SCA2012-08 1/12

LARGE PRESSURE DEPLETIONS IN ULTRA

DEEPWATER GOM RESERVOIRS CAN SIGNIFICANTLY
REDUCE NEAR WELL BORE PERMEABILITY AND
PORE VOLUME
John L. Shafer1, Kevin Vorhaben2, Greg N. Boitnott3
1
Consultant, Anacortes, WA, USA; 2Noble Energy, Houston, TX, USA
3
NER Inc., White River Junction, VT, USA

This paper was prepared for presentation at the International Symposium of the Society
of Core Analysts held in Aberdeen, Scotland, UK, 27-30 August, 2012

ABSTRACT
The ultra deepwater Gulf of Mexico (GoM) Miocene and Lower Tertiary sandstone fields
have pore pressures that often exceed 17,000 psi and may even exceed 25,000 psi. These
conditions of high initial pore pressures allow for the potential of large pore pressure
depletions to increase well productivity; however they can result in significant reduction
in pore volume and permeability. Decreasing reservoir pore pressure results in a
decreased oil viscosity which off sets, to some extent the reduction in rock permeability.

Laboratory measurements at true reservoir conditions including high pore pressures and
large pore pressure depletions are often available for pore volume compressibility.
However these test conditions are very difficult to meet when measuring transverse
permeability on vertical core plugs. To mimic reservoir conditions, the uniaxial rock
mechanics tests with flow are commonly performed at reduced external stresses by
keeping the pore pressure low and employing the concept of effective stress. However,
the effective stress path cannot duplicate the pore pressure depletion stress path for a
number of reasons. Thus, one can not directly convert permeability reduction obtained in
an effective stress test to reduction in permeability in a pore pressure depletion test.

The focus of this paper is to illustrate how to best predict permeability and pore volume
when optimal measurement conditions of transverse permeability on vertical core plugs
measurements are not available at true reservoir conditions of stress, pore pressure, and
temperature. Creep data are provided to indicate that one may also need to account for
the fact that laboratory measurements are completed on time scales ranging from a few
hours to a few days whereas reservoir field time scales are in terms of decades.

INTRODUCTION
For the Miocene and Lower Tertiary (LT) plays in the ultra deep water (water depths >
5,000 feet) Gulf of Mexico (GoM), wells are drilled to depths in excess of 25,000 feet
(7620m) and some have set depth records of about 35,000ft True Vertical Depth (TVD).
Reservoir engineering challenges include pore pressures that typically exceed 17,000
SCA2012-08 2/12

psia (117 MPa) and may approach 25,000 psi. Appraisal and development wells are very
expensive to drill and complete with costs well in excess of $100 million. Maximizing
well productivity will require large pore pressure depletions and competent rock. Thus
quantifying rock strength, pore compressibility, and permeability at reservoir conditions
of pore pressure depletion are very important.

The high initial pore pressures allow for the potential of large pore pressure depletions
and thus the potential for relatively significant changes in pore volume and permeability
during reservoir pore pressure depletion. These very high pore pressures in the ultra deep
water Gulf of Mexico cause unique challenges for rock and fluid property measurements.
Pore pressure depletion/drawdown reduces permeability due to compaction from increase
in net confining stress, and to increases in pore lining/throat grain volume (differential
swelling) [1, 2]. Previous data indicates that the magnitudes of these two effects are
potentially similar in the GoM LT rocks [3-6].

The intent of this paper is to draw reader’s attention to the potential impact of very high
pore pressures on pore volume and permeability measurements based on the authors’
recent experience. This is supported by prior literature. Our intent is also to draw
attention to the limitations in the capabilities of commercial testing laboratories that
impact reservoir characterization and modeling efforts in these cases.

The literature on the effect of pore pressure on permeability is limited but has been
addressed in a number of studies, with the earliest about 35 years ago.[4] Most of the
prior literature involves measurements made under hydrostatic stress [ 1, 2, 5, 6, 7, 8].
The effect of pore pressure on GoM LT rocks were first reported five years ago [6] with
hydrostatic measurements and three years ago with both hydrostatic and uniaxial
measurements [3, 4]. However, due to limitations in experimental technique, these
uniaxial measurements were conducted with vertical flow on vertical plugs as opposed to
the desired horizontal or transverse flow on a vertical plug which would better simulate
the reservoir flow direction with respect to stress.

Laboratory measurement of brine permeability as a function of confining stress and pore

pressures indicate that for these LT rocks, the routine core permeability measurements at
close to zero pore pressure may underestimate the in-situ rock permeability by 10% to
40% at maximum reservoir stress and pore pressure [3, 4, 5]. Also, the rate of
permeability decline is observed to be greater with pore pressure depletion as compared
to when pore pressure is kept constant (Figure 1 which is a replot of Figure #13 in [3]).

Data Input To Reservoir Simulators

Reservoir simulators require a table/plot of pore volume (PV) and permeability (perm)
[reduction] factors versus reservoir pore pressure. Over the past half dozen years, the co-
authors have been involved in obtaining PV and permeability versus stress for reservoirs
in the ultradeep GoM, Miocene and LT sands. These stiff rocks displayed uniaxial pore
volume compressibilities (PVC) typically between 1 and 4 microsips (1/1E06psi) and
SCA2012-08 3/12

displayed a permeability reduction of about 10% to 60% for an increase in effective axial
stress of 12,000 psi Typically the stiffer the rocks the less the percent of permeability
reduction, but not always. Also the pore volume compressibility was relatively constant
with increasing axial stress for at least the first 6,000 psi of effective drawdown, but for
the second 6,000 psi there sometimes was a marked increase in compressibility (e.g. pore
collapse) with further loading.
1.20
1.05

1.15 1.00
Normalized Permeability

Normalized PV or Perm
0.95
1.10

Pore Pressure Depletion 0.90

1.05

Constant Pore Pressure PV Factor

0.85
Perm Factor
1.00
0.80

0.95
0.75

0.90
0.70
0 2000 4000 6000 8000
0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000
Drawndown (psi) Axial Stress (psi)

Figure 1: Stress path affects permeability response Figure 2: Pore Volume and Perm Factor versus
axial Stress

Pore volume and permeability factor versus axial stress obtained from a uniaxial effective
stress (low pore pressure) compaction test for rock displaying PVC (2 microsips) and
permeability reduction of about 20% for 12,000 psi increase in axial stress is depicted in
Figure 2. For an effective stress (ES) test to simulate the reservoir pore pressure
depletion (PPD) one needs to assume that changes in axial stress can be equated to
change in pore pressure. This is not the case for stiff rocks as we will explain.

The data in Figure 2 were obtained simultaneously during an “effective stress” uniaxial
compaction test (low and constant pore pressure, with increasing axial stress). The use of
low pore pressure for the tests was required because of limitations in experimental
capabilities of Commercial Testing Laboratories. Due to a number of experimental
challenges, transverse permeability measurements during a pore pressure depletion
uniaxial compaction test at reservoir conditions of stresses, pore pressure, and
temperature is not routinely available and attempts to this point have been considered
developmental. While it is generally possible to obtain pore pressure depletion uniaxial
pore volume compressibilities at reservoir stresses, pore pressure, and temperature,
without flow measurements, set-ups for the addition of horizontal permeability
measurements are not established at this time.

Given that the PV/Perm factors versus axial stress or inferred pore pressure data shown in
Figure 2 are not obtained at reservoir conditions, both measurements need to be
converted to such. Rock property measurements of PVC, permeability, velocities, etc are
stress path dependent. Constant pore pressure tests are not equivalent to pore pressure
SCA2012-08 4/12

depletion as grain compressibility effects cause different loading trajectories that cannot
be described by a simple effective stress model. One can only mathematically convert
from one stress path to another if one assumes isotropic linear elastic behavior [3, 5], but
cannot predict the effect of the stress path differences on physical properties such as
velocity, permeability or porosity.

Effective Stress Law

To mimic reservoir conditions uniaxial rock mechanics tests with flow are performed at
low pore pressure and converted to reservoir stresses using the concept of effective stress.
An effective stress law is a means to convert two variables, confining stress and pore
pressure, into one equivalent variable. One such expression would be:

Rock property = f(pore pressure, confining stress) = f(Effective Stress) (1)

Effective stress = confining stress - effective stress coefficient*pore pressure (2)

Every rock property; e.g. permeability, compressibility, sonic velocities, etc., has its own
effective stress coefficient (ESC) [9, 10]. This coefficient is less than 1.0 for most rock
properties, but has been observed to be greater than 1.0 in some cases for permeability.
From linear poroelasticity, the magnitude of ESC for bulk volume is dependent on the
relative magnitude of the bulk modulus (or bulk compressibility) of the rock and grains.
Rocks that have a lower bulk modulus or higher bulk compressibility tend to be
characterized by ESC closer to a value of 1.0 for a wide range of properties [3,5].

Throughout this paper effective stress (low and constant pore pressure) uniaxial (radial
strain is held constant) pore or bulk compressibility are defined in Table 1. The external
stress (a tensor), is denoted by  and the subscript will indicate a type or directional
component (e.g. “a” for axial or vertical and “m” for average of axial and radial). Pp
denotes pore pressure and Vb is bulk volume & Vp is pore volume.

Table 1: Compressibility (microsips) and Resulting ESC Note: Assumes Cg = 0.20

Cpca Cbcm ESCbv Cpcm ESCpv Cpca = (1/Vp)[ Vp/a]
Cbcm = (1/Vb)[ Vb/m ]
1 0.6 0.67 2 0.90
Cpcm = (1/Vp)[ Vp/m]
4 2.1 0.90 8 0.98

Biot’s effective stress coefficient, “Alpha”, is the ESC for total volume change (bulk
compressibility) where:
Alpha or ESCbv = 1- [grain compressibility (Cg)/bulk compressibility (Cbcm)]. (3)

Carroll’s [11] effective stress law for pore volume is where:

ESCpv =1- [grain compressibility (Cg)/pore compressibility (Cpcm)]. (4)
SCA2012-08 5/12

For example using typical values for the lower compressibility (Cpca) samples (about 1.0
microsips) in our data base, if matrix or grain compressibility is 2.0E-07 (1/psi) and the
bulk compressibility to mean stress (Cbcm) is 6.0E-07 (1/psi), then Biot’s alpha or ESCbv
using equation #3 would be 1-(0.2/0.6) = 0.67 (Table 1).

To better illustrate the impact of ESC’s on effective stress PVC tests we have assumed a
set of stresses and pore pressures. Thus, Table 2 would indicate that a ESCbv of 0.67
would imply that the equivalent stresses for an effective stress test (low pore pressure) to
match reservoir stresses would increase the mean effective stress from 2,400 psi (ESC bv
=1.0) to 9,600 psi (ESCbv = 0.67). Thus to simulate 12,000 psi depletion, the axial stress
would increase from 9,600 psi to 21,600 psi.

Table 2. Conversion of Reservoir Stress to Effective Stresses

Reservoir Pressures & Stresses (psi) Effective Stresses (psi)
Pore pres. Total Vert. Total horiz. Total mean ESC Vertical Horiz. Mean
20,000 23,000 22,100 22,400 1.00 3,000 2,100 2,400
20,000 23,000 22,100 22,400 0.98 3,500 2,600 2,900
20,000 23,000 22,100 22,400 0.90 5,000 4,100 4,400
20,000 23,000 22,100 22,400 0.67 9,600 8,700 9,000

The corresponding pore volume compressibility (Cpcm) for the 1 microsip Cpca samples is
about is 2.0E-06 (1/psi), thus ESCpv using equation #4 is 1-(0.2/2.0) = 0.90. As we can
see from Table 2, this would imply that the equivalent stresses for an effective stress test
(low pore pressure) to match reservoir stresses would increase the mean effective stress
from 2,400 psi (ESCpv =1.0) to 4,400 psi (ESCpv = 0.90). Thus to simulate 12,000 psi
depletion the axial stress would increase from 5,000 psi to 17,000 psi. Tables #1 & #2
provides similar information for pore volume compressibility (Cpca) for the higher end
range of our samples, 4.0E-06 (1/psi).

The “standard” procedure for measuring Biot’s α or ESCbv at the beginning of a PPD test
leads to unrealistically high values. Biot’s α varies during a test – assuming incrementally
linear response, as shown in Figure 3A for a PPD PVC test on a rock sample with a pore
volume compressibility of about 1.5 microsips. The high values of bulk compressibility
and thus Biot’s Alpha are at stresses significantly below reservoir stress where core plug
sleeve conformance and strain calibration can be an issue. Thus at stresses above
reservoir stress the average bulk compressibility is about 0.55 microsips and the average
value of Biot’s Alpha is about 0.65 assuming that the grain compressibility is 0.20
microsips. According to Zimmerman [12] a more direct approach to obtain Biot’s Alpha
is given in the equation (εa)/ (-Pp) = α [(εa)/ (σa)]. The slope of axial strain (εa) versus
pore pressure as measured during pore pressure depletion PVC test or inferred during an
effective stress PVC test are presented in Figure 3B. The value of Biot’s Alpha
SCA2012-08 6/12

obtained from the ratio of these slopes is 2.08/2.94 or 0.7 that agrees well with that
calculated from the average value of the bulk compressibility in the PPD PVC test.
0.006
3.5 1.0

Effective Stress

2.8 0.9 PPD

0.005
Bulk Compresssibility (1/10 6 psi)

Cbcm Linear (Effective Stress)

Axial Strain
Linear (PPD)
Biot's Alpha

Biot's Alpha
2.1 0.8
0.004
y = -2.08E-07x + 6.70E-03
R2 = 9.95E-01
1.4 0.7

0.003
y = -2.94E-07x + 7.82E-03
0.7 0.6
R2 = 9.95E-01

0.002
0 0.5 20,000 18,000 16,000 14,000 12,000 10,000 8,000
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000
Effective Mean Stress (psi)
Pore pressure (measured or inferred) psi

Figure 3A: Bulk compressibility and Biot’s Figure 3B: Calculation of Biot’s Alpha from slopes
Alpha versus mean stress of axial strain versus pore pressure

To correctly apply an effective stress law, it is commonly assumed that one needs to
design the experiments to simulate true reservoir stress conditions for the desired
property. If one is most interested in PVC then ESCPV would be 1-(Cg/Cpcm). For higher
compressibility samples (Cpca ~ 4 microsip), ESCpv is expected to be close to 1, while for
lower compressibility samples (Cpca ~ 1 microsip), ESCpv can be significantly less than
1. Thus only for the lower compressibility samples (Cpca approaching 1 microsip) is there
a clear reason to apply effective stress coefficients in calculation of what stresses to use
in conducting a PVC test. For a new field, both bulk and pore volume compressibility
would not be known at the start of a rock mechanics program, thus one needs results from
several tests to correctly design remaining tests. Obviously if one simulates reservoir
behavior with uniaxial pore pressure depletion tests, then the uncertainties introduced by
application of effective stress laws do not come into play.

A few words of caution about attempting to predict a compressibility for one protocol
(e.g. pore pressure depletion) based on test results from another (effective stress, low pore
pressure). To do so requires adopting a model allowing one to predict deformations for
the loading path of interest. If one wants to assume that the elastic constants are isotropic
and independent of the loading path, then the poro-elasticity equations should be fine if
used properly. Our experience with consolidated sandstones is that the linear elastic
model typically does a good job of representing measured data provided stresses do not
approach the yield surface for the material [3, 5]. However, if you find for example that
the elastic constants measured using pore pressure depletion are systematically different
from the elastic constants measured during constant pore pressure testing (and/or if there
is anisotropy in elastic constants), the analysis needs to be done differently.

Pore Volume Compressibility

When flow and PVC are combined together in a single test, it is not possible to directly
SCA2012-08 7/12

measure the change in pore volume to directly calculate pore volume compressibility. In
such cases, pore volume strain and pore volume compressibility are commonly calculated
based on the bulk strain measurements subject to corrections to account for the effects of
grain compressibility (e.g. Zimmerman [12]). For the effective stress test converting
measured bulk compressibility to effective stress pore volume compressibility does not
provide the pore pressure depletion equivalent PVC which requires a separate conversion
that is also documented by Zimmerman [12].

Time Dependent Compaction (Creep)

During the past seven years on several projects, the authors [6] have utilized PVC testing
protocol such that at the end of the effective stress or pore pressure depletion PVC test,
representing about 12,000 psi depletion, the stresses and pore pressure were held constant
for up to 48 hours to record the strain. The measured change in pore volume during this
creep portion of the PVC test, while small, is still a significant fraction of the total non-
creep PV change. As shown in Figure 4, the total volumetric strain from the start of the
PVC test increased by about 32% in one day of creep. In 30 years of reservoir life total
volumetric strain would be extrapolated to increase by about 190%, thus nearly tripling
the delta PV observed during the non creep portion of the laboratory test. Thus we found
that accounting for creep at maximum depletion could increase the total volumetric strain
by about 50% (Cpca ~ 1E-06 psi-1) to 200% (Cpca ~ 4E-06 psi-1) and in some cases the
samples even crept into failure. The two parameter creep model [13] used to fit the creep
data shown in Figure 4 is given in equation #5.

100(creep ΔPV/total non creep ΔPV) = constant*Log10(1+creep duration/time constant) (5)

Thus accounting for creep may potentially triple the PV factor at large pore pressure
depletions (12,000 psi) in terms of reservoir time scale and because of the link between
PV and perm, this time dependent strain has potentially to significantly impact the
permeability reduction. While this model does an excellent job of fitting the observed
creep data for up to 48 hours, there is no guarantee that it will be accurate out to 11,000
days or about 30 years.

The magnitude of creep is dependent upon the temperature, stresses, and the fluid
saturations [13]. At elevated temperature and high pore pressures and stresses, these GoM
stiff rocks have the potential to experience stress corrosion cracking interaction (and/or
other rock-water interactions), resulting in higher PVC then 100% oil saturated samples
or tests conducted at ambient temperatures whether oil or brine saturated. Previous
studies by co-authors [6] has indicated that either 100% brine or connate water saturation
are equally effective at introducing these chemical effects on creep. The magnitude of
creep is also significantly impacted by the magnitude of the pore pressure depletion. The
relative contribution of creep at 6000 psi pore pressure depletion is about a third of that at
12,000 psi pore depletion. Also it is our observation that the magnitude by which the
pore volume compressibility increases (non-linear) with axial stress (ES test) or pore
pressure (PPD) correlates to the magnitude of the creep at the end of the loading ramp.
SCA2012-08 8/12

100(creep ΔPV/total non creep ΔPV) = constant*Log10(1+creep duration/time constant)

40
1.02
Creep Volumetric Strain (% of pre-crrep strain)

35
1.00
Creep data
30
Creep model 0.98

25

Perm Factor
0.96

20 0.94

15 0.92 Creep

10 0.90

5 0.88

0 0.86
0 0.2 0.4 0.6 0.8 1 1.2 0.988 0.990 0.992 0.994 0.996 0.998 1.000
PV factor
Creep Time (days)

Figure 4: Creep Modeling Figure 5: PV Factor versus Perm Factor

It is possible to obtain both PVC and permeability data with an effective stress uniaxial
PVC test at ambient temperature. As just mentioned, the measurement of creep at the
conclusion of an ambient temperature PVC test does not provide an accurate
measurement of creep (time dependent compaction) at reservoir conditions. Thus the
preferred and direct approach is to obtain pore pressure depletion PVC data with no flow
data at reservoir conditions of stresses, pore pressures, and temperature with core sample
at connate water saturation or 100% brine saturation and measure creep at the conclusion
of the PVC test. If this is not possible then a reservoir temperature effective stress PVC
test should be used. This approach requires effective stress PVC and permeability
measurements at ambient temperature on a companion core plug to obtain relevant
permeability data versus axial stress.

Permeability Factor
Because commercial testing laboratories typically can only measure permeability at low
and constant pore pressure, the measurements reported here were obtained from an
effective stress PVC test at constant pore pressure as illustrated in Figure 2. The main
experimental difficulty in measuring transverse permeability during a PPD test is being
able to port the inlet and outlet flow lines through core plug jacket in such a way to
maintain a well defined geometry of the flow boundary condition that is not affected by
changes in effective stress. This is not as much as an issue when the flow axis coincides
with the long axial of the core plug.

PV Factor versus Permeability Factor

A nearly linear relationship is observed if normalized permeability factor is plotted
against PV factor from the effective stress PVC data (Figure 5). Such a cross-plot is a
good QC for detecting if the viscosity of the oil (working fluid) used in flow
measurement is correctly accounting for temperature changes in the core sample. With
laboratory temperature fluctuations observed of nearly +/- 2F and a 1% change in
viscosity per 1°F, this small temperature driven viscosity variation has a large impact of
SCA2012-08 9/12

the reported perm when the perm reduction is only 10% for a 12,000 psi increase in axial
stress.

The PV Factor versus Permeability Factor for the creep portion of the PVC test is
highlighted in Figure 5. If one assumes that this PV to permeability relationship (Figure
5) is constant for the loading ramp, we can predict the Perm Factor for creep behavior to
account for the difference in time scale between lab tests and reservoir drawdown. A
close inspection of the plot of PV factor versus Permeability Factor shows that the slope
of the creep portion has a slightly steeper slope than the depletion portion, and thus this
assumption may understate permeability decline during creep.

RESULTS
Correcting PV and Permeability Factors to Reservoir Conditions
If both PV factor and Permeability factor are obtained from a single effective stress test,
then the PVC calculated from the measured Cbca will need to be converted to the
equivalent PPD PVC. This conversion typically represents about a 10 to 20% increase in
PVC and thus an equal magnitude decline in PV Factor will occur. If PV factor is
obtained from the preferred pore pressure depletion PVC test at reservoir temperature
then no conversion is required.

The second correction and the one with the far greater impact is for time dependent
portion of the compaction (creep) which has the potential to triple the predicted pore
volume compressibility and, by inference, the permeability reduction. However, as
already stated, creep data should come from PVC tests (either effective stress or
preferably PPD) conducted at reservoir temperature using core samples at connate water
saturation or 100% brine saturated. The magnitude of the creep contribution to the
effective PVC is dependent on the extent of pore pressure depletion. Thus our approach
has been to assume that the contribution of creep to the pore volume compressibility
exponentially increases from zero at zero depletion to the measured creep at 12,000 psi
depletion (Figure 4 and Table 3 within Figure 4). To be conservative we have only use
the projected creep out to a reservoir life of 100 days, about 100% of the pre-creep strain,
which is about 50% of the total creep for 30 years based on data presented in Table 3.

The permeability and PV factors presented in Figure 2 that were obtained from an
effective stress test at ambient temperature have been corrected for reservoir conditions
of temperature, stress path (PPD), and reservoir time scale. These corrections are based
on the methodology presented in this paper. Figures 6A and 6B summarize these
corrections. Figure 6A is a plot of inferred reservoir pressure versus PV Factors:
uncorrected PV Factor is corrected for increase in PVC resulting from ES to PPD
conversion, and then added to this creep correction. Figure 6B is a plot of inferred
reservoir pressure versus Permeability Factors; uncorrected Permeability Factor,
corrected for increase in PVC resulting from ES to PPD conversion, adding to this creep
correction, and finally adding in correction for ES to PPD permeability decline. These
estimates of corrections required on typical lab measurements are provided to focus
SCA2012-08 10/12

attention on the importance of obtaining accurate pore volume and permeability data at
reservoir conditions of temperature, stresses, and pressures and account for the difference
in time scale between lab and reservoir.

1.01
1.25

1.00

Perm or Viscosity Factor

0.99 1.00

PV Factor
0.98

0.75
0.97
Effective Stress
Effective Stress PPD
0.96
Pore Pressure Depletion (PPD) PPD + Creep 0.50

PPD + Creep PPD + Creep + Pp

0.95
Crude Oil Viscosity
Transmissibility
0.94 0.25
21,000 19,000 17,000 15,000 13,000 11,000 9,000 7,000 21,000 19,000 17,000 15,000 13,000 11,000 9,000 7,000
Inferred Pore Pressure (psi) Inferred Pore Pressure (psi)

Figure 6A: PV Factor versus Pore Pressure Figure 6B: Perm Factor versus Pore Pressure

Figure 6B also plots the relative viscosity response of a typical crude oil with decreasing
pressure but assuming constant reservoir temperature. For the first 10,000 psi pore
pressure depletion, the viscosity is decreasing faster than the permeability decline for this
example and thus the transmissibility of the rock is greater than at in-situ conditions.
Figure 6B also plots the relative transmissibility (permeability decline versus viscosity
decline) that indicates that the relative transmissibility peaks at about 7,000 psi pore
pressure depletion.

The stress path impact on permeability response correction was the subject of several
papers mentioned in the introduction. In some cases, we have found that the rate of
permeability decline for a pore pressure depletion tests was twice that of an effective
stress test (Figure 1). The core material that was used for this permeability versus stress
& pressure tests cited in these references is from the same sand unit as those for Figures
4, 6A & 6B. However, instead of doubling the rate of permeability decline we have
elected to use 150% decline. Please recall as stated in the introduction that these
laboratory tests had measured the change in vertical permeability during a uniaxial PVC
test on a vertical plug as opposed to a transverse permeability which would more
accurately represent reservoir conditions. Since non-hydrostatic changes in stress (e.g.
a  r ) will generally result in non-isotropic changes in the petrophysical properties
such as permeability, we have elected to discount the correction by half to be
conservative.

SUMMARY
The primary objective of this paper has been to illustrate that PV and Perm Factors
obtained from an ambient temperature effective stress (low and constant pore pressure)
uniaxial PVC test may greatly underestimate the PV and permeability decline during pore
SCA2012-08 11/12

pressure depletion that would actually occur in the reservoir at large drawdowns that
would be experience by the near-wellbore region.

We have shown that the two largest potential corrections to effective stress data that
impacts permeability decline are creep and the effect of pore pressure. Both of these are
likely to be influenced by rock mineralogy and location of ductile grains. Obviously
reservoir temperature PPD PVC laboratory tests with simultaneous transverse flow
measurements for permeability best simulate the reservoir conditions. If this is not
available then at a minimum an ambient temperature ES PVC test to obtain permeability
data coupled with a reservoir temperature PPD PVC lab test with creep but no flow
measurements is required to obtain effective pore volume compressibility data that
simulate reservoir time scale.

Given that we have very limited data on the magnitude of the ESCperm, [14] and it will be
rock specific, then the potential impact of pore pressure on permeability might best be
treated with a sensitivity analysis until such data does become more readily available.

Future Challenges
The following items are identified as deserving attention by the Core Analysis/Rock
Mechanics Service companies:

 There is a great need for transverse permeability measurements during PPD PVC tests
so estimated correction factors are not required.
 There is a need to obtain creep data versus extent of pore pressure depletion to project
PV factors out to reservoir time scale at all magnitudes of drawndowns.
 There is a need to extend ESCperm to inelastic regime (creep, pore collapse).
 Hopefully sometime in the near future we will be able to simulate with digital rock
data the impact of pore pressure on permeability that we can not conveniently
measure in the lab at true reservoir conditions.

Related Issue
Rock mechanics tests to obtain data for well completion are all effective stress tests with
zero pore pressure and typically assume that Biot’s effective stress coefficient, alpha, is
very near one, but the measured value at reservoir stress conditions are potentially
significantly lower, 0.7. This could potentially impact well completions rock mechanics
data interpretations. It should also impact the calibration of the log derived Poisson’s
ratio and Young’s modulus since compressional and shear velocities have different
ESC’s and thus effective stress collection of static and dynamic properties will not reflect
reservoir conditions of stresses and pore pressures.

ACKNOWLEDGEMENTS
The authors thank Noble Energy and NER for their support and approval to publish this
paper. Shafer and Boitnott acknowledge Terry Miller’s contributions during past decade.
SCA2012-08 12/12

REFERENCES
1. Al-wardy, W. and Zimmerman, R.W., 2004, “The Effective stress law for the
permeability of clay-rich sandstones”, J. of Geophy. Res., 109, B04203, 10 pages.
2. Warpinski, N.R., and Teufel, L.W., 1992, “Determination of the Effective Stress Law
for Permeability and Deformation in Low-Permeability Rocks,” SPE Formation
Evaluation, June, 123-131.
3. Boitnott, G. N., Miller, T. W., and Shafer, J. L., 2009, Pore-Pressure Depletion and
Effective Stress Issues in the Gulf of Mexico's Lower Tertiary Play, SPE 124790-PP,
SPE Annual Technical Conference and Exhibition held in New
Orleans, Louisiana, USA, 4–7 October 2009
4. Boitnott, G. N., Miller, T. W., and Shafer, J. L., 2009, Pore-Pressure Depletion Effects
and Permeability: Effective Stress Issues for High Pressure Reservoirs, SCA2009-05,
International Symposium of the Society of Core Analysts, Noordwijk, The Netherlands,
27-30 September.
5. Shafer, J., Boitnott, G., and Ewy, R., 2008, Effective Stress Laws for Petrophysical
Rock Properties, paper GG, Proceedings SPWLA 49th Annual Logging Symposium,
Edinburgh, Scotland, May 25-28
6. Shafer, J. and Fate, T., 2007, Coring And Core Analysis: Challenges Of Offshore Ultra
Deep Water Reservoirs, SCA2007-21, International Symposium of the Society of Core
Analysts,Calgary, Canada, 10-12 September
7. Zoback, M.D. and Byerlee, J. D., 1975, Permeability and effective stress, AAPG Bull.
V59, 154-158
8. Walls, J., and Nur, A., 1982, Pore pressure and confining pressure dependence of
permeability in sandstone, Proceedings of the 7th Formation Evaluation Symposium of
the CWLS, paper O, Canadian Well Logging Society, Alberta, Calgary.
9. Berryman, J. G. 1993, `` Effective-stress rules for pore-fluid transport in rocks
containing two minerals,'' Int. J. Rock Mech. Min. Sci. & Geomech. Abst. 30, 1165-1168.
10. Robin, P., 1973, “Note on Effective Pressure, Jr. of Geophysical Research, 78, n14,
pp2434-2437.
11. M. M. Carroll, 1980, Mechanical response of fluid-saturated porous materials. Proc.
Of the 15th Int. Cong. of theoretical & applied mechanics, Toronto. pp 251-262.
12. Zimmerman, R., 2000, Implications of Static Poroelasticity for Reservoir
Compaction, Proc. 4th North Amer. Rock Mech. Symp., A.A. Balkema, Rotterdam, pp.
169-72.
13. Schutjens, P.M.T.M, Blanton, T.L., Martin, J.W., Lehr, B.C., and Baaijens, M.N.,
1998 “Depletion –induced compaction of an overpressure reservoir sandstone: An
experimental approach,” SPE/ISRM 47277. Eurock ’98, Trondheim, Norway, 8-10 July,
pp53-65
14. Ewy, R.T., Bovberg, C.A., Hagin, P.N. and Shalz M.L., 2012, Permeability
Measurement with Ultra-High Fluid Pressure and Unequal Stresses, ARMA 12-656, 46th
US Rock Mechanics / Geomechanics Symposium held in Chicago, IL, USA, 24-27 June
2012.