic b=1.69, Di=0.43, EUR=7,837 MMscf

Exponential, Di =.221, EUR=4,354 MMscf

Field Data

1399 Days of Production

0.001

0.01

tDA

0.1

10

Actual PwD
Actual PwD'
Analytical PwD
Analytical PwD'

Finite Conductivity Fracture in 6 to 1 Rectangular Boundary at 1399 Days

Fig. 24 Field Case 3 GPA Type-curve match

0.10
0.0001

1.00

10.00

100.00

100

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Fig. 25 Field Case 3 Rate Cumulative Decline Curve

100
Jan-97

1000

10000

1/PwD

14
SPE 78695

PwD or PwD'

Gas Rate (Mscf/D)

0.0

0.5

1.0

Jan-02

Jan-03

Jan-04

Jan-05

Jan-06

0.02

0.04

0.06
QDA

0.08

0.1

0.12

Gas in Place = 2.20 BCF - Equivalent Area 43 acres - 490 Days

0.14

0.16

Jan-07

Simulation
Analytical

Fig. 26 Field Case 4 Decline Curve Analysis

Jan-01

Harmonic b=1, Di=0.77, EUR= 2,447 MMscf

Hyperbolic b=2.33, Di=5.17, EUR=4,261 MMscf

Exponential, Di =0.528, EUR=1,150 MMscf

Field Data

490 Days of Production

Fig. 28 Field Case 4 Rate Cumulative Decline Curve

100
Jan-00

1000

10000

1/PwD

0.01
0.000001

0.1

10

100

0.0001

0.001

tDA

0.01

0.1

10

Fig. 27 Field Case 4 GPA Type-curve match

0.00001

Actual PwD
Actual PwD'
Analytical PwD
Analytical PwD'

Uniform Flux Fracture in 5.5 to 1 Rectangular Boundary at 490 Days

100

SPE 78695
RESERVE ANALYSIS FOR TIGHT GAS
15

PwD or PwD'

P/z (psia)

P/z (psia)

6000

8000

10000

12000

14000

16000

18000

3
Gp (BCF)

1000

1500

2500
Gp (Bcf)

2000

3000

3500

4000

2
Gp (BCF)

8 day shut-in

Fig. 32 Field Case 3 Material Balance

1 day shut-in

Intial

P/z Material Balance - Field Case 3

Fig. 30 Simulation Case 2 Material Balance

500

Initial Pressure
Simulation Average Pressure
7-day final shut-in pressure
1-day final shut-in presure

P/z Material Balance - Simulation Case 2

4500

S. COX, J. GOLBERT, R. SUTTON, R. STOLTZ

Fig. 31 Field Case 2 Material Balance

27 day shut-in

4000

14 day shut-in

8 day shut-in

6000

8000

2000

2 day shut-in

1 day shut-in

Initial

2000

4000

6000

8000

P/z Material Balance - Field Case 2

Fig. 29 Simulation Case 1 Material Balance

Gp (Bcf)

4000

2000

1000

1000

3000

4000

5000

6000

2000

Initial Pressure
Simulation Average Pressure
7-day final shut-in pressure
1-day final shut-in presure

P/z (psia)

2000

3000

4000

5000

6000

P/z Material Balance - Simulation Case 1

P/z (psia)

16
SPE 78695