You are on page 1of 24

Submitted to AAPG book on gas

shales

Pore-to-regional-scale, integrated characterization workflow for unconventional


gas shales

Roger M. Slatt1, Paul R. Philp1, Neal O'Brien2, Younane Abousleiman1,


Prerna Singh1,3, Eric V. Eslinger4, Roderick Perez1, Romina Portas1,5,
Elizabeth T. Baruch5, Kurt J. Marfurt1 and Steven Madrid-Arroyo1

1
Conoco-Phillips School of Geology and Geophyscs, University of Oklahoma
2
Department of Geology, State University of New York, Potsdam
3
Current Address: Chevron-Texaco Inc.
4
Eric Geoscience Inc. and The College of St. Rose, Albany, New York
5
Current address: Conoco-Phillips, Inc.
ABSTRACT
Based upon recent studies of Barnett and Woodford gas shales in Texas and Oklahoma, a
systematic characterization workflow has been developed which incorporates litho- and
sequence-stratigraphy, geochemistry, petrophysics, geomechanics, well log, and 3D seismic
analysis. The workflow encompasses a variety of analytical techniques at a variety of geologic
scales. It is designed as an aid to identifying the potentially best reservoir, source, and seal facies
for targeted horizontal drilling. Not all of the techniques discussed in this paper have yet been
perfected, and cautionary notes are provided where appropriate.
Rock characterization includes: (1) lithofacies identification from core based upon fabric,
mineralogic (and chemical if possible) analyses; (3) scanning electron microscopy to identify
nano- and micro-fabric, potential gas migration pathways, and porosity types/distribution; (4)
determination of lithofacies stacking patterns; (5) geochemical analysis for source rock potential
and for paleoenvironmental indicators; (6) geomechanical properties for determining fracture
potential of lithofacies.
Well log characterization includes: (1) core-to-log calibration which is particularly critical with
these finely laminated rocks; (2) calibration of lithofacies and lithofacies stacking patterns to well
log motifs (referred to as 'gamma ray patterns' or GRP's in this paper); (3) identification and
regional to local mapping of lithofacies and GRP's from uncored vertical wells; (4) relating
lithofacies to petrophysical, geochemical and geomechanical.properties and mapping these
properties.
3D seismic characterization includes: (1) structural and stratigraphic mapping using seismic
attributes; (2) calibrating seismic characteristics to lithofacies and GRP's for seismic mapping
purposes; and (3) determining and mapping petrophysical properties using seismic inversion
modeling.
Integrating these techniques into a 3D geocellular model allows for documenting and
understanding the fine-scale stratigraphy of shales and provides an aid to improved horizontal
well placement. Although the workflow presented in this paper only relates to two productive gas
shales, we consider it to be more generically applicable.

INTRODUCTION
Recent discoveries of potentially vast global gas resources locked in shales has led to a
need to understand their stratigraphy for (1) regional to local correlations, (2) determining the
most favorable internal gas source and migration pathways, and (3) identifying the best strata for
horizontal well placement and artificial fracture treatment. Recent shale stratigraphic studies
(Bohacs and Schwalbach, 1992; Bohacs, 1998; Macquaker et al., 1998; Schutter, 1998; Almon et
al. 2002; Paxton et al, 2006; Loucks and Ruppel, 2007; Singh. 2008 ; Mazzullo et al, 2009) have
clearly demonstrated that shales are not usually stratigraphically homogenous, and that their
stratigraphic variability can be explained using well established sequence stratigraphic principles.
For the past few years, we have been evaluating some U.S. mid-continent gas shales---
principally the Barnett and Woodford shales---which has led to development of a systematic,
integrated characterization methodology or work flow. The work flow combines a variety of
analytical techniques to characterize these strata at a variety of scales (Figure 1). In this paper,
we present these techniques, provide some examples to demonstrate our findings, and point out
some potential pitfalls in their application. It is not our intent to present a comprehensive analysis
of individual shales we have been studying, but rather to provide examples of the techniques,
applications and results of our workflow approach for more generic application. This paper is
organized approximately according to the headings and subheadings in the work flow (Figure 1).
The first part of the workflow---regional tectono-stratigraphic aspects of these gas shales---has
been published, so is not repeated here (Montgomery et al., 2005; Pollestro, 2007).
Figure 1
.Flow chart for integrated characterization of unconventional gas shales.

ROCKS
Identify lithofacies in core and their properties
Techniques
Any study of gas shales should begin with the rocks, preferably whole core, but if not
available, cuttings, and if possible, outcrops. Core samples from both the Barnett and Woodford
shales were studied according to the methods of O'Brien and Slatt (1990)---that is, at the SEM
(scanning electron microscope), thin section, and hand specimen (core) scales. Cm-scale core
description of sedimentary structures, textures, and stratification styles form the primary basis of
core and thin section characterization. Mineralogic analyses were conducted by standard X-ray
diffraction (XRD) and Fourier Transform Infrared spectroscopy (FTIR) techniques, supplemented
by chemical analyses. These analyses, coupled with Total Organic Carbon content (TOC) by
combustion provide the basis for identifying lithofacies based upon compositional and fabric
features (Figure 1).
Example:
Singh (2008) used the above techniques to identify nine lithofacies (Figure 2 and Figure 3; Table
1) in some Barnett shale core. Similar lithofacies were identified by Hickey and Henk (2007). The
dominant minerals comprising the lithofacies are quartz, calcite, dolomite, clay minerals, feldspar,
muscovite and phosphatic grains (Figure 2). More recent, unpublished XRD analyses indicate the
only clay mineral is mixed layer illite/smectite with 70 to 95% illite layers.
Figure 2
.Nine lithofacies defined by Singh (2008). Each lithofacies lists the name, the average mineral composition
and illustrates the core and thin section characteristics. Numerical coding for figure 4 is: Lithofacies 1 is at
the top left (siliceous, non calcareous mudstone) and lithofacies 9 is at upper right (fossiliferous deposits).
Also shown is the lithofacies distribution in a cored well.

Figure 3
.Left column shows the distribution of lithofacies in a Barnett core (after Singh, 2008). GRP = gamma ray
pattern; red arrows to the left indicate upward increase in carbonate lithofacies; green arrows to the left
indicate upward decrease in carbonate lithofacies. The middle curve is the interpreted relative sea level curve
of Singh (2008) based upon GRP's; red arrows highlight interpreted deposits of a shallowing sea during
deposition; green arrows highlight interpreted deposits of a deepening sea. The right curve is the Residual
Hydrocarbon Potential (RHP) curve which shows trends of anoxic-to-oxic (red arrows) and oxic-to-anoxic
(green arrows) depositional environments. Note the close correspondence among the three sets of arrows,
indicating cyclic variations in water depth and oxicity levels, which we relate to eustatic sea level cyclicity.
Table 1
Nine
lithofacies in
Barnett Shale

At one end of the lithofacies spectrum is siliceous mudstone, which is enriched in clay
minerals and TOC (Figure 2; Table 1). Phosphatic deposits are also enriched in TOC. At the
other end of the spectrum are carbonate-rich deposits, which include micritic/limy mudstone,
fossiliferous deposits, concretions and dolomitic mudstone (Figure 2). Carbonate-rich lithofacies
contain sparse TOC (Table 1). Some of the carbonate is detrital and some is authigenic.

Porosity and permeability


Techniques
Porosity and permeability measurements are routinely made on core plugs of sandstone and
carbonate rocks, so the standard techniques are not discussed here. However, the credibility and
reproducibility of porosity and permeability measurements of shales using standard techniques is
complicated by small grain size, small pore throats, and small pores (Bowker, 2007). Also,
because shales are commonly fissile, it is often difficult to physically acquire an intact core plug
for analysis. Because lithofacies may be thinly interbedded , retrieval of a homogeneous sample
representative of a given lithofacies is often difficult.
Example:
One hundred and eighty-two(182) porosity and permeability core plug measurements of different
Barnett lithofacies failed to reveal any significant causal relationship (Figure 4). Even though
there is a vague positive relation between porosity and permeability, data points for the nine
different lithofacies appear to be randomly dispersed throughout the data cloud. This plot could
represent true values from undisturbed and uniform samples, a result of the difficulties in sample
collection and measurement described in the pitfalls section of this paper, or a combination.

Figure 4
.Porosity-permeability cross plot from 182 core plug samples from a Barnett Shale core. Different colors are
coded to different lithofacies. See Figure 2 for lithofacies 1-9.
SEM for pore types and networks
Techniques
Irrespective of the difficulty in confidently obtaining reliable, reproducible porosity and
permeability measurements from shales, pores and their connectivity can be directly observed
and to some degree, quantified. We have accomplished this on both Barnett and Woodford shale
samples using standard SEM techniques (O'Brien and Slatt, 1990), as well as by higher-
resolution field emission scanning electron microscopy (FE-SEM). In addition, the style and
degree of alignment of individual grains comprising shale samples were documented. Grain
morphologies and elemental compositions (for mineral identification of grains) were also
determined from energy dispersive x-ray (EDX) analysis.
Example:
Our studies concur with Loucks et al. (2009) conclusion that most pores in the Barnett Shale
siliceous mudstone are contained within organic particles rather than within the mudstone matrix.
We have also found that fossil fecal pellets, on the order of 100-300um in diameter comprise up
to 50% of the grains within some lithofacies (Figure 5A and Figure 5B). The pellets are enriched
in micro- to nano-scale pores, which are most visible under the FE-SEM at high magnifications
(pores approximately 100nm). Under the SEM/EDX, two types of fecal pellets are recognized:
calcium-rich pellets and calcium-phosphatic pellets (Table 2; Figure 5A and Figure 5B, Figure 6
and Figure 7), both of which lack the silicon and aluminum present in the shale matrix (Figure 7C
and Figure 7D), and which can occur together in the same pellet-rich laminae (Figure 6 and
Figure 7). Some calcium-phosphatic pellets are composed of crystalline phosphatic grains with
associated elemental fluorine, confirming the presence of the mineral apatite.

Figure 5
.Scanning Electron Micrographs of Barnett Shale. A. SEM of calcium- pellet morphology. Notice the
orientation of particles in matrix surrounding the pellet (arrows). B. SEM of calcium- phosphatic pellet
morphology. Notice the sharp contact of pellet edges with the adjacent matrix (arrow). C. Pyrite framboid (red
circle) cluster and quartz crystal (blue circle). D. Inside of framboid showing smaller pyrite crystals and
internal porosity.
Figure 6
.EDX dot maps of elemental distributions in Barnett Shale pellet zone. A. Backscatter mode – SEM image of
all pellet locations (high concentrations of dots outline the pellets). B. EDX image showing location of
calcium (Ca) in pellets (dots). Notice all pellets contain calcium. Dark area is matrix. C. EDX image showing
location of phosphorus (P) in some pellets. Notice those pellets also contain calcium (compare to 6B). D. EDX
image showing location of silicon (Si) distribution. Silicon is in the matrix and not in the pellets. The gross
mineral composition of this sample, determined by X-ray diffraction (XRD) is: Quartz 26%; Feldspar 6%;
Calcite 12%; Flourapatite 23%; Kerogen 9%; Pyrite/Marcasite 3%; Illite/muscovite 15%; Misc. 2%.

Figure 7
.Close-up view of two pellets in a shale matrix. A. The upper pellet is a calcite-phosphatic pellet and the lower
pellet is a calcite- pellet that has 'exploded' when hit with the electron beam. B. Same view with different
lighting. C. EDX distribution of Silicon (Si) highlighting the shale matrix. D. EDX distribution of Aluminum (Al)
highlighting the shale matrix. E. EDX distribution of Phosphorous (P) highlighting the upper pellet only. F.
EDX distribution of Calcium (Ca) highlighting both upper and lower pellets.
Table 2
Characteristics of pellets

An interesting feature of the calcium-pellets is that samples which have not been gold-
coated during preparation for analysis change their shape or appear to “explode" under a high
voltage (15 KV) electron beam (Figure 7A). Neither coated nor non-coated calcium-phosphatic
pellets change their shape when subjected to 15 KV voltage (Figure 7A), thus suggesting the
“explosion" is due to the generation of CO2 or other gas (methane?) in the calcite-rich pellets.
Although EDX analysis revealed the presence of elemental carbon, whether it is carbon
associated with carbonate or organic matter has not been ascertained.
Commonly, pore spaces that do occur within the shale matrix in both the Barnett and
Woodford shales appear elongate (Figure 8). An outcrop sample of Woodford Shale exhibits oil
droplets which appear to have partially migrated out of the shale matrix and into these elongate
pores during a hydrous pyrolysis experiment (Figure 9). It is possible that these elongate pores
provide early migration pathways for hydrocarbon molecules, as was determined by O'Brien et al.
(1996) from similar experiments performed on the petroliferous Monterey Shale, California. In the
case of the Monterey Shale, as well as with the Barnett and Woodford shale samples discussed
in this paper, we use the term 'microchannel' for these elongate pores to indicate their potential
as hydrocarbon migration pathways. These features superficially resemble parting planes that
have unloaded when the rock was brought to the ground surface. However, we have concluded
that they are not parting planes parallel to bedding, based upon the following observations: (1)
the microchannels do not extend across the entire viewing area of the sample on an SEM stub
nor at the core plug scale; (2) they are not perfectly horizontal and parallel to bedding; (3) they
form a stairstep pattern. They are interpreted to be primary openings in the original flocculated
clay sediment that remained after lithification, thus preventing perfect platy orientation.

Figure 8
.Scanning Electron Micrographs of Woodford shale. A. Microchannel in shale sample. Orange oval highlights
bacterium-like structure on a grain surface. B. Four arrows outline a "microchannel". C and D. Potential
hydrocarbon migration pathways between grains (red arrows).
Figure 9
.Scanning Electron Micrographs of Woodford Shale during hydrous pyrolysis experiment of heating to 350oC
for 4 days. A. Oil droplet in microfracture. B. Oil 'slick' near microfracture showing the contact between the oil
and the matrix. C and D. Oil droplets oozing from rock matrix into open microchannels.

Loucks et al. (2009) have suggested the pores within organic particles are generated during
thermal maturation of the organics. Here, we speculate a similar origin for the micro- to nano-
scale pores in the fecal pellets. It is conceiveable that pores are generated during maturation of
the organic particles within the pellets, and that generated gas first migrates through these pores,
then into and through the larger microchannels in the shale matrix. Thus, pellet-rich lithofacies
might be preferential zones of gas generation and primary migration.

Geochemistry for source rock potential and paleoenvironmental indicators


Techniques
ROCK-EVAL is a classical technique widely used for the characterization of source rock quality.
Among the many parameters that can be obtained from ROCK-EVAL analyses, two important
parameters in our shale studies are the (1) amount of extractable organic material in the source
rock, generally derived from kerogen breakdown (S1 peak on a gas chromatogram), and (2)
residual kerogen (S2 peak). The S1+S2 peaks normalized to TOC content of analyzed samples
provide the parameter referred to as the residual hydrocarbon potential (RHP) (Fang et al.,
1993).
We use RHP as a paleoenvironmental indicator. In order to apply the RHP parameter for this
purpose, it is first assumed that all of the samples are at similar levels of maturity, as was
determined for shales in our study area (samples from a single well are at similar depths of
burial). Therefore any change in S1 and S2 is not due to any maturity changes, but reflects
changes in the amount of preserved organic matter (kerogen). Therefore it can be expected that
under anoxic conditions in the depositional environment, larger amounts of organic matter (TOC)
will be preserved in the sediment and the S2 peak will be greater than under oxic conditions,
where less TOC is preserved and the S2 peak will be smaller. Thus, the calculated RHP value
[(S1 + S2/TOC)] will be larger for sediment deposited under anoxic conditions than for sediment
deposited under oxic conditions.
Analysis of eukaryotic biomarkers in the extractable material from the shales provides an
additional analytical tool for paleoenvironmental interpretation. Biomarkers have been used for
many years not only to interpret depositional environments, but also to determine source rocks,
maturity, extent of biodegradation and a number of other characteristics.
Examples:
Both the RHP and biomarker data were used as indicators of oxic and anoxic environments of
deposition. RHP values for a number of samples in a Barnett core revealed cyclicity between oxic
and anoxic conditions during sediment deposition (Figure 3): i.e. intervals of low RHP (app. 1.3-
1.5) alternating with periods of high RHP (1.6-2.2). Low RHP intervals generally correspond with
Singh's (2008) carbonate-rich lithofacies and higher RHP intervals are associated with the more
siliceous-organic mudstone lithofacies (Figure 3).
Biomarkers helped to determine the environment of deposition in a Woodford shale core.
The upper, quartzose part of this core contains higher concentrations of the eukaryotic
biomarkers, specifically C29 steranes, that are associated with oxic conditions (Figure 10). The
lower, more clay-rich part of the core contains fewer eukaryotic biomarkers,, indicating more
anoxic conditions

Figure 10
.Vertical distribution of geochemical biomarkers in a Woodford Shale core. Higher concentrations of
eukaryotic biomarkers-C29 steranes indicate more oxic conditions. Higher concentrations of Chlorobiaceae
indicate anoxic conditions (H2S rich conditions), probably indicative of a stratified water column The
horizontal dashed line corresponds to the boundary between clay-rich shale below and more quartz-rich
shale above.

Geomechanical properties of shales


Techniques:
Being able to visualize small-scale properties of shales is not sufficient to mechanically
characterize them . Measurements of applied force and displacement are required to quantify
shale properties, even when sample sizes are necessarily small owing to difficulties in sampling
mentioned above. Two critical properties that affect wellbore stability and hydraulic fracturing are
Young's modulus (E) and Poisson's Ratio (v). Experimental rock mechanics techniques that
measure these and related properties were used to test small samples of both Woodford and
Barnett shale (Abousleiman, et al., 2007 and 2009). Woodford shale samples were preserved at
the wellsite in non-reactive decane and mineral oil PG1 in order to prevent possible dessication
and rearrangement of grains before analysis. The technique combines a nano-indentor (Fig. 11)--
-which can measure displacements from 200nm to 10 um, and maximum applied forces <
1000uN by indenting drill cuttings-size (<1cm) shale samples (Ulm & Abousleiman, 2006;
Abousleiman et al., 2009)---, with laboratory-based Ultrasonic Pulse Velocity (UPV) and
mineralogic (XRD) measurements. Shown in Figure 11 is a schematic of a nano-indenter, the
force-displacement curve from a nano-indentation test, and an Atomic Force Microscope (AFM)
image of a nano-indention with dimensions of 4 um × 4 um on the surface of a Woodford sample.
From the peak load and the slope of the rebound curve, information about the material hardness
(or strength) and elastic modulus (or stiffness) can be obtained from the following equations:

Figure 11
.Schematic of the nano-indentor, a indentation mark on a Woodford Shale sample and a force-displacement
curve for the sample (from Abousleiman et al., 2009). See text for explanation.

(1)

(2)
Where H is the hardness of the material, M is the indentation modulus, P is the peak load of
the loading curve, h is the indentation depth, and Ac is the area of the contact surface between
the indentor cone and the indented material (Oliver & Pharr, 1992). The indentation moduli are
directly related to the engineering elastic properties of the material (Oliver & Pharr, 1992;
Delafargue & Ulm, 2004). For example, for an isotropic material, the indentation modulus can be
expressed in terms of the Young's modulus, E, and Poisson's ratio:

Examples:
Results of a nano-indentor test on a Woodford Shale sample are shown in Figure 11. Indentation
depth increased to 1000nm with an increasing indentation load to a peak load of almost 5mN,
and then the load was removed. The depth of indentation is a function of the micro- and nano-
scale strength properties of the rock, some of which are listed below, with examples.

1. Mineral composition: Elastic and pororelastic moduli and coefficients have been
estimated from the measurement of porosity, density and mineral composition. The
fracture frequency in a Woodford Shale core appears to be at least generally related to
overall mineral composition (Abousleiman et al, 2007 and 2009; Buckner, 2010), with
fractures being more abundant in a quartzose upper Woodford interval than in a lower,
more clay-rich cored interval (Figure 12). These observations were verified by fracture
distributions measured in a borehole image log in the corehole (Portas, 2009; Buckner,
pers. comm.. 2010)
2. Lithofacies: Filled fractures sometimes terminate at lithofacies boundaries. An example
in Figure 13 shows a fracture within a siliceous shale terminating within a phosphatic bed,
then re-emerging back into the siliceous shale which underlies the phosphatic bed. This
observation suggests that the more porous phosphatic bed may be able to absorb more
stress than the siliceous bed, and thus be less capable of fracturing.
3. Mineral crystal structure: FE-SEM microscopy revealed aligned, tensional
microfractures which resemble mineral fabric (i.e. crystal structure) (Figure 14),
suggesting that crystallographic planes of weakness, and thus mineralogy, might help
initiate micro-fractures.
4. Microchannels: As noted above, microchannels are common in the shales (Figure 8)
They are important not only as potential hydrocarbon migration pathways, but also
because they likely will affect geomechanical properties of the shales.
5. Rock Fabric (lineations/laminations/bedding): Sierra et al, (2010) have demonstrated
in the Woodford Shale core discussed above that small shale samples are weaker when
stress is applied parallel to laminations, then when stress is applied perpendicular to
laminations. This affect, which we deem to be very important to geomechanical
properties of shales, is most likely a result of the preferred orientation of fabric in most
shales.

Figure 12
.The left figure is an ECSTM mineralogy log of the behind-quarry well showing the higher quartz content
(yellow) in the upper Woodford and higher clay content (gray) of the middle Woodford Shale. Center figure is
an FMITM log of part of the well showing a fracture through thin-bedded strata. Right graph shows the
density of fractures per 2.5ft.,as measured from the FMITM log. Lower left figure shows a filled fracture (white
vertical plane) in the core.

Figure 13
.Shale core showing a near-vertical filled fracture offset and
separated by a phosphatic bed. The phosphatic bed appears more
porous than the adjacent shale beds, and thus may be more capable
of absorbing fracture stresses.
Figure 14
.Large figure is FE-SEM
Electron Micrograph at
100,000X magnification
showing linear tensional
partings, possibly aligned with
crystal structures. The three
smaller figures on the left are--
-from top to bottom---of a
propagated fracture tip at
200,000X, 300,000X and
600,000X magnification.

WELL LOGS
Core-to-log depth correction
Techniques:
In order to correlate well log properties to core characteristics, it is essential to first very carefully
and accurately determine a core-to-log depth correction. Conventional well logs such as the
gamma-ray log do not have the resolving power for detecting thin-bedded lithofacies in uncored
wells (Figure 15). As well, the human eye often can miss the fine-scale stratigraphy detected by a
borehole image log (Figure 15) (Davis et al., 2006). Therefore, without a precise core-to-log depth
correction, even if a gamma scan has been run on core, details of stratification may be
overlooked, which adds considerable uncertainty when attempting to relate geological
observation and laboratory-derived petrophysical properties with well log-derived properties.

Figure 15
.Depth calibration of calcite
concretions in core description
(yellow) and static and dynamic
FMITM log (white). Notice the
stratigraphic detail provided by the
FMITM log and the relatively flat
gamma ray log. In the core
description, concretions are yellow,
siliceous mudstone is gray, and
siliceous-calcareous mudstone is
brown.
Example:
We have found the best way to obtain a reliable core-to-log correction is to calibrate depths using
easily-identifiable lithofacies, such as calcareous concretions, which are visible on a borehole
image log (Figure 15) and which exhibit relatively high density and velocity (reciprocal of sonic
transit time) on logs. In the absence of an image log, radioactive lithofacies, such as phosphatic
rocks, can be used if they are thick enough to emit a detectable gamma ray log response. We
have also found that the core-to-log depth correction is not a constant length throughout a well,
thus corrections must be made at shorter stratigraphic intervals.

Determining lithofacies and their properties in logged, but uncored wells


Techniques:
Typical problems associated with calibration of well logs with core characteristics and then
prediction of lithofacies in uncored well intervals include: (1) the core-to-log depth correction
mentioned above, (2) well log insensitivity to thin beds, (3) well log insensitivity to some textural
and bedding features visible in core and sometimes on borehole image logs (Davis et al., 2006),
and (4) insufficient core to represent all facies types. However, there is sufficient contrast in
mineral composition and TOC content of the lithofacies we studied to affect gamma ray log
response, so this log was a primary tool for identification of lithofacies and their stacking patterns
in uncored wells.
A second, more automated approach that we tested to identify lithofacies in uncored wells
was a probabilistic clustering procedure (PCP) within a proprietary computer program (GAMLS
[Geologic Analysis via Maximum Likelihood System]). This approach is described by Eslinger and
Everett (2004, 2006 and this volume), so is not repeated here.
For relating geomechanical properties measured on shale samples to well log response,
neutron and density porosity (NPHI and DPHI) logs, and more advanced Element Capture
Sprectroscopy (ECS), Combination Magnetic Resonance (CMR), Formation Micro Imager (FMI)
and Sonic Scanner (MSIP) logs (Herron, 1986; Pemper et al., 2006)---all trademarks of
Schlumberger---were calibrated to core, then input to a theoretical GeoGenome model
(Abousleiman et al. 2007 and 2009) which estimates shale anisotropic elastic and poroelastic
properties from the combined rock-log data set (Ortega et al., 2007; Ortega et al., 2009).
Examples:
A comparison between core-described lithofacies and lithofacies predicted using the PCP method
shows good predictive capabilities for an interval of thin, compositionally diverse Barnett Shale
beds (Figure 16). The method was tested for automated correlations with a set of eleven Barnett
Shale wells forming a 42km (~24-mile) long cross section. Four well logs (RHOB, NPHI, GR, and
PEF) were used as clustering variables, resulting in the identification and well-to-well correlation
of major stratigraphic intervals (Figure 17).
Figure 16
.Gamma ray (GR) and resistivity logs of a portion of a
Barnett Shale core alongside a measured core
description and a predicted lithofacies distribution
based upon the probabilistic clustering procedure
described in the text. Grays are siliceous mudstones,
black/tourquise are calcareous mudstones, and
purple is phosphatic deposit.

Figure 17
.42km long well log cross section
using the PCP method (see text for
description) to identify rock groups
for correlation purposes. Gray =
siliceous, non-calcareous
mudstone; Red = phosphatic
deposit; green = calcareous
mudstone; blue = muddy
limestone.

Geomechanical properties calculated from the acoustic log revealed significant variations in
Young's Modulus, Poisson's Ratio and porosity between the less clay rich (20% from ECS log)
upper Woodford and the more clay-rich (32%) middle and lower part of the Woodford core (Figure
18). In general, Young's Modulus and Poisson's Ratio are lower for the more clay poor interval,
probably due to the higher porosity of that interval. However, at a finer scale (for example,
compare at 130ft. and 170ft. on Figure 18) the 130ft. clay rich zone exhibits a relatively low
Young's Modulus and high Poisson's Ratio and the 170ft. zone of similar composition exhibits the
opposite trend. Thus, the other lithologic factors, mentioned earlier, besides porosity, play a role
in geomechanical properties of similar lithologies.
Figure 18
.Depth plot of gamma ray log, ECSTM mineralogy log (yellow is quartz, gray is clay minerals), porosity (red =
density porosity; blue = neutron porosity), and log-derived Young's Modulus and Poisson's Ratio. Contact
between the upper and middle/lower Woodford is at 37m (122ft.) (red line). See text for explanation.

Define lithostratigraphy and stacking patterns in uncored wells for (sub)regional


mapping
Techniques:
Identifying the lithostratigraphy or vertical stacking of a set of shale lithofacies is not as
straightforward as it is for sandstones or carbonates owing to shales' finer grain size and degree
of stratification. However, once a good match between log and core depths is obtained for a well,
it is possible to relate subtle stratigraphic variations in log character to different lithofacies
stacking patterns. Once this match is accomplished, lithostratigraphy in uncored wells can be
(sub)regionally correlated and mapped.
Example:
Singh (2008) identified three distinctive lithofacies stacking patterns in Barnett cores: 1) upward
increase in carbonate lithofacies (upward decrease in siliceous lithofacies); 2) upward increase in
siliceous lithofacies, (upward decrease in carbonate lithofacies) and 3) no significant vertical
change in lithofacies (Figure 19). More TOC is associated with the siliceous lithofacies than with
the carbonate rich lithofacies (Table 1), so the corresponding three gamma-ray log patterns, here
termed GRP) are: 1) upward decrease, 2) upward increase, and 3) no vertical change in gamma
ray log response, respectively (Figure 19). Using these three gamma-ray log motifs, several
GRP's were identified and correlated on logs from 602 uncored, vertical wells for subregional
mapping in the area (Singh, 2008). For example, Figure 3 shows 13 GRP's in one Barnett core
and Figure 20 shows regional maps of two of these (GRP-4 and -5).
Figure 19
.Characteristics of the three gamma ray stacking patterns (GRP's): Upward-decreasing gamma ray, upward-
increasing gamma ray, and upward-constant gamma ray pattern. Each figure illustrates the gamma ray curve,
the vertical stacking pattern of lithofacies, and thin section photomicrographs of the lower and upper
lithofacies within each GRP.

Figure 20
.Two gross isopach thickness maps of GRP's. Color bars are thicknesses in feet. For GRP-4, red = 90ft. (30m)
and tourquise = 10 ft. (3m). . For GRP-5, red = 30ft. (10m) and dark blue = 4 ft. (1.3m) Black dots are cores
examined by Singh (2008).
Interpreting depositional history from GRP stacking patterns and related data
Technique:
At this stage of characterization, integration of the above techniques and results provide the
means to interpret the origin and depositional history of the shale strata being studied.
Example:
Singh (2008) interpreted the (1) upward-increasing carbonate/-decreasing TOC GRP as
indicating a progressive shallowing of water over the time interval of GRP deposition, (2) upward-
increasing siliceous/-increasing TOC GRP as indicating an upward deepening of water and (3)
uniform GRP as indicating aggradation during a time interval of unchanging depositional
environment. These three stacking patterns bear resemblance to the retrogradational,
progradational, and aggradational parasequence sets of VanWagoner et al. (1990; their Figure
3). This similarity, plus the sub-regional extent of the mapped GRP's (Figure 20) and their
correspondence with RHP geochemistry (Figure 3) indicate the depositional history of the shale
strata can be explained within the sequence stratigraphic context of cyclical fluctuations in
eustatic sea level.
The GRP with an upward-increase in carbonate lithofacies corresponds with oxic (low RHP)
conditions, as might be expected in a relatively shallow marine setting. The GRP with the upward-
increase in siliceous lithofacies corresponds with anoxic (high RHP) conditions, as would be more
likely in a deeper water setting. Although oxic/anoxic near-bottom water conditions are a first-
order effect of oceanic circulation and associated oxygenation levels, the cyclicity of the shale
strata (Figure 3) point to eustatic causes, as suggested from Loucks and Ruppel's (2007)
Mississippian eustatic sea level curve for the Barnett and related strata.
Unfortunately the lack of high-frequency age dates within the Barnett and Woodford shales
discussed in this paper precludes determining: (1) the hierarchy of eustatic cyclicity (i.e. 2nd oder,
3rd order, etc.), (2) whether these stacking patterns represent depositional sequences, systems
tracts, or parasequences (as defined by VanWagoner et al. (1990) and (3) the influence of
periodic tectonic activity on sub-regional depositional patterns (Borges, 2007, Singh, 2008, Abou
Elresh and Slatt, in press).

SEISMIC REFLECTION ANALYSIS


Techniques
The drilling of horizontal wells in the Barnett and Woodford shales has become the norm
(Montgomery, 2005). Although horizontal wells are more appropriate for production, vertical wells
provide hard information on the stratigraphy. Horizontal wells limit the ability to apply techniques
described above for regional and local mapping of strata that are stratigraphically beneath the
horizontal wells. It is possible that some productive zones that are stratigraphically beneath a
horizontal well could be missed. To extend stratigraphic analysis to deeper horizons, it is
imperative to establish relations between existing vertical wells and seismic reflection data,
preferably 3D seismic volumes. Ideally, a structural analysis should be conducted prior to
attempting seismic stratigraphic correlations of shales. Examples are provided below.
Example: Local structure
Figure 21A displays the major stratigraphic and structural features from a seismic section in part
of the Newark East Field. Some major faults appear to extend upward from the basement into the
Barnett Shale. The main fault, named the Mineral Wells fault, trends northeast-southwest (Figure
21B), which corresponds to the trend of basement features imaged from a horizontal tilt derivative
map, generated from high resolution aeromagnetic data (Figure 21C) (Elebiju et al., 2008).
Figure 21
.Top figure (A) is a pre-stack, time migrated seismic section showing interpreted major stratigraphic and
structural features in part of the Newark East Field. Lower left figure (B) is a horizon slice showing a major
northeast-southwest trending fault through the Barnett Shale (after Borges, 2007). Lower right figure (C) is a
horizontal derivative of the tilt derivative map (HD_TDR) beneath the Barnett Shale (Elebiju, 2008).

Smaller faults and fault-related karst on the unconformity surface which separates the
Barnett Shale from the underlying, water-bearing Ellenburger Group (and Viola) carbonates can
affect well performance. Water-encroachment into a wellbore can be a limiting factor in gas
production because the water moves upward into the reservoir from these faults and associated
fractures. Seismic attribute analysis of 3D seismic volumes in and south of the Newark East field
provides a means of imaging the details of unconformity fault- ,fracture- and related karst-
patterns and their effect on overlying shales (Baruch et al., this volume). Anomalously thick lower
Barnett intervals (potential sweet spots?) deposited over karst sinkholes have also been identified
and mapped using seismic attribute analysis (Baruch et al.,this volume).
Example: Stratigraphy and GRP stacking patterns
Variations in the compositional, petrophysical, and geomechanical properties of lithofacies implies
that some lithofacies would be more favorable horizontal drilling targets than others. Thus, it
would be desireable to seismically map individual lithofacies. However, this is not usually feasible
because individual lithofacies are often beneath seismic resolution. However, because lithofacies
are predictably stratified into the three possible GRP's, mapping the GRP's from seismic does
provide a technique for indirectly mapping individual lithofacies. In the study areas, individual
GRP's are 10-25m thick (Figure 3), which is at or near the resolution of our seismic volumes
(Perez, 2009; Baruch et al., this volume). For example, within 3D seismic volumes, internal
seismic reflections are present and mappable between the top of the upper Barnett and base of
the lower Barnett (Figure 22) (Borges et al., 2007; Baruch et al.; this volume). Perez (2009)
developed acoustic impedance profiles from well and seismic control and correlated GRP
horizons, which demonstrated that at least some GRP's could be resolved and mapped
serismically (Figure 23).
Figure 22
.3D seismically-mapped horizons in the southwest part of the Fort Worth Basin (Baruch et al. this volume).

Figure 23
.Seismic reflection and corresponding impedance sections from three Barnett wells showing correlations
with GRP's.

Integrating data sets for stratigraphic properties modeling and mapping


Techniques:
The principle objective of our stratigraphic characterization workflow is to construct a 3D model of
GRP stratification of an area of interest utilizing the techniques described in this paper. Once
constructed, various compositional, petrophysical and/or geomechanical properties of lithofacies
and/or GRP's can be input into the model for the purpose of selecting most suitable stratigraphic
intervals (lithofacies or GRP's) for horizontal drilling.
Example:
To demonstrate our objective, a 3D geological/petrophysical model for part of the Newark East
Field is shown in Figure 24. The model comprises 14 GRP's that were identified and correlated
from lithofacies-calibrated gamma ray, density, and sonic logs from 45 wells in a 3D seismic
volume area (Perez, 2009). Horizons used in the model are those that were mapped seismically.
The logs were upscaled (blocked) (Figure 25) in order to build the 3D model. Each resulting 3D
grid block is 76 x 76m (250 x 250ft.) in the horizontal direction, and the vertical dimension varies
according to thickness of each GRP. The resulting model consists of 463,643 cells enclosing an
area of approximately 59km2 (19 mi2) (Figure 24). The upscaled GRP's were then interpolated
using Sequential Gaussian Simulation (SGS) based upon variograms in order to estimate the
petrophysical properties at inter-well locations (Perez, 2009).
Figure 24
.A. is a 3D GRP model for an area of the Barnett Shale in Newark East Field. B. is a bulk density (RHOB) cross
section of the model. C. is a gamma ray (GR) cross section of the model. D. is a sonic transit time (delta-T)
cross section of the model.

Figure 25
.Interpreted gamma ray log and blocked/upscaled version based upon
GRP distribution and used for seismic inversion.
Figure 24A shows the resulting 3D distribution of the stacked GPR's, and figures 24 B-D
show selected cross-sections of some of the interpolated petrophysical properties. It is
particularly important to note that not only do properties vary stratigraphically among GRP's, but
they also vary within individual GRP's along their modeled length. This model demonstrates the
ability to integrate rock, well log, and seismic data into a coherent, geologically-realistic 3D model
that can be used for improved understanding of gas shale stratigraphy, and more importantly, as
an aid for stratigraphically-based placement of horizontal wells.

Potential Pitfalls in shale analysis


Based upon our studies, the following cautionary notes are provided with regards to
sampling and lithofacies analysis of shale cores.

1. Common procedure calls for sampling core at uniform depth intervals. Shale cores can
be stratified at a cm or even mm scale, and sometimes so subtly that fine stratification is
missed with the naked eye, and only detectable with high resolution logging tools such as
the borehole imager (Figure 15). Thus, it is possible to overlook some lithofacies if
sampling is conducted only at pre-set, equal stratigraphic intervals.
2. Both XRD and FTIR techniques determine mineral composition, but in fundamentally
different ways. It has become apparent from comparative studies that results using the
two analytical methods on the same sample mixes may be inconsistent in terms of the
reported absolute weight percentage values for various major minerals (calcite, clays,
quartz, etc.). These differences can be attributed to a variety of factors, including sample
preparation, machine conditions, and the manner in which quantitative analyses are
'calibrated' using standards. Accordingly, until the 'best' procedures for obtaining accurate
and consistent mineral analyses are determined, we treat reported values as semi-
quantitative as opposed to quantitative, and we provide mineralogical-based descriptive
conclusions more on 'trends' than on absolute percentages.
3. Macro- and micro-fabrics of shales influence rock strength, petrophysical and acoustic
properties, and gas migration. As such, fabric analysis at the SEM, thin section, and core-
size scales should be included in any shale characterization study (O'Brien and Slatt,
1991). It is insufficient to classify shales only on readily-obtained numerical parameters
such as mineral composition and porosity.
4. Sampling and analytical issues related to accurate porosity and permeability
determination have been mentioned earlier. At best, most such analyses currently
provide 'semi-quantitative' results with possibly significant error bars.
5. Because of the laminated character of gas shales, an accurate core-to-log depth
correction is critical before attempting any core-log petrophysical comparisons or
analyses.

CONCLUSIONS
A workflow has been developed which incorporates a variety of analytical techniques for
characterizing rock, well log, and seismic properties of gas shales at a variety of scales. The
objective is to integrate analyzed properties into a geologically-realistic, 3D stratigraphic model to
better understand the fine-scale stratigraphy of shales and as an aid to improved horizontal well
placement. Although the examples presented in this paper are from the productive Barnett and
Woodford shales, the workflow is intended to be for generic use. Some analytical techniques
currently are not perfected, so results should be used with caution.

ACKNOWLEDGEMENTS
We would like to acknowledge Devon Energy Co. for financial and data support for much of
this research, Ted Champagne of Clarkson University, Potsdam New York and Carol McRobbie
of SUNY-Potsdam for technical assistance with SEM and FE-SEM analyses, and Dennis Eberl,
U.S. Geological Survey, for XRD and chemical analysis of shales. This research was conducted
largely by graduate students in University of Oklahoma's College of Earth and Energy's Institute
of Reservoir Characterization and Poromechanics Institute.
REFERENCES CITES

• Abousleiman, Y., M. Tran, S. Hoang, C. Bobko, A. Ortega, and F.J. Ulm, 2007, Geomechanics Field and Lab
Characterization of Woodford Shale: The Next Gas Play: SPE paper 110120, SPE Annual Technical
Conference Exhibition, Anaheim, Calif, Nov. 11th-14th.
• Abousleiman, Y. M. Tran, S. Hoang, J. A. Ortega, and F. J. Ulm, 2009, GeoMechanics Field Characterization of
the Two Prolific U.S. Mid-West Gas Plays with Advanced Wire-Line Logging Tools: SPE paper 124428, SPE
Annual Technical Conference Exhibition, New Orleans, La, Oct. 4th-7th.
• Abou Elresh and Slatt, in press, High-Frequency Sequence Stratigraphic Analysis of the Barnett
• Shale, Johnson County, Fort Worth Basin, Texas, USA., AAPG Bull.
• Almon, W.R., W.C. Dawson, S.J. Sutton, F. G. Ethridge, and B. Castelblanco, 2002, Sequence stratigraphy,
facies variation, and petrophysical properties in deepwater shales, Upper Cretaceous Lewis Shale, south-
central Wyoming: Gulf Coast Assoc. Geol. Soc. Trans. 52, p. 1041-1053.
• Baruch, E. T., R.M. Slatt and K. J. Marfurt, (this volume). Seismic analysis of the Barnett Shale and Ellenburger
unconformity southwest of Newark East Field, Fort Worth Basin, Texas
• Bohacs, K., and J. Schwalbach, 1992. Sequence stratigraphy of fine-grained rocks with special reference to the
Monterey Formation, (in) J. Schwalbach, and K. Bohacs, eds. Sequence stratigraphy in fine-grained rocks:
Examples from the Monterey Formation: Pacific Section Society Economic Paleontologists and Mineralogists, v.
70, p 7-19.
• Bohacs, K.M., 1998, Introduction: Mudrock sedimentology and stratigraphy-challenges at the basin to local
scales, in J. Scheiber, W. Zimmerle, and P. Sechi (eds.): Shales and Mudrocks (v. 1): Basin Studies,
Sedimentology and Paleontology; Schweizerbartsche Verlagsbuchhandlung, Stuttgart, p. 13-20.
• Borges, G., 2007, Seismic sequence stratigraphy interpretation of the Barnett Shale at Newark East Field, Fort
Worth Basin, Texas, U.S.: Unpublished M.S. thesis, The University of Oklahoma, p.72.
• Bowker, K. A., 2007, Barnett Shale gas production, Fort Worth Basin: Issues and discussion, Amer. Assoc.
Petrol. Geol. Bull., v. 91, p.523-533.
• Chopra, S., and K. J. Marfurt, 2007, Seismic Attribute for Prospect Identification and Reservoir Characterization:
Society of Exploration Geophysicists Geophysical Development Series No. 11, 464B.
• Davis, R.J., C.R. Blume, and R.M. Slatt, 2006, Reservoir characterization: applications of electrical borehole
images (in) Slatt, R.M., N.C. Rosen, M. Bowman, J. Castagna, T. Good, R. Loucks, R. Latimer, M. Scheihing
and R. Smith (eds.): Reservoir Characterization: Integrating technologies and business practices, 26th Annual
GCSSEPM Foundation Bob F. Perkins Research Conference, Dec. 3-6, Houston, Tx. p. 465-512.
• Elebiju, O. O., G. R. Keller, and K. J. Marfurt, 2008, New structural mapping of basement features in the Fort
Worth basin, Texas, using high-resolution aeromagnetic derivatives and Euler depth estimates: Earth Scientist,
The University of Oklahoma, ConocoPhillips School of Geology and Geophysics, p. 46 – 48.
• Eslinger, E.V. and R.V. Everett, 2004, Net pay from petrophysical analysis of 29 Wilcox wells, SW Bonus Field,
Wharton County, Texas, using a multi-well and multi-dimensional clustering analysis coupled with a mineralogy-
based forward modeling procedure; AAPG Annual Convention, April, Dallas, TX (poster session).
• Eslinger, E.V. and R.V. Everett, 2006, A petrophysical study of reservoir quality and flow unit continuity in a
Lower Clearfork oil field, Lower Permian Dolostones, west Texas, using ten wells of variable age and data
quality: AAPG Annual Convention, June, Calgary, Alberta (poster session).
• Eslinger, E.V., B. Burdick, and J. Cooper, in prep., Reservoir characterization using probabilistic clustering as a
basis for classification.
• Ettensohn, F.R., 1998, Compressional tectonic control on epicontinental black shale deposition: Devonian-
Mississippian examples from North America, (in) J. Scheiber, W. Zimmerle, and P. Sechi (eds.), Shales and
Mudrocks (v. 1): Basin Studies, Sedimentology and Paleontology; Schweizerbartsche Verlagsbuchhandlung,
Stuttgart, p. 109-128.
• Fang, H, C. Jianyu, S. Yongchuan, and L. Yaozong, 1993, Application of organic facies studies to sedimentary
basin analysis: a case study from the Yitong Graben, China, Organic Geochemistry, v.20, p.27-42.
• Gonzalez, R. J., S.R. Reeves, E. Eslinger, and G.S. Garcia, 2008, Development and Application of an
Integrated Clustering/Geostatistical Approach for 3D Reservoir Characterization, SACROC Unit, Permian Basin:
SPE Paper 111453, presented, SPE/EAGE Reservoir Characterization & Simulation Conference, Abu Dhabi,
UAE, October 29, 2007.
• Herron, M., 1986, Mineralogy from geochemical well logging: Clays and Clay Minerals, vol. 34, pp. 204-213.
• Hickey, J. J. and B. Henk, 2007, LIthofacies summary of the Mississippian Barnett Shale, Mitchell 2 T.P. ims
Well, Wise County, Texas, Amer. Assoc. Petrol. Geol. Bull, v. 91, p. 437-443.
• Loucks, R. G. and S.C. Ruppel, 2007. Mississippian Barnett Shale: lithofacies and depositional setting of a
deep-water shale-gas succession in the Fort Worth Basin, Texas: Amer. Assoc. Petrol. Geol. Bull., v. 92, p.
579-601.
• Loucks, R.G., R. m. Reed, S.C. Ruppel, and D.M. Jarvie, 2009, Morphology, genesis, and distribution of
nanometer-scale pores in siliceous mudstones of the Mississippian Barnett Shale, Jour. Sedimentary esearch,
v. 79, p. 848-861.
• Macquaker, J.H.S., R.L. Gawthorpe, K.G. Taylor, and M.G. Oates, 1998, Heterogeneity, stacking patterns and
sequence stratigraphic interpretation in distal mudstone successions: Examples from the Kimmeridge Clay
formation, U.K., (in) J. Scheiber, W. Zimmerle, and P. Sechi (eds.), Shales and Mudrocks (v. 1): Basin Studies,
Sedimentology and Paleontology; Schweizerbartsche Verlagsbuchhandlung, Stuttgart, p. 163-185.
• Mazzullo, S.J. .W. Wilhite and I. W. Woolsey, 2009, Petroleum reservoirs within a spiculite-dominated
depositional sequence: Cowley Formation (Mississippian: Lower Carboniferous), south-central Kansas, Amer.
Assoc. Petrol. Bull. V. 93, p. 1649-1689.
• Montgomery, S. L., 2005, Mississippian Barnett Shale, Fort Worth basin, north-central Texas: Gas-shale play
with multi-trillion cubic foot potential: AAPG Bulletin, v. 89, no. 2, p. 155 – 175. 2008
• O'Brien, N., and R.M. Slatt, 1990, Argillaceous Rock Atlas, Springer-Verlag, N.Y., 141 p.
• O'Brien, N.R., Thyne, G., and Slatt, R.M. ,1996, Morphology of hydrocarbon droplets during migration: Visual
example from the Monterey (Miocene), California, Amer. Assoc. Petrol. Geol. Bull., v. 80, p. 1710-1718
• Ortega, J., F. J. Ulm, and Y. Abousleiman, 2007, The effect of the nanogranular nature of shale on their
poroelastic behavior: Acta Geotechnica, vol. 2, pp. 155-182.
• Ortega, J., F. J. Ulm, and Y. Abousleiman, 2009, The nanogrannular acoustic signature of shale, Geophysics,
vol. 74, pp. 65-84.
• Paxton, S.T., A,M. Cruse and A.M. Krystyniak, 2006, Detailed fingerprints of global sea-level change revealed
in Upper Devonian/Mississippian Woodford Shale of south-central Oklahoma, Amer. Assoc. Petrol eol. Search
and Discovery Article # 40211
• Pemper, R. A. Sommer, P. Guo, D. Jacobi, P. Longo, S. Bliven, E Rodriguez, F. Mendez and X. Han, 2006, A
new pulsed neutron sonde for derivation of formation lithology and mineralogy: SPE paper 102770, SPE 2006
Annual Technical Conference and Exhibition, San Antonio, Texas, Sept. 24th-27th.
• Perez, R., 2009; Quantitative petrophysical characterization of the Barnett Shale in Newark East Field, Fort
Worth basin, Texas, Unpublished M.S. thesis, The University of Oklahoma, Norman, Oklahoma, 125p.
• Pollastro, R. M., 2007, Geologic framework of the Mississippian Barnett Shale, Barnett – Paleozoic total
petroleum system, Bend arch-Fort Worth Basin, Texas: AAPG Bulletin, v. 91, p. 405 – 436.
• Portas,, A. R. M, 2009, Characterization and origin of fracture patterns in the Woodford Shale in eastern
Oklahoma for applications to exploration and development, Unpublished M.S. thesis, The University of
Oklahoma, Norman, Oklahoma, 110p.
• Ruppel, S.C. and R.G. Loucks, 2008, Black mudrocks: lessons and questions from the Mississippian Barnett
Shale in the southern Midcontinent, The Sedimentary Record, v. 6, June, p. 4-8.
• Schutter, S.R., 1998; Characterization of shale deposition in relation to stratigraphic sequence systems tracts:
(in) J. Scheiber, W. Zimmerle, and P. Sechi (eds.), Shales and Mudrocks (v. 1): Basin Studies, Sedimentology
and Paleontology; Schweizerbartsche Verlagsbuchhandlung, Stuttgart, p. 79-108.
• Sierra, R., M.H. Tran, Y. N. Abousleiman, and R. M. Slatt, 2010, Woodford Shale mechanical properties and
impacts of lithofacies, 44th U.S. Rock Mechanics Symposium and 5th U.S.-Canada Rock Mechanics
Symposium, Salt Lake City, ARMA10-461.
• Singh, P., 2008, Lithofacies and Sequence Stratigraphic Framework of the Barnett Shale, Northeast Texas.
Unpublished PhD thesis, The University of Oklahoma, Norman, Oklahoma, 181 p.
• Slatt, R. M., E.V. Eslinger, and S.K. Van Dyke, 2009, Acoustic and petrophysical properties of a clastic
deepwater depositional system from lithofacies to architectural elements' scales, Geophysics, vol. 74, March-
April, WA35-WA50.
• Ulm, F. J., and Y. Abousleiman, 2006, The nanogranular nature of Shales: Acta Geotechnica, vol. 1, pp. 77-88,
2006.
• Van Wagoner, J. C., R. M. Mitchum, K. M. Campion, and V. D. Rahmanian, 1990, Siliciclastic Sequence
Stratigraphy in well logs, cores, and outcrops: Concepts for High-Resolution correlation of time and facies:
AAPG Methods in Exploration Series, no. 7, p. 55.

You might also like