SPE-170972-MS

A Practical Petrophysical Approach for Brittleness Prediction from

Porosity and Sonic Logging in Shale Reservoirs
Xiaochun Jin,1 and Subhash Shah, the University of Oklahoma; Jerome Truax, Linn Operating;
Jean-Claude Roegiers, the University of Oklahoma

Copyright 2014, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Amsterdam, The Netherlands, 27–29 October 2014.

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
Brittleness has been used as one of the important descriptors for the characterization of unconventional
shale reservoirs. The degree of brittleness in shale reservoirs is determined upon the basis of its
mineralogical composition, which can be obtained from mineralogical logging tools (such as ECS™,
FLeX™, GEM™, Litho Scanner™), or XRD tests in the laboratory. Generally, measurements of miner-
alogical brittleness are obtained from physical sources and lead to relatively reliable interpretation results.
However, mineralogical logging is expensive and not commonly available in the shale play. Alternatively,
brittleness can also be calculated from dynamic Young’s modulus and Poisson’s ratio, but the absence of
shear slowness in some wells restricts its wide application. Internal friction angle based brittleness can
give similar interpretation results as the preceding two methods, but its accuracy depends highly on the
quality of correlations. It is observed that the curves of the three different brittlenesses demonstrate similar
shapes. Therefore, we have attempted to build correlations between mineralogical brittleness and porosity
or sonic compressional slowness for typical shale plays (Woodford, Barnett, and Eagle Ford shale), and
have proven their validity with the data obtained from wells not included in the development of
correlations. Applications of these findings include: (1) enabling the possibility of evaluating brittleness
in plays lacking mineralogical and shear sonic loggings, thus reducing the quantity of laboratory testing,
(2) inspiring operators to develop in-house correlations of brittleness for shale gas plays, and (3)
investigating similar correlations in emerging unconventional oil plays, such as Granite Wash.
Keywords: Brittleness · Unconventional shale · Hydraulic fracturing · Elemental logging · Tight oil

Introduction
Hydraulic fracturing has played a significant role in commercializing tight reservoirs, but field experience
suggests that not all targets rich in TOC (Total Organic Content) are good fracturing candidates [Britt and
Schoeffler, 2009; King, 2012]. In order to maximize the efficiency of hydraulic fracturing, integrated

1
Xiaochun Jin completed this work when he was Ph.D. student at the University of Oklahoma in Norman. He is a Geomechanics Specialist at BP Global Wells
Organization in Houston.
2 SPE-170972-MS

Figure 1—Brittle and ductile rock behavior from post failure curve of stress versus strain. This is one of the direct methods to distinguish between
brittle and ductile rocks. Brittleness is difficult to quantify and uneconomical for large sample sizes.

geomechanics, petrophysics and well logging analysis have been applied in unconventional reservoir
characterization. The major parameters included in this integrated analysis are: TOC, thermal maturity,
mineralogy, fracture density and orientation, thickness, porosity, permeability, relative permeability,
saturation, density, PVT behavior, static and dynamic geomechanical properties (such as Young’s
modulus and Poisson’s ratio), uniaxial and triaxial compressive strength, internal friction angle, cohesion,
Biot’s coefficient, tensile strength, fracture toughness, closure stress, proppant embedment, propped and
unpropped fracture conductivity, brittleness, shale stability and compatibility with injected completion
fluid. Brittleness is selected here as one critical parameter for investigation. In recent years, brittleness has
been included as an essential parameter in petrophysics reports on shale reservoirs. Improvement of
brittleness evaluation in shale formation can help better appraise its potential to be a good measure for
hydraulic fracturing [Rickman et al., 2008], However, there is no universally standardized and accepted
definition or measurement method for brittleness [Yang et al., 2013].
Brittleness is defined as a material property that ruptures or fractures with little or no plastic flow
[Jackson, Mehl and Neuendorf, 2005]. Additionally, materials with higher brittleness exhibit the follow-
ing characteristics [Hucka and Das, 1974; Singh, 1986; Jarvie et al., 2007; Rickman et al., 2008; Wang
and Gale, 2009; Yarali and Kahraman, 2011; Heidari et al., 2013]:
● Failing immediately at peak stress: from the stress-strain tests, it is observed that the brittle rock
fails immediately when peak stress reaches rock strength, while the ductile rock does not fail
immediately, but continues to absorb energy, and finally fails at large strain, as is shown in Fig.
1.
● Low values of elongation: when brittle and ductile samples with the same dimensions are pulled
under the same tensile stress, the brittle sample is elongated less than the ductile one before failure.
● Fracture failure: when brittle and ductile samples with the same dimensions are struck by the same
hammer with the same force, the brittle sample fractures easier than the ductile one.
● Formation of fine particles: when brittle and ductile samples with the same dimensions are struck
by the same hammer at the same force, fine particles are formed more easily in brittle sample.
● Higher ratio of compressive and tensile strength: brittle materials exhibit low tensile strength and
high compressive strength, which means they fail immediately once stress reaches the rock
strength.
● Higher resilience: a deformed sample with high resilience will have difficulty recovering its
original shape, so this type of material is brittle.
SPE-170972-MS 3

● Higher internal friction angle: the internal friction angle defines the ease with which a material can
slip along the fracture surface. The lower the internal friction angle, the more easily the material
can slip along the fracture surface. The brittle rock exhibits high internal friction angle that means
it does not slip easily.
● Formation of cracks during indentation: given brittle and ductile samples with the same dimen-
sions, cracks are more likely to form in brittle rock during an indentation test.
● Higher Young’s modulus and lower Poisson’s ratio: brittle material experiences less axial strain
and lateral strain because of its high Young’s modulus and low Poisson’s ratio.
● Higher percent of quartz and carbonate: quartz and carbonate are brittle minerals, and formations
with a high percentage of brittle minerals are prone to be brittle.
● Lower porosity: Internal friction angle can be correlated with porosity. Low porosity corresponds
to high internal friction angle, thus high brittleness.
● Drill cuttings: drilling in a brittle formation generates larger cuttings.
● Big coal blocks in coal mining. It is observed that big coal blocks form easily in a brittle coal field.
● Lower rate of penetration in drilling: drilling in brittle formations creates large cuttings, which are
difficult to transport to surface. In addition, brittle formations are much stronger, and are not easily
drilled; thus, high brittleness reduces rate of penetration.
There are more than twenty formulations for brittleness in the literature, which are summarized in
Table A-1. Generally, measurements of mineralogical brittleness are obtained from physical sources and
lead to relatively reliable interpretation results. Although there are disagreements in brittleness definitions,
researchers have attempted to seek some practical and reliable definitions [Rickman et al., 2008; Britt and
Schoeffler, 2009]. Results from geomechanical experiments have shown that the brittleness values of the
same sample by different definitions can vary vastly because these definitions are not uniform and
outdated [Yang et al., 2013]. The mineralogical brittleness ignored calcite and dolomite, while Young’s
modulus and Poisson’s ratio based brittleness was not properly normalized. It is impossible to have the
same or similar values of brittleness by different methods because they are based on different physical
mechanisms. However, it might be possible to obtain similar interpretation results by independent
methods.
The purpose of brittleness evaluation is to identify which formation is of low, medium, and high
brittleness, whether the shapes of logging curves obtained by different definitions are similar for the same
well. For instance, it was proven that internal friction angle based brittleness B18, Young’s modulus and
Poisson’s ratio based brittleness B19, and mineralogical brittleness B22 gave logging curves with different
magnitudes, but similar shapes for the same Barnett shale well, which ensured similar interpretation
results, as shown in Fig. 2 [Jin et al., 2014]. Similar results are observed in other shale plays, such as
Woodford, Eagle Ford, Granite Wash, etc. This benchmark work proves the feasibility to interpret
brittleness by different methods. In addition, brittlenesses B18, B19, B22 are relatively easily accessible2 in
the field when compared with rock mechanical methods. One can use one, two, or all of them to
characterize rock brittleness, depending on what the data are availability.
Because the development of brittleness correlations uses B18, B19, B22, their formulas are included in
the following.
The internal friction angle based brittleness B18 is defined as
(1)

where ␸ is the internal friction angle.

2
Internal friction angle is included in LAS file as “FANG”; Weight fractions of minerals can be obtained from mineralogical logging or XRD tests; Young’s
modulus and Poisson’s ratio can be calculated with density logging, compressional and shear slowness.
4 SPE-170972-MS

The Young’s modulus and Poisson’s ratio based

brittleness B19 is defined as
(2)

where E n and ␯ n are normalized dynamic

Young’s modulus and Poisson’s ratio, and are de-
fined below:
(3)

(4)

where Emin and Emax are the minimum and max-

imum dynamic Young’s modulus for the investi-
gated formation, vmin and vmax are dynamic mini-
mum and maximum Poisson’s ratios for the
investigated formation, E and v are dynamic
Young’s modulus and Poisson’s ratio along the
depth. Emin, Emax, vmin, vmax are constants, and E and
v are variables. B19 indicates that formation with
higher Young’s modulus (E) and lower Poisson’s
ratio (v) is of higher brittleness.
The mineralogical brittleness B22 is defined as3
(5)
Figure 2—Comparison of different brittleness indices for Barnett shale
(the major unit in the vertical direction is 100 ft)
where WQFM/WTot is the weight fraction of sili-
cates (quartz, feldspar, and brittle mica); WCarb/WTot
is the weight fraction of carbonate minerals4 consisting of dolomite, calcite, and other brittle carbonate.
Although the three brittleness definitions B18, B19, B22 have proven to be applicable for brittleness
evaluation, sometimes there is a shortage of logging data due to expense, tool limitation, and environ-
mental restrictions. Mineralogical logging is one of the most expensive logging services, and is not
available for all wells. In addition, a complete set of sonic logging data is not commonly available:
sometimes there is only compressional sonic slowness, but no shear slowness because it is more
expensive. Furthermore, the internal friction angle in the LAS file is not measured directly, but derived
with correlations. The quality of correlations will affect the accuracy of B18. Therefore, it is of great value
to build correlations between mineralogical brittleness and other less-expensive and more commonly
available parameters.
It has been proven from laboratory studies that brittleness decreases with increasing porosity [Heidari
et al., 2013]. In addition, the internal friction angle is correlated with porosity and compressional slowness
[Weingarten and Perkins, 1995; Chang, Zoback and Khaksar, 2006]. Compressional slowness can be
correlated with porosity [Mavko, Mukerji and Dvorkin, 2009], and rocks of different porosities exhibit
different compressional slownesses. Furthermore, Fig. 2 proves that curves of B18, B19, B22 are similar.
Therefore, it is possible to correlate mineralogical brittleness with porosity and compressional slowness.

3
Weight fractions of silicate minerals, carbonate, dolomite, and calcite are indicated as “WQFM”, “WCAR”, “WDOL”, and “WCLC”, respectively.
4
In LAS file, for the carbonate minerals, if there are WCARB, WDOL and WCLC, WCARB represents all other carbonate minerals except calcite, and all of
them should be included in brittleness calculation; if there are only WCLC and WDOL, but no WCARB, include both of them in brittleness calculation.
SPE-170972-MS 5

The objectives of this paper are to: (1) develop new correlations of mineralogical brittleness with
neutron porosity or compressional slowness for Woodford, Barnett, and Eagle Ford shales, including two
global correlations with data obtained from these three typical shale reservoirs; (2) verify the new
correlations with the mineralogical brittleness of independent wells; (3) verify whether similar correla-
tions exist in other shale plays, such as Granite Wash, other Woodford shale wells, etc.; and (4) develop
the ability to predict brittleness of shale wells without shear slowness and mineralogical loggings.

Prediction of Brittleness with Neutron Porosity for Typical Shale

Reservoirs
Porosity is an important parameter in reservoir development, and is divided into two categories: total
porosity and effective porosity [Tiab and Donaldson, 2011]. Fluid flow in porous medium is governed by
effective porosity, but rock mechanical properties, such as Young’s modulus, Poisson’s ratio, brittleness,
etc. are (or partly) controlled by total porosity. Because internal friction angle is correlated with porosity
[Chang, Zoback and Khaksar, 2006], and increasing porosity reduces brittleness are found from the
laboratory study [Heidari et al., 2013], a correlation between brittleness and porosity may be developed
if sufficient source data is available.
Both neutron and density porosities are available in most LAS files. In shale reservoirs, it is found that
mineralogical brittleness is not correlated well with the density porosity (DPHI), but is associated well
with the neutron porosity (NPHI). In the present study, three separate correlations between mineralogical
brittleness and neutron porosity are developed for Woodford, Barnett, and Eagle Ford shales. One global
correlation between mineralogical brittleness and neutron porosity with available data from the three shale
plays is also developed. The LAS files also include enhanced neutron porosity data. It is observed that the
enhanced neutron porosity is also linearly correlated with neutron porosity. Once the correlation between
brittleness and neutron porosity is obtained, one can derive the correlation between brittleness and
enhanced neutron porosity. In the following sections, the “porosity” means neutron porosity, and
“brittleness” means mineralogical brittleness B22.

Correlation of Brittleness with Porosity for Woodford Shale

Woodford shale is located in south-central Oklahoma, covering approximately 11,000 square miles. The
burial depth ranges from ~6,000 ft to ~11,000 ft. The thickness ranges from ~120 to ~220 ft across the
play. It is a Devonian-age shale bounded by limestone above and undifferentiated strata below [Council
and Consulting, 2009]. The Woodford shale formation can be divided into Upper, Middle, and Lower
members [Hester, Schmoker and Sahl, 1990]. Well logging data from four wells of Woodford shale were
selected for developing the correlation between mineralogical brittleness and neutron porosity, which was
plotted in Fig. 3.
The vertical resolution of the logging data is 0.5 ft. Because the lithology may not change every 0.5
ft, the tracks of raw logging data are averaged over every 1.5, 3.0, 5.0, 10.0, 15.0, and 20.0 ft. The average
processing method can smooth the curves, ensure similar interpretation results, and help us get better
correlations. After examining correlation quality and similarity between processed and raw logging
curves, 10 ft interval was selected for data processing. Similar data processing methods was also applied
to the development of correlations for other plays.
The correlation of brittleness with porosity for Woodford shale then was then developed,
(6)

where ␾ is neutron porosity, R2 is coefficient of determination.

B23 is a good correlation in terms the value of R2.
6 SPE-170972-MS

Figure 3—Correlation of brittleness with porosity for Woodford shale.

Figure 4 —Correlation of brittleness with porosity for Barnett shale.

Correlation of Brittleness with Porosity for Barnett Shale

Barnett shale is located in the Fort Worth Basin of north-central Texas, covering about 5,000 square miles.
The burial depth varies from ~6,500 ft to ~8,500 ft. The average thickness ranges from ~100 to ~600 ft
across the play. It is a Mississippianage shale bounded by Marble Falls limestone above and Chappel
Limestone below [Council and Consulting, 2009]. The Barnett shale formation can be divided into Upper
and Lower members [Fisher et al., 2002]. Well logging data from three wells of Barnett shale were
selected for the development of correlation between mineralogical brittleness and neutron porosity, and
were plotted in Fig. 4.
The correlation of brittleness with porosity for Barnett shale is
(7)
SPE-170972-MS 7

Figure 5—Correlation of brittleness with porosity for Eagle Ford shale.

B24 is an excellent correlation in terms of high value of R2.

Correlation of Brittleness with Porosity for Eagle Ford Shale

Eagle Ford shale is a late Cretaceous-age shale located in south Texas with thickness ranges from ~100
to ~300 ft. The burial depth varies from ~4,000 ft to ~14,000 ft. It can be divided into the Upper and Lower
Eagle Ford, and the Upper Eagle Ford is thicker than the Lower Eagle Ford. It is bounded by the Austin
Chalk above and Buda formation below. Both gas and oil are produced in Eagle Ford Shale, but only gas
play is studied here. The high fraction of carbonate in shale makes it brittle and a good fracturing
candidate [Inamdar et al., 2010]. Well logging data from five wells of Eagle Ford shale were selected for
developing the correlation between mineralogical brittleness and neutron porosity, and were plotted in
Fig. 5.
The correlation of brittleness with porosity for Eagle Ford shale is
(8)

B25 is an excellent correlation in terms of high value of R2.

Prediction of Brittleness with Compressional Sonic Slowness for Typical

Shale Reservoirs
The equation for B19 was derived with density, compressional and shear slowness. The tracks of B19 and
B22 are similar (Fig. 2). In addition, the association of mineralogical brittleness with porosity is proven
in section 2, and porosity can be correlated with compressional slowness [Mavko, Mukerji and Dvorkin,
2009]. Therefore, mineralogical brittleness might be correlated with density, compressional, and shear
slowness. However, after examining the cross plots of mineralogical brittleness with density, compres-
sional and secondary slowness, it was found that only compressional sonic slowness was correlated well
with mineralogical brittleness. Therefore, three correlations of mineralogical brittleness with compres-
sional slowness were also developed for Woodford, Barnett, and Eagle Ford shales and are presented in
the following.
8 SPE-170972-MS

Figure 6 —Correlation of brittleness with compressional slowness for Woodford shale.

Figure 7—Correlation of brittleness with compressional slowness for Barnett shale.

Correlation of Brittleness with Compressional Sonic Slowness for Woodford Shale

Well logging data shown in Fig. 6 from two wells of Woodford shale were selected for investigating the
correlation between mineralogical brittleness and compressional slowness (DTC). The correlation is as
follows:
(9)

Brittleness B26 is not good in terms of low R2 value, but shows a linear trend. In order to improve the
quality of B26, one needs data from additional wells in Woodford shale. Unfortunately, not many wells
have both mineralogical and sonic logs. In the section of field application, one similar correlation of
mineralogical brittleness and DTC was found in Woodford shale well.
SPE-170972-MS 9

Figure 8 —Correlation of brittleness with DTC for Eagle Ford shale.

Figure 9 —Correlation of brittleness with porosity for Woodford, Barnett and Eagle Ford shales

Correlation of Brittleness with Compressional Sonic Slowness for Barnett Shale

Well logging data from three wells of Barnett shale were selected for developing the correlation between
mineralogical brittleness and DTC, and were plotted in Fig. 7. The correlation is as follows:
(10)

B27 is an excellent correlation in terms of high value of R2.

Correlation of Brittleness with Porosity for Eagle ford Shale
Well logging data from five wells of Eagle Ford shale were selected for exploring the correlation between
mineralogical brittleness and DTC, and were plotted in Fig. 8.
The correlation is defined as follows:
10 SPE-170972-MS

Figure 10 —Correlation of brittleness with compressional slowness DTC for Woodford, Barnett and Eagle Ford shales.

Figure 11—Correlation of brittleness with compressional slowness DTC for Barnett and Eagle ford shales.

(11)

B28 is a good correlation in terms of the value of R2.

Global Correlations of Brittleness with Porosity and Compressional

Slowness
From the above analysis, it is concluded that mineralogical brittleness can be linearly correlated with
neutron porosity and compressional slowness for Woodford, Barnett, and Eagle Ford Shales. It might be
valuable and possible to develop a global correlation between mineralogical brittleness and compressional
slowness, mineralogical brittleness, and neutron porosity by integrating data from the three plays. If
similar correlations as last two sections are observed, to some extent, it proves that there might be a
physical relationship between mineralogical brittleness, neutron porosity, and compressional slowness.
SPE-170972-MS 11

Figure 12—Prediction of brittleness with local correlation (B24) and

global correlation (B29) for Well A of Barnett shale. The major vertical
Figure 13—Prediction of brittleness with local correlation (B27) and
gridline is 100 ft.
global correlation (B30) for Well B of Barnett shale. The major vertical
gridline is 100 ft.

Global Correlation of Brittleness with Porosity

Mineralogical brittleness and neutron porosity data from 12 wells of Woodford, Barnett, and Eagle Ford
shales are selected for the development of a global correlation between mineralogical brittleness and
porosity. The data are plotted in Fig. 9. The developed correlation is as follows:
(12)

B29 is an excellent correlation in terms of high value of R2.

The trend in Fig. 9 proves that the correlation between brittleness and neutron porosity is a global
relationship for all shale reservoirs. For other shale plays not being studied here, such as Marcellus,
Haynesville, Fayetteville shale, etc., one may develop similar correlations of brittleness with the available
data, or use B29 to conduct a qualitative evaluation if not enough logging data is at hand.

Global Correlation of Brittleness with Sonic Compressional Slowness

Mineralogical brittleness and DTC data from all 10 wells of Woodford, Barnett, and Eagle Ford shales are
selected for the correlation and are plotted in Fig. 10. The correlation is as follows:
(13)
12 SPE-170972-MS

Figure 14 —Comparison of Woodford shale data with the entire well data in the plot of mineralogical brittleness versus neutron porosity. As the wells
discussed earlier, neutron porosity has a strong correlation with mineral-derived brittleness in the Woodford section and throughout the logged
interval in this well.

Figure 15—Comparison of Woodford shale data with the entire well data in the plot of mineralogical brittleness versus compressional slowness.
Compressional slowness also has a strong correlation with mineral-derived brittleness in the Woodford section and throughout the logged interval
in this well.

The correlation given by B30 is good, but not as good as for B29. It can be observed in Figs. 9 and 10
that mineralogical brittleness is better correlated with neutron porosity than compressional slowness. The
correlation B30 is a global correlation that can be applied as a global correlation to evaluate brittleness of
other shale plays without mineralogical logging.
The B30 correlation can be improved by eliminating Woodford shale data. Employing only the Barnett
and Eagle Ford shale data, an improved global correlation of brittleness with DTC was developed as
shown below:
(14)
SPE-170972-MS 13

Figure 16 —Logging Data of one granite wash well. Eleven beds are depicted and the neutron, DTC, and ROP are examined separately in each with
respect to their relationship to mineralogical brittleness.

B31 is a good correlation in terms of the value of R2.

If one has adequate logging data from Woodford shale, a better global correlation between brittleness
and compressional slowness can be developed.
Correlation Verification
The brittleness correlations developed here were independently verified with the data from wells not
included in the development of correlations.
Case I: Predicting brittleness with neutron porosity in Barnett shale play
The data from Well A of Barnett shale is shown in Fig. 12. Local brittleness B24 and global brittleness
B29 were calculated, and compared against mineralogical brittleness B22. The results indicate that both B24
and B29 can predict brittleness in Barnett shale equally well. The tracks of GR (gamma ray), CAL (caliper),
and RT (true resistivity) are also included to verify the interpretation of results.
Case II: Predicting brittleness with compressional slowness in Barnett shale play
The data from Well B of Barnett shale is shown in Fig. 13. Local brittleness B27 and global brittleness
B30 were calculated, and compared against mineralogical brittleness B22. The results indicate that both B27
14 SPE-170972-MS

Table 1—Differences in the Correlations of Mineralogical Brittleness and Neutron Porosity, Compressional Slowness for Woodford Shale Wells
Correlation Name Figure # Slope Intercept

Mineralogical Brittleness Vs. Neutron Porosity Fig. 3 -1.531 0.858

Fig. 14 -1.417 0.887
Mineralogical Brittleness Vs. Compressional Slowness Fig. 6 -0.012 1.492
Fig. 15 -0.0117 1.515

Figure 17—Correlation of brittleness with neutron porosity for granite wash. In this granite wash log example, neutron porosity shows a definite
correlation with mineral-derived brittleness. The correlation is generally stronger when beds are considered separately.

and B30 can predict brittleness in Barnett shale equally well. The tracks of GR, CAL, and RT are also
included to verify the interpretation of results.
Field Applications
The correlation verification ensures the validity of the concept that developing correlations between
mineralogical brittleness and compressional slowness, neutron porosity. Now, we have another Woodford
Shale well, Granite Wash well at hands. We are going to check whether similar correlations can be found
from these data, how big the differences are.
Case I: Application in Woodford shale play
In a Woodford well in central Oklahoma, the usual strong correlation exists between brittleness and
porosity logs, as shown in Fig. 14 for neutron porosity and Fig. 15 for compressional slowness. Blue
triangle points on the plot represent several thousand feet of log, while brown circular points represent
only the Woodford. The similarities and differences between Figs. 14 and 15, Figs. 3 and 4 are shown in
Table 1. It is observed that the correlations of mineralogical brittleness and neutron porosity have similar
slopes and intercepts, the same for the correlations of mineralogical brittleness and compressional
slowness.
Case II: Application in Granite Wash play
Granite Wash play spanning the Texas Panhandle into Oklahoma approximately 160 miles long and 30
miles wide has been producing hydrocarbons since the late 1950’s. The applications of advanced
horizontal drilling and multi-stage hydraulic fracturing technologies have resulted in increasing produc-
tion rate and net return of revenue. This section verifies that similar brittleness correlations can be
SPE-170972-MS 15

Figure 18 —Correlation of brittleness with DTC for granite wash. In this granite wash log example, compressional slowness demonstrates a decent
correlation with mineral-derived brittleness (R2ⴝ0.32). It is observed that the correlation varies considerable in slopes, when beds are considered
separately.

observed in less tight formations, and assist the design of hydraulic fracturing. Similar analyses were
performed in a Texas Panhandle well penetrating the Granite Wash, depicted in the log in Fig. 16. Neutron
porosity and compressional slowness show a good correlation with brittleness as shown respectively in
Figs. 17 and 18. The log spans four potential reservoirs and several other beds over a few hundred feet.
In these figures, plotted points are color coded for the eleven formation layers. The correlations are not
as tight as those of the shale plays, due to the largely variable mineralogy of the granite wash beds.
In conclusion, Figs. 17 and 18 prove that similar correlations can be found in other plays, such as
Granite Wash plays. The derivation of similar correlations should not be limited to tight shales: it can be
other less tight or non-tight plays.
Discussion and Conclusions
After conducting integrated geomechanics and petrophysics analysis on logging data from twelve plus
wells, the following new correlations were developed: shale specific correlations between mineralogical
brittleness and neutron porosity, and mineralogical brittleness and compressional sonic slowness for the
Woodford, Barnett, and Eagle Ford shales; two global correlations between mineralogical brittleness and
neutron porosity, and mineralogical brittleness and compressive slowness for the three plays. The
correlations have been verified against mineralogical brittleness of two independent wells from Barnett
shale. The concept has been applied to study mineralogical brittleness in granite wash, and other
Woodford shale wells. Similar relationships between mineralogical brittleness and compressional slow-
ness or neutron porosity were observed. The new correlations enable the possibility of evaluating
brittleness in plays lacking mineralogical and shear sonic logging, thus reducing the quantity of laboratory
testing, and inspiring the operators to develop in-house correlations with their own database. Furthermore,
the application in granite wash plays proved the possibility to obtain similar correlations in tight oil plays.
Generally, it is considered that high gamma ray value corresponds to high clay content and low
brittleness, but this is not universally true. Attempts were made to develop correlation between B22 and
gamma ray, but failed. It will also be valuable to compare mineralogical brittleness with internal friction
angle measured by Mohr-Coulomb experiments in laboratory, if sufficient data is available. In addition,
we are not stating that brittleness is an indicator for hydraulic fracturing, but one parameter that should
be considered in fracturing design.
16 SPE-170972-MS

Acknowledgement
The authors would like to thank Prof. Ahmad Ghassemi, Prof. Roger Slatt, Yawei Li of the University of
Oklahoma, and Dr. Zhengwen Zeng of BP for contributing to the valuable discussions. The principal
author also thanks the financial support provided by the University of Oklahoma.

References
Altindag, R. (2003), Correlation of specific energy with rock brittleness concepts on rock cutting,
Journal of the South African Institute of Mining and Metallurgy, 103(3), 163–171.
Andreev, G. E. (1995), Brittle failure of rock materials: test results and constitutive models, Taylor
& Francis.
Bishop, A. W. (1967), Progressive Failure-with special reference to the mechanism causing it, paper
presented at Geotechnical Conference, Oslo, Norway.
Britt, L., and J. Schoeffler (2009), The geomechanics of a shale play: what makes a shale prospective,
paper presented at SPE Eastern Regional Meeting, Charleston, West Virginia, USA, September 23-25.
Chang, C., M. D. Zoback, and A. Khaksar (2006), Empirical relations between rock strength and
physical properties in sedimentary rocks, Journal of Petroleum Science and Engineering, 51(3), 223–237.
Copur, H., N. Bilgin, H. Tuncdemir, and C. Balci (2003), A set of indices based on indentation tests
for assessment of rock cutting performance and rock properties, Journal-South African Institute of Mining
and Metallurgy, 103(9), 589 –599.
Council, G. W. P., and A. Consulting (2009), Modern shale gas development in the United States: A
primer, US Department of Energy, Office of Fossil Energy.
Fisher, M., C. Wright, B. Davidson, A. Goodwin, E. Fielder, W. Buckler, and N. Steinsberger (2002),
Integrating fracture mapping technologies to optimize stimulations in the barnett shale, paper presented
at SPE Annual Technical Conference and Exhibition, San Antonio, Texas, September 29 - October 2.
Hajiabdolmajid, V., and P. Kaiser (2003), Brittleness of rock and stability assessment in hard rock
tunneling, Tunnelling and Underground Space Technology, 18(1), 35–48.
Heidari, M., G. Khanlari, M. Torabi-Kaveh, S. Kargarian, and S. Saneie (2013), Effect of porosity on
rock brittleness, Rock Mechanics and Rock Engineering, 1–6.
Hester, T. C., J. W. Schmoker, and H. L. Sahl (1990), Log-derived regional source-rock character-
istics of the Woodford Shale, Anadarko Basin, Oklahoma.
Honda, H., and Y. Sanada (1956), Hardness of coal, Fuel, 35, 451–461.
Hucka, V., and B. Das (1974), Brittleness determination of rocks by different methods, International
Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, 11(10), 389 –392.
Inamdar, A., T. Ogundare, R. Malpani, W. Atwood, K. Brook, A. Erwemi, and D. Purcell (2010),
Evaluation of Stimulation Techniques Using Microseismic Mapping in the Eagle Ford Shale, paper
presented at Tight Gas Completions Conference, San Antonio, Texas, November 2-3.
Jackson, J. A., J. P. Mehl, and K. K. Neuendorf (2005), Glossary of geology, Springer.
Jarvie, D. M., R. J. Hill, T. E. Ruble, and R. M. Pollastro (2007), Unconventional shale-gas systems:
The Mississippian Barnett Shale of north-central Texas as one model for thermogenic shale-gas assess-
ment, AAPG bulletin, 91(4), 475–499.
Jin, X., S. Shah, J.-C. Roegiers, and B. Zhang (2014), Fracability Evaluation in Shale Reservoirs-An
Integrated Petrophysics and Geomechanics Approach, paper presented at SPE Hydraulic Fracturing
Technology Conference, Society of Petroleum Engineers, the Woodlands, TX, February 4-6.
King, G. (2012), Hydraulic fracturing 101: what every representative, environmentalist, regulator,
reporter, investor, university researcher, neighbor and engineer should know about estimating frac risk and
improving frac performance in unconventional gas and oil wells, paper presented at SPE Hydraulic
Fracturing Technology Conference, The Woodlands, TX, USA, February 6-8.
SPE-170972-MS 17

Lawn, B., and D. Marshall (1979), Hardness, toughness, and brittleness: an indentation analysis,
Journal of the American ceramic society, 62(7-8), 347–350.
Mavko, G., T. Mukerji, and J. Dvorkin (2009), The rock physics handbook: Tools for seismic analysis
of porous media, Cambridge University Press.
Protodyakonov, M. (1962), Mechanical properties and drillability of rocks, paper presented at
Proceedings of the 5th Symposium on Rock Mechanics, Minnesota, May.
Quinn, J. B., and G. D. Quinn (1997), Indentation brittleness of ceramics: a fresh approach, Journal
of Materials Science, 32(16), 4331–4346.
Rickman, R., M. Mullen, J. Petre, W. Grieser, and D. Kundert (2008), A practical use of shale
petrophysics for stimulation design optimization: all shale plays are not clones of the barnett shale, paper
presented at SPE Annual Technical Conference and Exhibition, Denver, Colorado, USA, September
21-24.
Sehgal, J., Y. Nakao, H. Takahashi, and S. Ito (1995), Brittleness of glasses by indentation, Journal
of materials science letters, 14(3), 167–169.
Singh, S. (1986), Brittleness and the mechanical winning of coal, Mining Science and Technology,
3(3), 173–180.
Tiab, D., and E. C. Donaldson (2011), Petrophysics: theory and practice of measuring reservoir rock
and fluid transport properties, Gulf professional publishing.
Wang, F., and J. Gale (2009), Screening criteria for shale-gas systems, Gulf Coast Association of
Geological Societies Transactions, 59, 779 –793.
Weingarten, J., and T. Perkins (1995), Prediction of sand production in gas wells: methods and Gulf
of Mexico case studies, Journal of Petroleum Technology, 47(7), 596 –600.
Yagiz, S. (2009), Assessment of brittleness using rock strength and density with punch penetration
test, Tunnelling and Underground Space Technology, 24(1), 66 –74.
Yang, Y., H. Sone, A. Hows, and M. Zoback (2013), Comparison of Brittleness Indices in Organic-
rich Shale Formations, paper presented at 47th US Rock Mechanics/Geomechanics Symposium, Amer-
ican Rock Mechanics Association.
Yarali, O., and S. Kahraman (2011), The drillability assessment of rocks using the different brittleness
values, Tunnelling and Underground Space Technology, 26(2), 406 –414.
18 SPE-170972-MS

Appendix

Table A-1—Selected Expressions of Brittleness

Formula Variable declaration Test Method Reference

B1 ⫽ (Hm-H)/K H and Hm are macro and micro- Hardness test [Honda and Sanada, 1956]
hardness, K is bulk modulus
B2 ⫽ q␴c q is percent of debris (⬍0.6 mm Proto impact test [Protodyakonov, 1962]
diameter); ␴c is compressive
strength
B3⫽ ␧ux⫻100% ␧ux is unrecoverable axial strain Stress strain test [Andreev, 1995]
B4 ⫽ (␧p- ␧r)/ ␧p ␧p is peak of strain, ␧r is residual [Hajiabdolmajid and Kaiser, 2003]
strain
B5 ⫽ ␶p- ␶r/ ␶p ␶p and ␶r are peak and residual of [Bishop, 1967]
shear strengths
B6 ⫽ ␧r/ ␧t ␧r and ␧t are recoverable and total [Hucka and Das, 1974]
strains
B7 ⫽ Wr/Wt Wr and Wt are recoverable and
total strain energies
B8⫽ ␴c/␴t ␴c and ␴t are compressive and Uniaxial compressive strength and
B9⫽ (␴c - ␴t)/(␴t ⫹ ␴c) tensile strength Brazilian test
B10 ⫽ (␴c␴t)/2 [Altindag, 2003]
B11 ⫽ (␴c␴t)0.5/2
B12⫽H/KIC H is hardness, KIC is fracture Hardness and fracture toughness [Lawn and Marshall, 1979]
toughness test
B13 ⫽ c/d c is crack length, d is indent size Indentation test [Sehgal et al., 1995]
for Vickers indents at a
specified load; empirically
related to H/KIC
B14⫽Pinc/Pdec Pinc and Pdec are average [Copur et al., 2003]
increment and decrement of
forces
B15⫽ Fmax/P Fmax is maximum applied force on [Yagiz, 2009]
specimen, P is the
corresponding penetration.
B16 ⫽ H⫻E/KIC2 H is hardness, E is Young’s Hardness, stress strain, and [Quinn and Quinn, 1997]
modulus, KIC is fracture fracture toughness test
toughness
B17 ⫽ 45°⫹ ␸/2 ␸ is internal friction angle Mohr circle or logging data [Hucka and Das, 1974]
B18 ⫽ Sin␸
B19 ⫽ (En⫹vn)/2 En and vn are normalized dynamic Density and sonic logging data [Jin et al., 2014]
Young’s modulus and dynamic
Poisson’s ratio
B20⫽ (Wqtz)/WTot Wqtz, is the weight of quartz, WTot Mineralogical logging or XRD in [Jarvie et al., 2007]
is total mineral weight. the laboratory
B21⫽ (Wqtz⫹Wdol)/WTot Wqtz and Wdol are weights of [Wang and Gale, 2009]
quartz and dolomite, WTot is
total mineral weight.
B22⫽ (WQFM⫹WCarb)/WTot WQFM is weight of quartz, [Jin et al., 2014]
feldspar, and mica; WCarb is
weight of carbonate minerals
consisting of dolomite, calcite,
and other carbonate
components. WTot is total
mineral weight.