Professional Documents
Culture Documents
This paper was prepared for presentation at the 2007 SPE Annual Technical Conference and The mechanical and poromechanical properties of Woodford
Exhibition held in Anaheim, California, U.S.A., 1114 November 2007.
shale were measured at four different scales, from field well
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
logs to standard rock testing to the penny-size samples of
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to IDSTD and down to drill cuttings scales using the nano-
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 indentation. Furthermore, the innovative nano-indentation
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
techniques for rock testing have allowed the construction of a
for commercial purposes without the written consent of the Society of Petroleum Engineers is GeoGenomeTM simulation model which can estimate and
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 determine macroscopic rock properties based on porosity,
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, Texas 75083-3836 U.S.A., fax 01-972-952-9435.
packing density, and mineralogy. The simulated moduli and
parameters using this model showed excellent agreement
Abstract when compared to both lab and field log results.
Woodford shale is emerging as one of the major gas
formations in the US Midwest. Despite its tremendous Introduction
potential, existing data on the Woodford shale geomechanics Woodford shale formation, deposited during the lower
characterization are limited at best. In this work, a well in the Mississippian and upper Devonian period in anaerobic marine
Woodford shale formation, 200 feet deep, was cored and environment, is found throughout the central part of US
logged in Oklahoma, USA. The resulting retrieved preserved Midwest. The formation has long been known to be one of the
cores were lab tested using standard acoustic techniques and major source rocks of the region and is recently emerging as
triaxial testing for shale mechanical and poromechanical one of the major gas plays with publicly announced rate up to
characterization in terms of compressibility, strength, pore 10 MMCFD and estimated reserve from 2.0 to 3.0 BCF per
pressure coefficient, Youngs modulus, Poissons ratio, etc. In well.1 The relatively low permeability of this formation,
addition, shale mechanical parameters alteration when in estimated to be 45 nD by pressure pulse decay method, makes
contact with drilling muds and fracturing fluids were hydraulic fracturing a vital process for the production from the
measured using Brazilian tests and the innovative Inclined formation. However, strangely enough, the mechanical and
Direct Shear Testing Device (IDSTD). Finally, mechanical poromechanical properties of the formation which are
Woodford shale parameters were also measured and correlated essential for drilling as well as fracturing operations are little
with field log results, using samples a tiny as drill cuttings (a known. Moreover, the environment sensitive and reactive
few millimeters in size) with the newly emerging nano- shale samples used for mechanical characterization on some
indentation rock characterization techniques developed in the previous studies attempting to characterize this shale gas play
GeoGenome Industry Consortium. This newly developed were collected from a ditch on an outcrop which makes these
methodology for rock testing and shale chracterization, part of studies not comprehensive in terms of field and laboratory
the nanotechnology wave, showed excellent results when experimental preserved shale sample quantity and quality.2
compared with shale acoustic laboratory measurements and
log data and results. Standard Testing
In this geomechanics comprehensive study, 200 ft of well-
Despite a relatively high quartz content as shown by XRD and preserved shale were cored from a well in the Woodford
Elemental Capture Spectroscopy (ECS) log results, the formation located at the Northern flank of the Arbuckle uplift
Woodford shale does exhibit clear transversely isotropic (Pontatoc County, Oklahoma). The preserved cores were
mechanical characteristics, from Youngs moduli to Poissons mechanically characterized through comprehensive standard
ratios and other mechanical parameters. Moreover, IDSTD lab measurements including Ultrasonic Pulse Velocity (UPV),
and Brazilian tensile tests on the preserved Woodford samples triaxial tests, and tensile split tests (Brazilian tests) which
exposed to different drilling and hydraulic fracturing fluids allow obtaining important poromechanical parameters such as
showed that fluid effects play an important role on both Youngs modulus, Poissons ratio, Pore Pressure Coefficient,
compressive strength and tensile strength of the shale despite compressive strength, and tensile strength, in addition to
the fact that the Woodford clay content mainly composes of measurements of shale failure parameters such as cohesion
illite and chlorite. and friction angle.
2 SPE 110120
agreement with average errors in the non-clay, clay, and parameters. One interesting observation is that the Pore
kerogen portion of 4.5%, 5%, and 9%, respectively. Pressure Coefficients, despite the relatively high anisotropy
degree in Youngs moduli and Poisson ratios, show only a
Porosity small deviation between 1 and 3 which for most engineering
Figure 9 illustrates the porosity excluding bounded water and purposes can be neglected. The results also show a reasonable
isolated pores that was derived from the suite of logs. In agreement between the sonic log obtained moduli and UPV
addition to log porosity, 5 core samples were chosen for obtained moduli.
laboratory porosity measurement using density method with
kerogen already removed from the matrix and the results are To ensure the accuracy in measurement, the velocity of
reported in Table 2. Comparison of the lab measured porosity compressional wave propagating at 30o with respect to the axis
with the log derived porosity at the same intervals shows that of symmetry was also measured for samples 131-0 and 185-
the log porosities are consistently greater by 1.4% than the lab 10 and compared with the velocity predicted using obtained
porosities, as shown in Table 3. moduli of the same samples. Figure 16 and Figure 17 show
excellent agreement between the predicted velocities at 30o
Shale Mechanical Characterization Using Logs propagation angle and the measured velocities at the same
The mechanical properties needed for both hydraulic angle, confirming the accuracy of previous measurements of
fracturing and wellbore stability simulation, i.e. Youngs elastic moduli.
modulus, E, Poissons ratio, , and the Pore Pressure
Coefficient, , are obtained from sonic log and the results are 2. Triaxial Test: Quasi-Static Moduli and Compressive
shown in Figure 10 to 12. In addition, the stiffness coefficients Strength
C44, C55, and C66 are also obtained and shown in Figure 13. It One 2 4 sample from depth interval 166-0.5 to 166-5
can be seen from Figure 13 that the stiffness coefficients C44 = was chosen for drained triaxial test at 5000 psi confining
C55 C66 which is a confirmation of the Woodford shale pressure. The optimum axial strain rate which controls the
natural transverse isotropy. The unconfined compressive magnitude of pore pressure buildup at failure and
strength, UCS, can be obtained from log porosity using consequently the accuracy of measured sample strength was
Lashkaripour and Dusseault correlation9 and the results are selected using PCORE-3D simulator, a poromechanics
shown in Figure 14. Other methods of correlating UCS to analytical solution base software developed at the
sonic log and neutron log data can be found in the works of Poromechanics Institute, University of Oklahoma.14-16 Table 6
Onyia and McNally.10- 11 reports the strength, compressibility, Youngs modulus, and
Poissons ratio obtained at 50% strain.
It is noteworthy that the log obtained mechanical properties E,
, , and UCS, not only have large uncertainty on their During loading, four small load and unload loops were
accuracy but also are insufficient for a complete wellbore conducted, as shown in Figure 18, to obtain the small strain
stability simulation which requires the construction of the mechanical properties and the results are shown in Table 7.
Mohrs failure envelope, and the measurements of cohesion Additionally, the P wave and S wave propagating normal to
and friction angle. Unfortunately, correlations for estimation bedding were measured at each 2000 psi increment in axial
of the friction angle, a factor essential for creating the Mohrs stress to evaluate the variation of dynamic moduli with applied
failure envelope, have not been well established.12 In addition, stress and the results are displayed in Table 8. Comparison of
there are no correlations available for estimating the tensile the small strain moduli and the dynamic moduli shows that the
strength and shale fracture toughness, which are vital dynamic moduli can be correlated to quasi-static moduli using
parameters for hydraulic fracturing design and in general a correction factor of 1.15. This correction factor will facilitate
different from the compressive strength. Thus, despite the the use of log obtained moduli in many well bore stability
great convenience of using logs for formation characterization, analysis problems which generally requires the quasi-static
obtaining essential parameters for drilling and hydraulic moduli. The sample after failure displays an angle of 38o
fracturing activity still has to rely on laboratory rock between the failure plane and the sample sides as shown in
mechanics testing and results. Figure 19, suggesting a friction angle of 14 degrees. The
Mohrs failure envelope was then constructed using the
Standard Rock Mechanics Testing ultimate strength and this friction angle as shown in Figure 20.
1. Ultra-Sonic Pulse Velocity Measurements, UPV
Up to this stage, UPV measurements13 were performed on five 3. Tensile Split Test (Brazilian Test): Tensile Strength
samples using setup shown in Figure 15. XRD mineralogy The complete design of hydraulic fracturing requires not only
quantification was also conducted on small pieces extracted the knowledge of elastic moduli and Poissons ratios but also
from these samples. Table 4 summarizes the measured the rock strength, in particular the tensile strength of the
velocities. From these velocities, the dynamic Youngs formation which differs from the compressive strength
moduli, E1, E3, Poissons ratios, 1, 3, shear modulus, G, and obtained from triaxial or uniaxial test. In this work, the
Pore Pressure Coefficients, 1, 3 of the five samples were Brazilian test method17-18 was used to investigate the tensile
obtained and the results are shown in Table 5. Again, the strength of two samples obtained from the same depth of 166-
Woodford shale transverse isotropy characteristics can be 7. Fluid effects on tensile strength were also investigated to
observed from almost all of the obtained moduli and further complete of the characterization. The samples were
both 1 inch in diameter and 0.5 inches in thickness, within the
4 SPE 110120
recommended range by both ASTM and ISRM19-20. The load to the apparent bedding planes of the well core retrieved.
rate was 200 N/s to meet with ISRM standard. The first Figure 23 gives an example of the load-displacement curve
sample was tested after taken out from the PG1 oil which allows the extraction of the indentation moduli.21-22
preservation container. On the other hand, the second sample Table 11 shows the summary of hardness and indentation
was immersed in a very common hydraulic fracturing fluid moduli of the Woodford shale samples. The results show that
used for Woodford shale, well-site available water, 6 hours the anisotropy characteristic can also be observed from both
before testing. The results of the two tests given in Table 9 the hardness and indentation moduli. Results from these tests
shows that the water based fracturing fluid weaken the sample added to the results from over 20,000 indentation tests on
tensile strength by 15% compared to the non-reactive shale and sandstone conducted previously were used to
fracturing fluid. develop a GeoGenome model relating the macroscopic
moduli and parameters such as E, , to the shale mineral
New Rock Mechanics Testing Method composition, packing density, and porosity.
The IDSTD Test: Fluid Effects on Rock Strength
Four tiny cylindrical shale samples with dimensions of 0.8 2. Results and Calibration at All Scales
inches in diameter and 0.3 inches in thickness, were prepared The developed GeoGenome model23 was implemented in
from the retrieved core intervals at166-9.5 to 166-10.5. the Quantitative GeoGenome Simulator, QGGMS, which
The first two samples were subjected to non-reactive mineral not only allows the users to simulate poromechanical
oil (PG1) circulation for three hours and tested at confining properties of shale and sandstones but also capable of
pressures of 1000 psi and 2000 psi to establish the baseline for performing Monte Carlo simulation for the output based on
analysis of drilling fluid effects on sample strength. The other the uncertainties of the input data.24
two samples were subjected to a circulating water based mud,
designated as mud A, for three hours before loading, also Figure 24 to 26 show the comparison between log moduli, lab
under confining pressures of 1000psi and 2000 psi. Testing moduli, and GeoGenome simulated moduli for five samples
results are summarized in Table 10. with UPV measurements. The QGGMS simulated moduli
are calculated based on the XRD mineralogy and the porosity
The Mohr failure envelopes are presented in Figure 21. The obtained from lab measurements. The results show not only
results show that, despite a predominant presence of less excellent agreement between QGGMS simulated moduli
reactive illite and chlorite, the water base mud A clearly with both UPV moduli and log moduli but also the advantage
weakens the shale samples. Consequently, choice of drilling of using simulated GeoGenome model over the sonic log
fluid should deserve a better attention while drilling in deep since the limited setup of the logging sonde only allows E3, 3,
Woodford formations. and 3 to be obtained. Table 12 shows that the average and
standard deviation of the difference between QGGMS
Figure 22 shows the four stress-strain curves with almost simulated E1 and UPV E1 is 0.05 GPa and 2.29 GPa,
identical slopes which confirm similar shale properties of the respectively, while those of E3 are 0.8 GPa and 1.27 GPa.
four samples. It is interesting that although mud A clearly Only UPV moduli and QGGMS simulated moduli from
weakens the strength of the shale, it practically does not XRD mineralogy were compared in Table 12 since log moduli
change the stiffness moduli. are average properties of 2 ft intervals within the log
resolution.
Nano Technology in Shale Testing
1. Nano-Indentations: Shale Mechanical Characteristics Taking a further step to the field application, the log porosity
Shale/rock with its complicated natural composition and its and ECS mineralogy in the form of an LAS file was imported
nano-granular structure4 was studied at their smallest to QGGMS and the simulated moduli was compared with
components using the nano-indentation technique in order to log moduli and UPV moduli as illustrated in Figure 26 and 27.
obtain an understanding of their mechanical behavior and A statistical analysis was conducted for the difference between
characteristics. The results from this scale of testing when sonic log moduli and QGGMS simulated moduli from ECS
incorporated into a statistical approach have made it possible mineralogy since both ECS and sonic log has a resolution of
to study the interrelation between mechanical properties of approximately 2ft. On average the difference between sonic
natural composite material, rock/shale, and its mineral log E3 and 3 and QGGMS simulated values from ECS
composition.5 mineralogy is 1.98 GPa and 0.03 respectively as shown in
Table 13.
In this work, a total of 3000 indentation tests have been
conducted on 10 Woodford cubical shaped shale samples with Conclusions
6mm sides. One face of each sample is subjected to 600 In this work, a comprehensive mechanical characterization of
indentations, using an environment and computer controlled Woodford shale, the major source rock and the emerging gas
CSM nano-indenter in the Department of Civil and play of the Mid West, has been performed at all scales ranging
Enviornmental Engineering at Massachusetts Institute of from log scale to conventional lab testing to nano scale. The
Technology (M.I.T). Before nano-indenting, the samples were mechanical and poromechanical properties obtained from this
polished to a fine 0.1 micron roughness on the indented face. comprehensive study will facilitate future activities in
The indentations were conducted on samples cut and polished Woodford formation:
on two orthogonal directions, parallel with and perpendicular
SPE 110120 5
70
90
110
Depth (ft)
130
150
170
190
210
50 (ft)
Young's Modulus
E3 (Mpsi)
100
Depth (ft)
150 (ft)
150
200
200 (ft)
Cij (Mpsi)
0.2 0.25 0.3 0.35 0.4 0 0.5 1 1 .5
50 50
100 100
C44
D e p th (ft)
D e p th (ft)
C55
C66
150 150
200 200
Figure 11 - Poisson's ratio from acoustic log. Figure 13 - Stiffness coefficients from acoustic log.
70
100 90
110
150
Depth (ft)
D e p th (ft)
130
150
200
170
250
190
210
300
Figure 12 - Pore pressure coefficient, , from acoustic log. Figure 14 - Unconfined compressive strength from log.
10 SPE 110120
= 14o
Figure 25 - Biot's Pore Pressure Coefficients in direction Figure 27 - Young's modulus and Poisson ratio in direction
parallel and perpendicular to bedding planes. perpendicular to bedding planes.
Table 3 Comparison between log measured porosity and lab measured porosity
Depth (ft) Log Porosity Core Lab Porosity Difference
131 0.16 131 0.16 0
155 0.21 154-12 0.19 -0.02
166 0.16 166 0.16 0
175 0.19 175-0.5 0.17 -0.02
186 0.21 185-10 0.18 -0.03
Average of Difference -0.014
Table 6 - Strength, compressibility, and quasi-static moduli of Woodford shale sample at depth of 166'
Sample 166-2 to 166-6
Strength (psi) 18000
E (kpsi) 660
0.3
Compressibility (psi-1) 1.59 x 10-6
14 SPE 110120
Table 7 - Young's modulus from small strain measurements for Woodford shale sample at depth of 166
Axial Stress Cycling
(psi) E (M psi)
1250 1.57
2300 1.53
3700 1.62
5500 1.62
Table 8 Dynamic Young's modulus and Poisson ratio at different axial stresses for Woodford shale sample at depth of 166
Axial Stress Acoustic
(psi) E (M psi)
1250 1.80 0.32
2300 1.81 0.32
3700 1.82 0.33
5500 1.83 0.33
Table 9 - Woodford shale Brazilian tensile strength when exposed to PG1 and tap water
Fluid Type Tensile Strength (psi)
PG1 1180
Tap Water 997
Table 10 Woodford shale IDSTD strength after exposure to PG1 and Mud A
Fluid Sample Conf. P (psi) Deviatoric Strength (psi)
Mud A 1 1000 6309
Mud A 2 2000 7674
PG 1 1 1000 7970
PG 1 2 2000 8493
Table 12 - Average and standard deviation of the difference between UPV and QGGMS moduli from XRD mineralogy
Property Average of Difference Stdv of Difference
E1 -0.058 2.29
E3 0.812 1.27
1 -0.038 0.023
3 -0.006 0.026
Table 13 - Average and standard deviation of difference between sonic log and QGGMS moduli from ECS mineralogy
Property Average of Difference Stdv of Difference
E3 1.98 2.60
3 0.03 0.02