SPE 159172

Petrophysical Properties of Unconventional Low-Mobility Reservoirs (Shale

Gas and Heavy Oil) by Using Newly Developed Adaptive Testing Approach
Hamid Hadibeik, The University of Texas at Austin; Dingding Chen, Mark Proett, Sami Eyuboglu, Halliburton
Energy Services; Carlos Torres-Verdn, The University of Texas at Austin
Copyright 2012, Society of Petroleum Engineers

This paper was prepared for presentation at the 2012 SPE Annual Technical Conference and Exhibition held in San Antonio, Texas, USA, 810 October 2012.

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 have not been
reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect any position of the Society of Petroleum Engineers, its
officers, or members. Electronic reproduction, distribution, or storage of any part of this paper 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 SPE copyright.


Abstract
Pressure testing in very-low-mobility reservoirs is challenging with conventional formation-testing methods. The main
difficulty is the over-extended buildup times required to overcome wellbore and formation storage effects. Possible wellbore
overbalance or supercharge are additional complicating factors in determining reservoir pressure. This paper addresses the
above technical complications and estimates petrophysical properties of low-mobility formations using a newly developed
adaptive testing approach.
The adaptive testing approach employs an automated pulse-testing method for very-low-mobility reservoirs and uses
short drawdowns and injections followed by short pressure stabilization periods. Measured pressure transients are used in an
optimized feedback loop to automatically adjust subsequent drawdown and injection pulses in order to reach a stabilized
pressure as quickly as possible.
The automated pulse data is used to determine supercharge effects, formation pressure and mobility via analytical models
by analyzing the entire pressure sequence. A genetic algorithm estimates additional reservoir parameters, such as porosity
and viscosity, and confirms results obtained with analytical models (reservoir pressure and permeability). The modeled
formation pressure exhibits less than 1% difference with respect to true formation pressure, while the accuracy of other
parameters depends on the number of unknown properties. As a faster method to estimate reservoir properties, a direct neural
network regression of pulse-testing data was also investigated.
Synthetic reservoir models for low-mobility formations (M <1 D/cp) which included the dynamics of water- and oil-
base mud-filtrate invasion that produce wellbore supercharging were developed. These reservoir models simulated the pulse-
testing methods, including an automated feedback optimization algorithm that reduces the testing times in a wide range of
downhole conditions. The reservoir models included both simulations of underbalanced and overbalanced drilling conditions
and enabled the development of new field testing strategies based on a priori reservoir knowledge. The synthetic modeling
demonstrates the viability of the new pulse-testing method and confirms that difficult properties, such as supercharging, can
be estimated more accurately when coupled with the new inversion techniques.

Introduction
Formation pressure is a fundamental key to assess the hydrocarbon yield of a reservoir. Without an estimate of the formation
pressure, there is a great deal of uncertainty in a fields development and the investment required. Virtually all the methods
used to calculate the net amount of recoverable hydrocarbon are highly dependent on the initial formation pressure (Snyder
1971; Sullivan et al. 1988; Mason 1987; Bennett et al. 1975). Field development optimization also depends on formation
pressure estimates to verify reservoir depletion and delineate the producing intervals connectivity.
There have been attempts to find the fundamental properties of tight sand, shale gas, and heavy oil reservoirs (Dastidar et
al. 2007; Abu Omokaro et al. 2011; Shabro et al. 2011; Kundert et al. 2009; Galford et al. 2000). However, rarely reported in
literature is a study on the pressure transient analysis methods applied to packer and probe-type formation testing for these
types of reservoirs yielding the true formation pressure. When a typical drawdown and buildup test is applied, the pressure
transient takes too much buildup time to resolve using conventional analysis or a history match to be of practical value in
these very-low-mobility reservoirs. With the introduction of the unconventional automated pulse-test method for low-
mobility formations (Hadibeik et al. 2012), it is possible to obtain a pressure response that can be used to determine the initial
reservoir pressure and permeability in a practical time frame, usually less than 1 hour. The pressure transient analysis can
2 SPE 159172
follow the same concept as a normal well-test analysis (Rees et al. 2011; Proett et al. 1994; Nakano et al. 2009).
In addition, the test results can be further analyzed with optimization method and inverse algorithm to yield more
information about the reservoir properties (Torres-Verdn et al. 2004; Angeles et al. 2007; Angeles et al. 2010; Alpak et al.
2002; Elshahawi et al. 2008). However, these methods are usually time-consuming and cannot be used when the number of
unknowns becomes large. To reduce the computation time in inverse processing, several methods can be used to speed up the
CPU time calculations such as the streamline method (Hadibeik et al. 2011). To optimize a large vector of parameters, a
genetic algorithm is used to match the pulse-test feature and obtain the reservoir and in-situ fluid properties. Finally, the
automated pulse-test technique incorporated with the optimization method can resolve the reservoir properties in tight
formations.

Problem Statement
There are two main difficulties in low-mobility testing. First, the pressure response is so slow in these reservoirs, which takes
several hours or days for a pressure buildup to become sufficiently stabilized to detect infinitely acting radial flow and verify
formation pressure. In many cases, the pressure drawdown is difficult to control and results in a phase change causing multi-
phase flow in the tool flow lines, making it harder or even impossible to reach the stabilized pressure. Second, the mud-
filtrate invasion in the near well bore region creates the supercharged effect, which causes the stabilized pressure to be
different from the true initial reservoir pressure (Figure 1). Hadibeik et al. 2009 studied in depth the effect of dynamic mud-
filtrate invasion in different filtrate invasion scenarios. The supercharging problem usually is not a concern in the reservoirs
with higher Mobilities (i.e., >10 md/cp) where the recorded pressure is actually a good estimate of the true reservoir pressure
(Figure 2). Depending on the drilling situation (overbalanced or underbalanced drilling); this measured pressure is higher or
lower than the initial reservoir pressure in tight formations. In the previous study, the first two problems (slow buildup &
phase change) were addressed by introducing the automated pulse-test technique using a pressure feedback method to
monitor the pressure response and adaptively alter the flowrate of the test (Hadibeik et al. 2012). In this paper, the third
problem is solved by incorporating the supercharge model (Proett et al. 1996 and 2004) to correct the measured pressure
based on the available bottom-hole data.

Figure 1: Mudcake usuall y does not form in very low-mobility (tight formation) reservoirs, since the formation permeability is on the
same order of magnitude or lower than the mudcake permeability. Therefore, the supercharged effect is high and the measured
sandface pressure by tool is much larger than the reservoir pressure in the case of overbalanced drilling situation.

Figure 2: Mud-filtrate invasion is complete and mudcake forms full y in a permeable zone; therefore, most of the pressure drop is
across the mudcake, and the sandface pressure is very close to the far-field reservoir pressure, which is assumed to be the ini tial
reservoir pressure.
SPE 159172 3
Automated Pulse-Test Method
This method uses a smart feedback loop in the pressure-testing design to monitor the reservoir pressure response and
determine if an injection or a drawdown (i.e., production) pulse will be performed. After an initial drawdown or injection is
performed, there is a no flow period where the pressure transient is monitored and the feedback algorithm determines if the
next flow period will be an injection or drawdown in addition to the flow rate to be used. Typically, the flow and no-flow
periods are of equal time durations, but this is not a requirement. The subsequent series of injection and production tests
continue until a stabilized pressure is determined. The method starts from a pulse, a drawdown or an injection test, depending
on the drilling situation (overbalanced or underbalanced drilling). In many cases, the well bore balance is unknown, but the
method will still work even if the initial pulse is not in the ideal direction. In the simplest case, during the shut-in (i.e., no
flow period) a pressure gauge measures the pressure at the beginning and at the end of the shut-in test. If the pressure
decreases (the second pressure becomes less than the first one), this indicates that the pressure is declining toward the
reservoir pressure, and the next move should be a drawdown test. Otherwise, if the pressure is increasing in the shut-in
section, the reservoir pressure is higher than the current pressure, and there is a need for an injection test to move the pressure
closer to the reservoir pressure. The average slope during the shut-in periods can also be used and is less susceptible to noise
but the simple two-point method is used for illustration. The iterations continue until the maximum number of iterations has
been performed. Figure 3 illustrates the algorithm of this automated pulse-testpulse-test technique. Figure 4 depicts the
pressure response of an automated pulse-testpulse-test method obtained from the straddle packer. Table 1 presents the
reservoir properties and Table 2 through 4 shows the tester properties. The numerical model used in this study is UTFECS, a
previously validated compositional multi-phase near wellbore finite-difference fluid flow simulator (Pour 2011, Angeles
2008). As an alternative, automated pulse-testing could also be implemented with a fixed pattern of pulse sequences. Pre-job
designs can be simulated over the dynamic range of uncertainty concerning reservoir parameters of interest, and in-situ
optimization of pulse parameters can be performed through a feedback sequential classification routine using a look-up table
built from the pre-job designs.

Formation properties Values

Permeability[D] 1
Anisotropy (k
v
/k
h
) 1.0
Porosity [%] 10
Fluid viscosity [cP] 1.0
Fluid compressibility [psi
-1
] 3.0x10
-6

Formation pressure [psi] 20,000
Table 1: Synthetic tight sand reservoir model.
Properties Values
Flow line fluid compressibility [psi
-1
] 3.0x10
-6
Flowline volume (cc) 37,000
Wellbore diameter [ft] - ([in.]) 0.71 (8.5)
Packer height [ft] 4.2
Packer equivalent probe radius [in.] 25
Pulse rate [bbl/D] - ([cc/sec]) 1.0 (1.84)
Pulse time [sec] 30

Properties Values
Flow line fluid compressibility [psi
-1
] 3.0x10
-6
Flow line volume [cc] 250
Wellbore diameter [ft] 0.71
Oval probe height [in.] 9
Oval probe width [in.] 1.75
Oval probe radius [in.] 2.04
Probe geometric factor 1.2
Oval probe equivalent radius [in.] 1.49
Pulse rate [bbl/D]-([cc/sec]) 0.009 (0.0184)
Pulse time [sec] 30
Table 3: Oval probe, tester properties.
Table 2: Straddle packer, tester properties.
Properties Values
Flow line fluid compressibility [psi
-1
] 3.0x10
-6
Flow line volume [cc] 160.0
Well bore diameter [ft] 0.71 (8.5)
Probe radius [in.] 0.328
Probe geometric factor 0.6
Probe equivalent radius [in.] 0.239
Pulse rate [bbl/D] - ([cc/sec]) 0.0011 (0.002)
Pulse time [sec] 30
Table 4: Standard probe, tester properties.

4 SPE 159172

Figure 3: Automated pulse-testpulse-test algorithm and its pressure feedback. The number of repetitions of the feedback loop can
be set to a default number or a certain pressure bound criteria, meaning if the change in pressure was lower than a certain bound,
the iterations are stopped and the final no-flow shut-in is observed.

Figure 4: Automated pulse-testpulse-test feature. The initial reservoir pressure is 20,000 psi, and the overbalance pressure is 1,000
psi; it was assumed that during this test, the invasion no longer continues. A complete pulse-testpulse-test with a single drawdown
was performed before automation in this overbalanced drilling situation. Each section of the test (injection or pumpout) is followed
by a shut-in period. The pulse durations and injection or pumpout rates should be chosen properl y. In this plot, the flow rate of each
new action (injection or pumpout) is half of its previous action. The pulse duration of injection or drawdown test is set to constant of
30 seconds, followed by a 60-second shut-in period. With any prior reservoir information, it is possible to choose the best flow-rate
and pulse duration with optimization technique to start with, and help the pressure feedback loop to achieve the stabilized pressure
faster.

Supercharge Effect
The term supercharge is defined when the near wellbore pressure is different from the initial formation pressure which is
caused by an overbalanced pressure (the mud-filtrate invades the reservoir) or underbalanced drilling condition (the reservoir
bleeds into the wellbore). This effect makes the formation pressure near the borehole wall much higher or lower than the far-
field pressure in tight formations. The supercharging effect can be measured by adding an observation pressure gauge after
setting the packer- or probe-type formation tester. In the model, mud-filtrate invasion continues during the test. It is
interesting to note that after setting the packer- or probe-type formation tester, the pressure at the packer or probe location
declines even without performing a test (Figure 5). Figure 6 shows the automated pulse-testpulse-test with two hypothetical
observation probes. As expected, the stabilized pressure with the supercharge effect is different from the formation pressure,
SPE 159172 5
however, it is still beneficial to use the automated pulse-test method to reach the stabilized pressure (Figure 7, 8, and 9),
since the stabilized pressure is used to find the true initial formation pressure.


Figure 5: For the first synthetic case, the straddle packer was set in the middle of the formation after 1-day of mud-filtrate invasion.
An observation pressure gauge was placed 2-ft above the packer, one at the edge of the packer and one at the center of the packer.
Although the invasion continues, the pressure at the center of the packer declines without performing any test. It turns out the
supercharge pressure propagates into the reservoir, when the packer seals its locating section from the filtrate invasion. This
pressure decline depends on the reservoir mobility and its porosity.


Figure 6: Automated pulse-testpulse-test method for a straddle packer with a pressure gauge outside of the packer wall. With
superposition the amplitude response of pressure at the observation probe becomes larger with time even though the pulse signal
amplitude declines with time.


Figure 7: Comparison of an automated pulse-testpulse-test with a single pulse-testpulse-test for the straddle packer in the synthetic
reservoir of Table 1. The observation probe for the automated pulse-testpulse-test and the single pulse-testpulse-test is shown in
0 500 1000
2.03
2.04
2.05
2.06
2.07
2.08
2.09
2.1
x 10
4
Time [min]
P

[
p
s
i
]
2ftabovethepacker
Attheedgeofthepacker
Centerofthepacker
0 5 10 15 20 25 30
1.8
1.9
2
2.1
2.2
x 10
4
Time [min]
P

[
p
s
i
]
Pressuregaugeatthepackerlocation
Pressureattheobservationgauge
atadistancefromthesource
0 5 10 15 20 25 30
1.8
1.9
2
2.1
2.2
x 10
4
Time [min]
P

[
p
s
i
]
0 5 10 15 20 25 30
1.8
1.9
2
2.1
2.2
x 10
4
Time [min]
P

[
p
s
i
]
Pressuregaugeatthepackerlocation
Pressureattheobservationgauge
atadistancefromthesource
0 5 10 15 20 25
1.75
1.8
1.85
1.9
1.95
2
2.05
2.1
2.15
x 10
4
Time [min]
P

[
p
s
i
]


Q
+
=Q
-
, BD Stab. T =360.3 min
Observation Probe
Automated, BD Stab. T =14.64 min
Observation Probe
6 SPE 159172
black and green respecti vel y. The automated pulse-testpulse-test reaches a limit of 0.35% of a stabilized pressure approximatel y 19
times faster than a single pulse-testpulse-test.

Figure 8: Comparison of automated pulse-testpulse-test method with a single pulse-testpulse-test technique for an oval pad tester
in the synthetic reservoir of Table 1. In this case, both methods reach the stabilized pressure on the same time-frame order of
magnitude. Automated pulse-testpulse-test method could be faster if (a) initial flowrate was chosen appropriatel y, and (b) reduced
number of iterations was applied , which results in the final buildup occurring sooner.


Figure 9: Automated pulse-testpulse-test and a single pulse-testpulse-test for the standard probe in the first synthetic reservoir.
Pulse-testPulse-test Data Analysis
When a stabilized pressure is determined with the automated pulse-testpulse-test, the data can be further analyzed to obtain
reservoir properties such as mobility, and a correction made to the stabilized pressure to account for the supercharge effect to
obtain the true formation pressure. In this section, two methods are proposed for this analysis. The first method is a fast
analysis to obtain the reservoir permeability and reservoir pressure after correcting the supercharge effect with a spherical
flow model. The second method is a robust inversion algorithm to estimate reservoir parameters and other unknown
properties. Furthermore, this detailed optimization method is used to validate the findings from the first method and yields
additional information.

Derivative Analysis and Supercharge Model
The essential information can be obtained from the automated pulse-test data by using a combined method of supercharge
modeling and a conventional pressure-transient analysis. Despite the fact that the pulse-test method reaches pressure stability
much faster than a conventional reservoir test, the presence of mud-filtrate invasion results in a supercharged final pressure in
the tight formations. The consequence is that the stabilized pressure can be very different from the reservoir pressure. In this
case, there is a need for a model to account for the supercharging effect. The model should also handle both overbalanced and
underbalanced situations. For the overbalanced case the sandface pressure will stabilize at a higher pressure than the initial
formation pressure. The opposite is true for the underbalanced case where the stabilized test pressure is lower than the actual
initial formation pressure. The model developed in this study has been tested successfully for these cases. Below is a brief
summary of these developments:

0 5 10 15 20 25
1.98
2
2.02
2.04
2.06
2.08
2.1
2.12
2.14
x 10
4
Time [min]
P

[
p
s
i
]


Q
+
=Q
-
, BD Stab. T =18.7 min
Observation Probe
Automated, BD Stab. T =24.5 min
Observation Probe
0 5 10 15 20 25
1.9
1.95
2
2.05
2.1
2.15
2.2
x 10
4
Time [min]
P

[
p
s
i
]


Q
+
=Q
-
, BD Stab. T =21.6 min
Observation Probe
Automated, BD Stab. T =20.8 min
Observation Probe
SPE 159172 7
Overbalanced Drilling
This model can be used as a fast analysis on jobsite evaluations to determine the reservoir pressure and permeability as:
o

t
=

14696 ( )
2 ( ( ))
bu
f
p ibu
q t
k
r P P t
, (1)
where k
f
is the formation permeability, is the fluid viscosity, q
bu
(t) is the mud-filtrate invasion rate during the buildup
period, P
ibu
is the initial pressure at the start of the buildup period, P(t) is the pressure changing with time, r
p
is the probe
equivalent radius, and
o
is the shape factor. Because the reservoir fluid near the wellbore is dominated by the mud-
filtratemud-filtrate invasion, the fluid viscosity can be assumed to be the mud-filtratemud-filtrate.

The invasion rate during the buildup period can be calculated as:
= ( )
bu fl fl
dp
q t c V
dt
. (2)
where c
fl
is the flowline fluid compressibility, and V
fl
is the flowline volume which is known a priori for each formation-tester
tool. Storage is defined to be the product of flowline fluid compressibility (c
fl
) and flowline volume (V
fl
):
C =c
fl
V
fl
. (3)

Form the early time of the final shut-in test, the mud-filtrate invasion rate can be determined as follows (see Proett et al.
1996):
o
=
1
( ( ))
ibu
dp
P P t
dt
, (4)

where o is a constant; knowing the pressure during buildup period, and its derivative, o can be calculated as:
o
=

1
( ( ))
ibu
t
dp
dt
P P t
. (5)
The calculation of formation permeability is determined by:
o

t o
=
14696 1
( )
2
fl fl
f
p
c V
k
r
o

t o
=
14696 1
( )
2
fl fl
f
p
c V
k
r
. (6)
The supercharge pressure (AP
sc
) is defined as the difference between sandface pressure (P
ss
) and formation pressure (P
f
):

t |
A = =
2
4
14696 ( )
2
m f
sc ss f
f w
q k t
P P P Ln
hk cr
, (7)
in tight sand formation, there is no mudcake present; therefore, sandface pressure (P
ss
) is the same as mud hydrostatic
pressure (P
mh
), and q
m
is the filtrate loss.
The velocity of the fluid near the wellbore is defined as:


= ( )
14696
f ss sb
m
e e
k P P
S
r
, (8)
which is the disturbance caused by the pad element blocking the seepage of the mud around the source;
e
is the element
shape factor, and r
e
is the local geometric correction for non-spherical effects.

Combining (7) and (8), formation pressure (P
f
) is obtained as:
|
=
2
4
( ) ( )
w f
f mh mh sb
e e w
r k t
P P P P Ln
r cr
, (9)
where P
sb
is the final stabilized pressure at the end of the buildup test. The faster this stabilization can be achieved, the faster
and more accurate the formation pressure can be determined. The automated pulse-test helps to achieve P
sb
faster than
conventional methods.

Figure 10 shows a draw of the situation and parameters. Figure 11 presents the results of automated pulse-test under the
supercharge effect. Figures 12 through 14 illustrate the derivative analysis in conjunction with the supercharge model to
obtain true reservoir pressure and permeability.
8 SPE 159172

Figure 10: True reservoir pressure is estimated by appl ying the supercharge model in tight sand formations. Gi ven the wellbore
bottom-hole pressure, by measuring the stabilized pressure at the wellbore, it is possible to determine the formation pressure. The
benefit of the automated pulse-test method is to yield the stabilized pressure faster than conventional pressure testing methods.


Figure 11: Semi-log data anal ysis of a pulse-test performed in a synthetic formation with a packer-type formation tester. Assuming
that the formation pressure and its permeability are unknown, then these properties are calculated based on the pressure-transient-
anal ysis technique and the supercharge model.


Figure 12: Semi-log data anal ysis of a pulse-test performed in a formation with an unknown pressure and permeability. This figure
shows the final build-up (shut-in) period after finishing the pulse-tests. Invasion rate is calculated from the early section of the final
shut-in test. Storage is determined by Equation 3, and the rate of mud-filtrate invasion is obtained from the final shut-in section of
the test from Equation 2.

10
-6
10
-4
10
-2
10
0
10
2
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
2.12
x 10
4
P

[
p
s
i
]
Time [min]
3x10
-6
c
t
[psi
-1
]
10 | ||
3x10
-6
c
t
[psi
-1
]
10 | ||
110 Q [cc/min]
37.0 V
fl
[liter]
110 Q [cc/min]
37.0 V
fl
[liter]
Overbalance pressure: 1000 psi
80 100 120 140
0
100
200
300
400
500
600
700
A
P

[
p
s
i
]
Time [min]
y =8.94x +609.9
R =6x10
-5
6 4
3 10 232.7 6.98 10
fl fl
C c V

= = =
24
bu
d p
q C
dt
A
=
4
24 6.98 10 8.94 0.149 / bbl D

= =
80 100 120 140
0
100
200
300
400
500
600
700
A
P

[
p
s
i
]
Time [min]
y =8.94x +609.9
R =6x10
-5
6 4
3 10 232.7 6.98 10
fl fl
C c V

= = =
24
bu
d p
q C
dt
A
=
4
24 6.98 10 8.94 0.149 / bbl D

= =
SPE 159172 9

Figure 13: Calculation of the reservoir permeability from the shut-in periods of the pulse-test data. Any intermediate shut-in period
can be anal yzed to obtain the reservoir permeability from the slope of its tangent line. The two different shut-in periods pressure
data was analyzed and the permeability is close to that of the reservoir model.

Figure 14: The supercharge model applied to the extended shut-in section of the automated pulse-test successfull y yields the true
reservoir pressure (P
f
) with equation 9. However, the conventional anal ysis (interception of the tangent line on the pressure data
with pressure axis) results in an inaccurate report for the initial reservoir pressure. This result shows the importance of integrati on
of the two models (automated pulse-testing with the supercharge model).

Underbalanced Drilling.
The same concept can be applied to the underbalanced drilling situation. The only difference is that, ideally, there is no mud
filtration in the underbalanced drilling situation; therefore, the pressure transient analysis should be applied to the pumpout
(drawdown) tests combined with the supercharge model instead of shut-in tests to determine the formation permeability. The
other difference is the way the automated pulse-test is performed. In the underbalanced drilling situation, the reservoir
pressure near the wellbore is lower than the initial reservoir pressure; therefore, the automated pulse-test can start with an
injection instead of a pumpout test to speed up the process. Figures 15 and 16 illustrate the analysis to obtain the reservoir
pressure and permeability in an underbalanced drilling situation. The results indicate that this simple and fast approach to
obtaining the reservoir parameters is a reasonable approach, as demonstrated in a synthetic reservoir case. However, this
method should apply equally as effective for the field test data.
10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
10
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
2.12
x 10
4
P

[
p
s
i
]
Horner Time
162.6
bu
q
k
mh

=
2
162.6 0.149 1
1.9
294.27 42
nd
k D

= =

P =-294.27log(Horner T) +20975
P =-224.24log(Horner T) +20575
Eq.s of tangential lines
1
162.6 0.149 1
2.5
224.24 42
st
k D

= =

10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
10
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
2.12
x 10
4
P

[
p
s
i
]
Horner Time
162.6
bu
q
k
mh

=
2
162.6 0.149 1
1.9
294.27 42
nd
k D

= =

P =-294.27log(Horner T) +20975
P =-224.24log(Horner T) +20575
Eq.s of tangential lines
1
162.6 0.149 1
2.5
224.24 42
st
k D

= =

10
0
10
1
10
2
10
3
10
4
10
5
10
6
10
7
10
8
10
9
10
10
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
2.12
x 10
4
P

[
p
s
i
]
Horner Time
P =-111.05 log(Horner T) +20382
Conventionalanalysis:
P
int
=20382psi

Superchargemodel:
20003.3
f
P psi =
2
0.355 4 0.0019 14.6/ 60
( )
2.06 0.1 1 3 6 0.355
Ln
e


21000 (21000 20464.5)
f
P =
2
4
( ) ( )
f w
f mh mh sb
e e w
k t r
P P P P Ln
r cr |
=
10 SPE 159172

Figure 15: Calculation of permeability from a pulse-test data obtained from a synthetic reservoir with k = 10
-5
mD. The calculated
permeability is very close to that of the actual value.

Figure 16: Calculation of reservoir pressure with the supercharge model for an underbalanced drilling. It turns out that the anal ysis
of an underbalanced situation is the same as an overbalanced case if the pressure (y-axis) is reversed.

Genetic Algorithm Optimization and Inversion
The developed pressure analysis method combined with the supercharge model is a good approximation for a fast estimation
of reservoir properties on jobsites. Further detailed evaluation can be performed to provide additional reservoir parameters
and confirm the obtained parameters with previous method. The genetic algorithm (Figure 17) which is based on
evolutionary computation theory is used for optimization and feature matching of pressure data. In this approach an initial
population of guesses for the set of unknown reservoir parameters will be used to reproduce the automated pulse-test
measurements. An objective cost function calculates the difference between the measurements and each of the simulated tests
and is based on discrepancy of individual guess. The genetic algorithm offers a new population of parameter sets with use of
genetic operators such as ranking, selection, mutation and crossover to minimize the cost function. If the reproduced
measurements matched features of the automated pulse-test reasonably and within a limited error bound or other stopping
criteria were satisfied, the evolutionary computation can be terminated. Compared to conventional optimization methods,
genetic algorithm eliminates the need for calculating the gradient, and is capable of providing multiple solutions over the
selected range of whole parameter space simultaneously.

10
-4
10
-3
10
-2
10
-1
10
0
10
1
10
2
1.94
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
x 10
4
P

[
p
s
i
]
Time [min]
1
162.6
DD
st
q
k
mh

=
162.6 0.5 1
7912 42

=

4
1
2.4 10
st
k mD

=
P =-7912 log(T) -1681.3
R
2
=0.999
10
0
10
2
10
4
10
6
10
8
10
10
1.94
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
x 10
4
P

[
p
s
i
]
Horner Time
P =-100.23 log(Horner T) +19804
Supercharge model:
4
2
0.355 4 2.4 10 14.39/ 60
19500 (19500 19785.3) ( )
2.06 0.1 1 3 6 0.355
19928.4
f
f
P Ln
e
P psi

=

=
10
0
10
2
10
4
10
6
10
8
10
10
1.94
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
x 10
4
P

[
p
s
i
]
Horner Time
P =-100.23 log(Horner T) +19804
Supercharge model:
4
2
0.355 4 2.4 10 14.39/ 60
19500 (19500 19785.3) ( )
2.06 0.1 1 3 6 0.355
19928.4
f
f
P Ln
e
P psi

=

=
SPE 159172 11

Figure 17: The genetic algorithm is a stochastic global search method that mimics the metaphor of natural biological evolution. In
the natural biological evolution, the ones who survi ve are the fittest to produce (hopefull y) better and better approximations to a
solution. This optimization method is used to find the best fit for the parameters the reservoir model. As an alternative to the
previous anal ytical method, reservoir parameters can also be estimated through inverse processing by evol vi ng unknown formation
parameters with genetic algorithm and numerical models.

Overbalanced Drilling
As an example, the synthetic field data was used to do the feature matching and obtain the reservoir and in-situ fluid
properties in tight oil formations. Table 6 and 7 show the true reservoir parameters and the inverted properties obtained from
feature matching with the genetic algorithm optimization for the synthetic reservoirs of case 1 and 2 of Table 5. Figure 18
and 19 illustrate the measurements and reproduced measurements from the genetic algorithm.
| [%] 10
[cp] 1.0
c
t
[psi
1
] 3x10
6

K
1
[D] 1.0
K
2
[D] 0.01
Table 5: The physical properties of two synthetic homogenous reservoirs with different permeabilities.

MeasurementsPressure InvertedPressure
[psi] [psi]
20976 20904
19798 19787
19863 19890
20470 20487
20466 20474
20185 20208
20222 20262
20377 20413
20384 20416
20532 20556
20526 20539
20454 20459
20462 20468

Figure 18: Automated pulse-test data obtained for the synthetic tight sand model of case 1 with a straddle-packer formation tester.
Four parameters (formation pressure, permeability, in-situ fluid viscosity, and reservoir porosity) were assumed to be unknown and
feature matching algorithms were used to generate the pressure response. The method successfull y estimates the reservoir
parameters. The cost function has 13 pressure points, which are at the beginning and at the end of each test. Decreasing the
number of unknown parameters reduces the uncertainty and improves the feature matching. The overall advantage was that the
reservoir parameters in a tight formation with 1D permeability could be determined with a test that was approximatel y 15 minutes.

GenerateInitialPopulationof
ReservoirParameters
EvaluateObjectiveFunction
=80%
=30%
=60%
Stop
Yes
No
GAOperator
Selection
Crossover
Mutation
Old
New
NextGeneration
FeatureMatching
Satisfied?
0 5 10 15
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
x 10
4
Time [min]
P

[
p
s
i
]


Match Curve with Genetic Algorithm
Simulated Field Test
12 SPE 159172
MeasurementsPressure InvertedPressure
[psi] [psi]
20938.0 20936.9
19672.2 19657.5
19680.1 19659.9
20287.2 20263.8
20290.3 20265.5
20904.4 20879.6
20900.1 20878.3
20589.1 20569.7
20591.0 20570.0
20744.7 20723.8
20743.1 20723.3
20665.7 20646.1
20665.7 20646.0

Figure 19: Automated pulse-test data obtained from the synthetic tight sand model of Case 2 with a straddle-packer formation
tester. It turns out that reducing the number of parameters decreases the uncertainty of the inverted parameters.
ReservoirParameters Actual Reproduced
P
f
[psi] 20,000 19,925
K(D) 1.0 2.9
(cp) 1.0 1.02
|(%) 10.0 11.2



ReservoirParameters Actual Reproduced
P
f
[psi] 20,000 19984.3
K(D) 0.01 3.52x10
2
Table 6: Estimated reservoir parameters and the actual values
for Figure 18. There are 4 unknowns to be inverted for this
reservoir.
Table 7: Estimated reservoir parameters and the actual values
for Figure 19. The formation pressure and reservoir permeability
are assumed unknown.

Underbalanced Drilling
In the underbalanced situation, the pressure is 500 psi less than the formation pressure. A straddle packer is used to perform
the test with the properties summarized in Tables 8 and 9. Figure 20 shows the results of inversion with the genetic
algorithm to obtain the reservoir properties.

Straddle Packer, Tester Properties
Flowline fluid compressibility [psi
-1
] 3.0x10
-6

Flowline volume [bbl] 0.0
Wellbore diameter [ft] - ([in]) 0.71 (8.5)
Packer height [ft] 4.2
Packer equivalent probe radius [in] 25
Pulse rate [bbl/D] - ([cc/sec]) 1.0 (1.84)
Pulse time [sec] 30
Table 8: The flowline volume is zero to account for zero wellbore storage.

Type of formation tester Volume [cc]
Straddle packer 37,000
Oval pad 250
Standard Probe 160
Table 9: The amount of flowline volume, which results in the storage effect for each formation-tester type.


0 5 10 15 20
1.96
1.98
2
2.02
2.04
2.06
2.08
2.1
x 10
4
Time [min]
P

[
p
s
i
]


Match Curve with Genetic Algorithm
Simulated Field Test
SPE 159172 13
MeasurementsPressure InvertedPressure
[psi] [psi]
19578.5 19531.2
20773.5 20715.3
20775.9 20770.8
20170.9 20157.5
20155.4 20155.4
19558.3 19556.2
19586.4 19557.1
19859.6 19855.0
19866.4 19854.5
19717.6 19705.0
19705.2 19701.5
19779.7 19696.6
19779.6 19703.1
19778.6 19704.6


Figure 20: Inversion of simulated field test for the synthetic reservoir model of case 2. The estimated pressure is closer to the true
formation pressure than the overbalanced cases. One reason is that the overbalanced pressure in previous cases was 1,000 psi
above the formation pressure; the underbalanced pressure is 500 psi less than the formation pressure, which is 500 psi closer to
the formation pressure.
Calibration Transfer for Fast Inversion
Calibration transfer is the basic procedure dealing with the difference between the lab measurements or simulated ones
(called the first instrument) and field measurements (called the second instrument). Generally, it converts the measurements
of the second instrument to the measurements of the first instrument through a transformation or standardization algorithm
in order to have a common base for analysis. Applied to pulse-test data inversion, the numerical method could simulate the
field experiments more closely by including considerably detailed geometric and additional boundary conditions with cost of
high intensity computation that would limit its use in standard practice. This shortcoming could be overcome through a
robust mapping, which compensates all borehole environmental factors and generates analytically equivalent measurements
that can be processed with a faster inversion algorithm. As a substitute for the field data the results generated from a
numerical simulations can be used as input of a transformation algorithm, and matched to cases generated from analytical
simulations to transform the output of the model. The transformation algorithm is implemented with a neural network (NN),
as shown in Figure 21, taking pressure feature points simulated with numerical and analytical methods as NN inputs and
outputs, respectively. In this application the pulse rate, pumpouts and shut-in interval are optimized with genetic algorithm,
first on the selected examples, and set to the same for each transformation pair of numerical and analytical simulations.
Moreover, the pulse sequence requires a fixed pattern, i.e., same number of injection and drawdown tests in order, applied to
field tests. Table 10 provides a summary of synthetic reservoir parameters used for training, validating and testing the
transformation algorithm. Figure 22 presents an example of improved estimation of reservoir parameters with analytical
inversion. The feature pressure points are converted from numerical simulation through calibration transfer before using the
analytical model.

0 2 4 6 8 10
1.95
2
2.05
2.1
x 10
4
Time [min]
P

[
p
s
i
]


Match Curve with Genetic Algorithm
Simulated Field Test
14 SPE 159172
Figure 21: This plot conceptually shows the neural network transformation algorithm to convert feature pressure points (12 points
in this example) of numerical simulations, which are close analogue for field tests, to the same number of feature pressure points
obtained from anal ytical simulations. Note that the supercharge effect observed in numerical simulations is compensated through
transformation which allows fast inversion under anal ytical near-ideal conditions.
Properties K (D) (cp) |(%) P
Reservoir
P
Bottomhole
Case 1 1.0 1.0 5 17,000 17,300
Case 2 0.1 10 15 17,000 17,300
Case 3 0.01 10 20 17,000 17,300
Case 4 10.0 1.0 10 17,000 17,300
Case 5 1.0 10 15 17,000 17,300
Case 6 0.01 1.0 10 20,000 20,300
Case 7 0.001 1.0 10 20,000 20,300
Case 8 0.1 1.0 10 20,000 20,300
Case 9 1.0 1.0 10 20,000 20,300
Case 10 10 1.0 10 20,000 20,300
Case 11 10.0 1.0 5 15,000 15,300
Case 12 1.0 10 15 15,000 15,300
Case 13 0.01 10 20 15,000 15,300
Case 14 1.0 1.0 5 15,000 15,300
Case 15 0.1 10 15 15,000 15,300
Case 16 10.0 1.0 5 10,000 10,300
Case 17 1.0 10 10 10,000 10,300
Case 18 0.1 10 15 10,000 10,300
Case 19 0.01 10 20 10,000 10,300
Case 20 1.0 1.0 5 10,000 10,300
Case 21 10.0 1 5 12,500 12,800
Case 22 1.0 1 5 12,500 12,800
Case 23 1.0 10 10 12,500 12,800
Case 24 0.01 10 20 12,500 12,800
Case 25 0.1 10 15 12,500 12,800
Table 10: Various synthetic low-mobility reservoir models used to perform the training, validation and testing to obtain the
transformation function. The bottomhole pressure is higher than the reservoir pressure, therefore, the well was drilled in an
overbalanced condition. Various values of porosity, permeability, fluid viscosity, and formation pressure are used to encompass
large reservoir parameters to build a robust transformation function.

Figure 22: Inversion of synthetic pulse-testing data with anal ytical model by matching the target feature pressures converted from
the numerical simulations. Compared to inversion with numerical input directl y, inversion through calibration transfer produces
better estimation of actual reservoir parameters.

Neural Network Inversion
Neural network inversion was investigated as an alternative to genetic algorithm inversion with flow models. In NN
inversion, simulated pulse parameters and feature pressure points are used as inputs to predict reservoir parameters directly.
Figure 23 illustrates the concept of NN inversion. Figures 24 explains the typical pulse parameters and Figure 25 illustrates
how these parameters can be optimized through simulation for training neural network. Figures 26 and 27 show the
predictions on formation pressure, fluid mobility and reservoir porosity by using analytical simulation inputs and numerical
simulation inputs respectively. The results show that the analytical simulation based input gives a more accurate prediction
on outputs because of its nature of static system response. The nonlinearity of an underlying mapping function is higher with
use of numerical simulation results because of the variation of dynamic factors such as supercharge and geometric grid
1 2 3 4 5 6 7 8 9
1.4
1.42
1.44
1.46
1.48
1.5
1.52
1.54
x10
4
Time (min)
P
r
e
s
s
u
r
e

(
p
s
i
)
Pij1
Pbd1
Pbd2
Pdd2
Pdd3
Pbu2
Pbu3
Pdd1
Pij2
Pij3
Pbu1
Pstb
Pulse Feature Points

Target(psi) Match (psi)
Pdd1 14505 14519
Pbu1 14562 14572
Pij1 15420 15433
Pbd1 15352 15360
Pdd2 14871 14876
Pbu2 14899 14902
Pij2 15020 15023
Pbd2 15014 15016
Pdd3 14961 14961
Pbu3 14965 14964
Pij3 15000 14999
Pstb 14996 14978

Reservoir Parameters
Actual Estimation
FP (psi) 15000 14953
MOB (md/cp) 0.0001 0.000152
Cf (1/psi) 3.00E-06 3.04E-06
Cm(1/psi) 3.00E-07 2.99E-06
Vf (cc) 37000 36878
Phi (%) 10.00 10.10
Cd 0.1150 0.1116

SPE 159172 15
meshing. However, as demonstrated in Figures 24 and 25, the results in both cases are promising if an adequate number of
simulations, which are tool type based, are used to build the direct inversion model. As a very fast solution method, NN
inversion can provide reservoir parameter predictions in seconds compared to an analytical model based inversion taking
minutes and numerical inversions implementation taking hours with the genetic algorithm.


Figure 23: Neural network inversion of pulse-testing data. The inputs of neural network include pulse parameters and feature
pressure points of pulse-test. The outputs of NN are multiple reservoir parameters such as initial pressure, fluid mobility, formation
porosity, flow line volume and fluid compressibility.


Figure 24: Typical pulse parameters. Q denotes flow rate, low case dd and ij represents drawdown and injection test respecti vel y,
and numerical number 1, 2, 3 indicates sequential index for each acti vity. Note that several flow rate ratios are used as actual pulse
parameters through optimization.


Figure 25: Optimized pulse parameters as a function of reservoir parameters and used as partial inputs for NN training. The gi ven
examples are simulated with straddle parker, and NN direct inversion models can be built respecti vel y based on tool types.
2 4 6 8 10 12
1.45
1.46
1.47
1.48
1.49
1.5
1.51
1.52
1.53
1.54
x10
4
Time (min)
P
r
e
s
s
u
r
e

(
p
s
i
)
Qdd2
Qdd3
Qij3
Qdd1
Qij1
BUDT DDPT
Qij2
DDPT: Drawdown and injection pulse time
BUDT: Buildup and builddown time between pulses
Qdd2/Qij1: Flow rate ratio of Qdd2 over Qij1
Qij2/Qdd2: Flow rate ratio of Qij2 over Qdd2
Qdd3/Qij3: Flow rate ratio of Qdd3 over Qij3
Qij3/Qdd3: Flow rate ratio of Qij3 over Qdd3
2 4 6 8 10 12 14
1.6
1.65
1.7
1.75
x10
4
Time (min)
P
r
e
s
s
u
r
e

(
p
s
i
)

VIS=1.0 (cp), PHI=0.10, MOB=0.01
VIS=1.0 (cp), PHI=0.05, MOB=0.001
VIS=10 (cp), PHI=0.15, MOB=0.0001
VIS=10 (cp), PHI=0.15, MOB=0.00001
VIS=10 (cp), PHI=0.20, MOB=0.000001

DDPT BUDT Qdd2/Qij1 Qij2/Qdd2 Qdd3/Qij2 Qij3/Qdd3
(sec) (sec) (%) (%) (%) (%)
MOB=0.01 13.61 30.00 68.24 26.24 28.35 61.88
MOB=0.001 24.43 56.35 45.65 11.76 69.29 12.12
MOB=0.0001 10.00 48.12 71.76 75.29 74.24 34.00
MOB=0.00001 34.35 34.12 49.18 11.05 18.12 33.29
MOB=0.000001 39.76 30.82 68.94 49.18 82.71 67.88
16 SPE 159172


Figure 26: NN Prediction of initial formation pressure, fluid mobility and reservoir porosity using anal ytically simulated pulse-test
parameters. The same examples as shown in Table 10 are used for training, validating and testing the direct inversion model. Very
accurate prediction is achieved in these cases.


Figure 27: NN Prediction of initial formation pressure, fluid mobility and reservoir porosity using numerically simulated pulse-test
parameters. The same examples as shown in Table 10 are used for training, validating and testing the direct inversion model.
Reasonabl y good prediction is achieved in these cases.

Conclusions
The analysis of the automated pulse-test method shows that it is possible to obtain the initial formation pressure and mobility
in very low-mobility reservoirs. In the presence of mud-filtrate invasion, the supercharge model was used to successfully
yield the formation pressure. Further analysis on the pressure data using different inversion methods were performed to
compare the speed of inversion method, determine if additional reservoir parameters could be determined and compare the
accuracy of the results.
Pressure transient analysis considering the supercharge effect helps to find the reservoir pressure and permeability. The
analysis can be done using an early shut-in tests during the automated pulse-test method for the overbalanced drilling
situations and during drawdown tests for the underbalanced drilling cases.
Using several inversion methods the additional formation properties were determined from the pulse-testing data. For
example, the estimated reservoir pressure had less than 0.5% error from the actual data.
The extended shut-in period and the stabilization pressure can be used to determine the actual average reservoir pressure.
The conventional testing analysis reported a large error when predicted the average reservoir pressure due to the
supercharge effect.
To reduce the CPU calculation time to analyze the data, the neural network transformation is successfully used to
calibrate the numerical model with the analytical model. The same approach can be done to transform the field data into
the analytical model for further analysis.
0 5 10 15 20 25
0.5
1
1.5
2
2.5
x 10
4
F
o
r
m
.

P
r
e
s
s
u
r
e

(
p
s
i)


0 5 10 15 20 25
10
-10
10
-5
10
0
M
o
b
i
li
t
y

(
m
d
/
c
p
)
0 5 10 15 20 25
0
0.1
0.2
0.3
0.4
Case Index
P
o
r
o
s
it
y
NN Inversion
Analy. Simulation
0 5 10 15 20 25
0.5
1
1.5
2
2.5
x 10
4
F
o
r
m
.

P
r
e
s
s
u
r
e

(
p
s
i
)
0 5 10 15 20 25
10
-10
10
-5
10
0
M
o
b
i
l
i
t
y

(
m
d
/
c
p
)
0 5 10 15 20 25
0
0.1
0.2
0.3
0.4
Case Index
P
o
r
o
s
i
t
y
SPE 159172 17
To remove the effect of noise on the feedback pressure loop in the automated pulse-test method, it is possible to use the
change in the slope of the line to make the decision about the next pressure test instead of observing the two pressure
end-points at the shut-in interval.

Nomenclature
q
bu
(t) : Invasion rate, [bbl/D] c
t
: Compressibility, [psi
-1
]
q
m
: Filtrate loss, [bbl/D] V
fl
: Flowline volume, [cc]
Q : Pulse rate, [cc/min] r
p
: Probe equivalent radius, [in]
P
ibu
: Initial pressure at the start of buildup test, [psi] r
e
: Local geometric correction, [in]
P(t)

: Flowline volume, [psi ] |

: Porosity, [%]
P
sc
: Supercharged pressure, [psi] : Viscosity [cp]
P
ss
: Sandface pressure, [psi] k : Absolute permeability, [D]
P
f
: Formation pressure, [psi]
e
: Element shape factor, [ ]
P
mh
: Mud hydrostatic pressure, [psi]
o
: Shape factor, [ ]
P
sb
: Final stabilized pressure, [psi] o : Constant factor
M : Mobility, [D/cp]

Acronyms
BD : Build-down Stabilization RDT : Reservoir Description Tool
DD : Drawdown Test DPS : Dual Probe Section
BU : Build-up Test OBM : Oil-base Mud
INJ : Injection Test WBM : Water-base Mud
References
Abu Omokaro, F., Musharfi, N., Soliman, W., Rourke M., and Eid, M., 2011. Challenges of Wireline-Formation Testing and Fluid
Sampling in Tight, Low-Permeability Gas Reservoirs: A Case Study from Saudi Arabia. Paper SPE 141840 presented at the SPE
Middle East Oil and Gas Show and Conference, 25-28 September, Manama, Bahrain.
Alpak, F., Torres-Verdn, C., Sepehrnoori, K., and Fang, S., 2002. Numerical Sensitivity Studies for the J oint Inversion of Pressure and DC
Resistivity Measurements Acquired with In-Situ Permanent Sensors. Paper SPE 77621 presented at the SPE Annual Technical
Conference and Exhibition, 29 September-2 October, San Antonio, Texas.
Angeles, R., 2008. Simulation and Interpretation of Formation-Tester Measurements Acquired in the Presence of Mud-Filtrate Invasion,
Multiphase Flow, and Deviated Wellbores. Ph.D. dissertation, The University of Texas at Austin, Austin, Texas, USA.
Angeles, R., Torres-Verdn, C., Lee, H., Alpak, F., and Sheng, J., 2007. Estimation of Permeability and Permeability Anisotropy from
Straddle-Packer Formation-Tester Measurements Based on the Physics of Two-Phase Immiscible Flow and Invasion. SPE Journal,
12, 3.
Angeles, R., Torres-Verdn, C., Hadibeik, A., and Sepehrnoori, K., 2010. Estimation of Capillary Pressure and Relative Permeability from
Formation-Tester Measurements Using Design of Experiment and Data-Weighing Inversion: Synthetic and Field Examples. Journal
of Petroleum Science and Engineering, 75, 1-2.
Bennett, E.N., and Forgerson, C.D., 1975. Predicting Reserves and Forecasting Flow Rates of Relatively Tight Gas Wells Using Limited
Performance Data. Journal of Petroleum Technology, 27, 5.
Dastidar, R., Sondergeld, C., and Rai, C., 2007. An Improved Empirical Permeability Estimator from Mercury Injection for Tight Clastic
Rocks. Petrophysics, 48, 3.
Elshahawi, H., Venkataramanan, L., McKinney, D., Flannery, M., Mullins, O., et al., 2008. Combining Continuous Fluid Typing, Wireline
Formation Testers, and Geochemical Measurements for an Improved Understanding of Reservoir Architecture.SPE Reservoir
Evaluation and Engineering, 11, 1.
Galford, E., and Marschall, D., 2000. Combining NMR and Conventional Logs to Determine Fluid Volumes and Oil Viscosity in Heavy-
Oil Reservoirs. Paper SPE 63257 presented at the SPE Annual Technical Conference and Exhibition, 1-4 October, Dallas, Texas.
Hadibeik, H., Proett, M., Chen, D., Eyuboglu, S., Torres-Verdn, C., Pour, R., 2012. Formation-Tester Pulse-testing in Tight Formations
(Shales and Heavy Oil): Where Wellbore Storage Effects Favor the Determination of Reservoir Pressure. Paper SPE 155037
presented at the SPE Americas Unconventional Resources Conference, Pittsburgh, Pennsylvania, USA, 57 J une 2012.
Hadibeik, H., Pour, R., Torres-Verdn, C., Sepehrnoori, K., and Shabro, V., 2011. 3D Multiphase Streamline-Based Method for
Interpretation of Formation- Tester Measurements Acquired in Vertical and Deviated Wells. Paper SPE 146450 presented at the SPE
Annual Technical Conference and Exhibition, 30 October-2 November 2011, Denver, Colorado, USA.
Hadibeik, H., Proett, M., Torres-Verdn, C., Sepehrnoori, K., and Angeles, R., 2009. Wireline and While-Drilling Formation-Tester
Sampling With Oval, Focused, and Conventional Probe Types in the Presence of Water- And Oil-Base Mud-Filtrate Invasion in
Deviated Wells. Paper SPWLA 86800 presented at the SPWLA 50th Annual Logging Symposium, J une 21 - 24, The Woodlands,
Texas, USA.
Kundert D., and Mullen, M., 2009. Proper Evaluation of Shale Gas Reservoirs Leads to a More Effective Hydraulic-Fracture Stimulation.
Paper SPE 123586 presented at the SPE Rocky Mountain Petroleum Technology Conference, 14-16 April, Denver, Colorado.
Mason, R.A., 1987. Estimating Reserves for High-Rate Devonian Shale Oil and Gas Completions with Limited Production Data. SPE
Reservoir Engineering, 2, 3.
Nakano, C., Pinto, A., Marcusso, J ., and Minami, K., 2009. Pre-Salt Santos Basin-Extended Well Test and Production Pilot in the Tupi
Area-The Planning Phase. Paper SPE 19886 presented at the Offshore Technology Conference, 4-7 May, Houston, Texas.
Pour, R., 2011. Development and Application of a 3D Equation-Of-State Compositional Fluid-Flow Simulator in Cylindrical Coordinates
18 SPE 159172
for Near-Wellbore Phenomena. The University of Texas at Austin, December, Austin, TX.
Proett, M., Waid, M., Heinze, J., and Franki, M., 1994. Low Permeability Interpretation Using a New Wireline Formation Tester "Tight
Zone" Pressure Transient Analysis. Paper presented at the SPWLA 35
th
Annual Logging Symposium.
Proett, M., and Chin, W., 1996. Supercharge Pressure Compensation Using a New Wireline Testing Method and Newly Developed Early
Time Spherical Flow Model. Paper SPE 36524 presented at the SPE Annual Technical Conference and Exhibition, 6-9 October,
Denver, Colorado, USA.
Proett, M., Seifert,D., Chin, W., Lysen, S., and Sands, P., 2004, Formation Testing In The Dynamic Drilling Environment, Paper SPWLA
N, presented at the SPWLA 45th Annual Logging Symposium, J une 69, Noordwijk, The Netherlands.
Rees, H., Foot, J ., and Heddle, R., 2011. Automated Pressure Transient Analysis with Smart Technology. Paper SPE 144327 presented at
the SPE Digital Energy Conference and Exhibition, 19-21 April, The Woodlands, Texas, USA.
Sullivan, S.A., Poston, S.W., and Piper, L.D., 1988. Using Short-Term Pressure Buildup Tests for Reserves Estimation in Tight Gas
Reservoirs. Paper SPE 17707 presented at the SPE Gas Technology Symposium, 13-15 J une, Dallas, Texas.
Shabro, V., Torres-Verdn, C., and J avadpour, F., 2011. Numerical Simulation of Shale-Gas Production: From Pore-Scale Modeling of
Slip-Flow, Knudsen Diffusion, and Langmuir Desorption to Reservoir Modeling of Compressible Fluid. Paper SPE 144355 presented
at the North American Unconventional Gas Conference and Exhibition, 14-16 J une, The Woodlands, Texas, USA.
Snyder, R, 1971. A Review of the Concepts and Methodology of Determining "Net Pay". Paper SPE 3609 presented at the fall meeting of
the Society of Petroleum Engineers of AIME, 3-6 October, New Orleans, Louisiana.
Torres-Verdn, C., Habashy, T., Dussan, E., and Alpak, F., 2004. J oint Inversion of Transient-Pressure and Time-Lapse Electromagnetic
Logging Measurements. Petrophysics, 45, 3.