You are on page 1of 25

Rock classification in the Eagle AUTHORS

Ford Formation through Shahin Amin ~ Enhanced Oil Recovery


Business Unit, Occidental Petroleum

integration of petrophysical, Corporation, Houston, Texas;


shahin_amin@oxy.com

geological, geochemical, and Shahin Amin is a staff petrophysicist at


Occidental Petroleum Corporation in Houston,
Texas. He received his B.Sc. and M.Sc. degrees
geomechanical characterization in petroleum engineering from Texas A&M
University. While at Texas A&M University,
Shahin Amin, Matthew Wehner, Zoya Heidari, he was a research associate at Multi-Scale
and Michael M. Tice Formation Evaluation of Unconventional and
Carbonate Reservoirs. His research interests
include petrophysics, formation evaluation,
and rock classification in unconventional
ABSTRACT reservoirs.

Integration of geomechanical, geological, geochemical, and pet- Matthew Wehner ~ Veratek Design,
rophysical characterization is critical to enhance production from Escondido, California; nwgeologist@
organic-rich mudrocks. This paper introduces an integrated rock gmail.com
classification method in the Eagle Ford Formation shales and Matthew Wehner is currently a research and
marls in southern Texas, consisting of organic-rich fossiliferous development acoustics engineer at Veratek
marine shale deposited during the Late Cretaceous. A joint in- Design. He previously worked for University
Lands as a geoscience associate and earlier as
version of triple-combo, spectral gamma-ray, and elemental
a geologist and data analyst at W.D. Von
capture spectroscopy logs was initially conducted to estimate Gonten Laboratories. He holds both a B.S.
volumetric concentrations of minerals, porosity, and fluid satu- degree and Ph.D. in geology from Texas A&M
ration. Effective elastic properties such as Young’s modulus and University. While at Texas A&M University,
Poisson’s ratio as well as minimum horizontal stress were then Matthew was a ConocoPhillips Research
estimated as a function of depth. Rock classification was finally Fellow in 2009 and 2010. His research interests
performed based on geologic texture and geochemical pro- include unconventional reservoirs, x-ray
perties, as well as well-log–based estimates of petrophysical and fluorescence, geochemistry, compositional
data analysis, sequence stratigraphy,
geomechanical properties.
quantitative stratigraphy and its application in
The introduced method was applied to two wells located in basin analysis, and unconventional reservoir
the oil window of the Eagle Ford Formation (average gas–oil characterization.
ratio of less than 2000 SCF/STB). The results of the integrated
rock classification demonstrated similar organic richness and Zoya Heidari ~ Hildebrand Department of
petrophysical properties in both wells. However, the geo- Petroleum and Geosystems Engineering, The
University of Texas at Austin, Austin, Texas;
mechanical rock classification as derived from in situ stress
zoya@utexas.edu
profiles suggests higher proportions of the rock class with better
Zoya Heidari is an associate professor in the
completion quality (55% of net thickness in one well vs. 34% in
Hildebrand Department of Petroleum and
the other well). A 90-day report on the cumulative oil pro-
Geosystems Engineering at The University of
duction of this well shows an additional 11,000 bbl (i.e., 20% Texas at Austin. Before joining The University of
Texas at Austin, she was an assistant professor
at Texas A&M University in College Station
Copyright ©2021. The American Association of Petroleum Geologists. All rights reserved. from September 2011 to August 2015. Zoya
Manuscript received December 22, 2016; provisional acceptance January 9, 2018; revised manuscript was the founder and director of the Texas A&M
received May 25, 2018; revised manuscript provisional acceptance December 10, 2018; 2nd revised Joint Industry Research Program on Multi-
manuscript received May 1, 2019; 2nd revised manuscript provisional acceptance June 19, 2019; 3rd
revised manuscript received November 27, 2019; 3rd revised manuscript provisional acceptance
Scale Formation Evaluation of Unconventional
December 11, 2019; 4th revised manuscript received February 23, 2020; final acceptance July 28, 2020. and Carbonate Reservoirs from 2012 to 2015.
DOI:10.1306/12222016520

AAPG Bulletin, v. 105, no. 7 (July 2021), pp. 1357—1381 1357


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
She has been the founder and director of The more oil production) compared to the second well. This ob-
University of Texas at Austin Industrial Affiliates servation was accounted for by geomechanical properties and
Research Program on Multi-Scale Rock Physics the distribution of rock classes that differentiate reservoir quality
since 2016. She holds a Ph.D. (2011) in and production.
petroleum engineering from The University of
Texas at Austin. Zoya is one of the recipients of
the 2020 Society of Petrophysicists and Well
Log Analysts Young Professional Technical
Award, the 2019 European Association of INTRODUCTION
Geoscientists and Engineers Arie van Weelden
Award, the 2019 American Institute of Mining, Reliable rock classification in organic-rich mudrocks should con-
Metallurgical, and Petroleum Engineers sider petrophysical, compositional, and geomechanical properties.
Rossiter W. Raymond Memorial Award, the Not only does the total thickness of the good-quality unconven-
2019 Society of Petroleum Engineers (SPE) tional reservoir need to be considered but so does the spatial
Distinguished Membership Award, the 2019 distribution of the good-quality reservoir (impacting efficiency of
departmental teaching award from the
hydraulic fracking, as well as hydrocarbon and water production).
Hildebrand Department of Petroleum and
Geosystems Engineering, the 2017 SPE Cedric Properties such as storage capacity, fluid saturation, and mineralogy,
K. Ferguson Medal, the 2016 SPE Regional along with quantity and type of organic matter (i.e., kerogen type),
Formation Evaluation Award, and the 2015 SPE are important factors in formation evaluation of organic-rich
Innovative Teaching Award. Her research mudrocks. Furthermore, inclusion of elastic properties in rock
interests include petrophysics, borehole classification is essential for the assessment of hydraulic fracturing
geophysics, rock physics, inverse problems, performance. However, because of the complexity in matrix
completion petrophysics, and reservoir composition and pore structure of organic-rich mudrocks, as-
characterization of unconventional and
sessment of these parameters is challenging and requires proper
carbonate reservoirs.
integration of core and well-log data.
Michael M. Tice ~ Department of Ideally, a rock classification would include both inorganic and
Geology and Geophysics, Texas A&M organic geochemical data. Inorganic geochemical data (i.e., x-ray
University, College Station, Texas; mtice@ fluorescence [XRF] or elemental capture spectroscopy [ECS]
geos.tamu.edu logs) are useful because they are quantitative and can be cor-
Michael M. Tice is an associate research related with mineralogy (as validated with x-ray diffraction
scientist in the Department of Geology and [XRD]). Rock-Eval and LECO total organic carbon (TOC) can
Geophysics at Texas A&M University. He be used to quantify and characterize the organic components in
earned his Ph.D. in geology from Stanford
mudrocks. Analysis of trace element concentrations (typically by
University in 2006 and worked as a
postdoctoral researcher at the California XRF) in the rocks provides insights into paleoproductivity and
Institute of Technology from 2005 to 2007. paleoredox conditions at the time of deposition of mudrocks
Michael’s research interests focus on (Wright et al., 2010). Redox-sensitive trace metals, which are less
sedimentary geology and geochemistry, soluble under reducing conditions, are enriched in oxygen-
geobiology, and water–rock interactions. depleted sedimentary facies. However, some trace elements re-
quire both reducing conditions and organic matter to accumulate
ACKNOWLEDGMENTS appreciably in shales. Thus, elements such as molybdenum (Mo),
vanadium (V), and uranium (U) provide information about the
The work reported in this paper was funded by
redox state in the water column during deposition, as well as the
the Joint Industry Research Program on Multi-
Scale Formation Evaluation of Unconventional quality (and availability) of organic matter (Lewan, 1984; Brumsack,
and Carbonate Reservoirs. We thank the man- 2006; Tribovillard et al., 2006).
agement and technicians at the W.D. Von Gonten Geoscientists have introduced different techniques to classify
Laboratories for their support and for making sedimentary rocks based on similarities in particular properties.
their facility available for the experimental work Among previous rock classification studies, Archie (1952) classified
reported in this paper. Special gratitude goes carbonates based on their pore structure. Archie’s porosity-
to Andrew Russell for his technical support based classification was among the earliest to differentiate lime-
related to x-ray diffraction measurements.
stones based on petrophysical properties for the selection of

1358 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
reservoir-quality rocks (Archie, 1952). The increase classification across the United States shale plays such
in availability and use of thin-section images, par- as Marcellus Shale and Bakken Formation (Wang and
ticularly for interpretation of depositional environ- Carr, 2013; Bhattacharya et al., 2016). The choice of
ments, shaped the texture-based rock classification classification method can impact the assessment of
schemes such as that of Folk (1962) and Dunham these challenging reservoirs.
(1962). Folk (1962) introduced a detailed texture- Production of hydrocarbons from organic-rich
based classification that incorporated grain size, mudrocks is not simply dependent on petrophysical
roundness, sorting, and packing as well as grain properties. To maximize wellbore–reservoir contact
composition using microscopic images. In a similar and create high-permeability pathways for hydro-
approach, Dunham (1962) classified carbonate rocks carbon flow, practices such as horizontal drilling and
based on the ratio of grain to matrix, into two main hydraulic fracturing are widely used. Hence, geo-
groups: grain supported and mud supported. Cur- mechanical evaluation in organic-rich mudrocks is
rently, both the original and modified versions of the required to model fracture geometry as part of the
carbonate classification of Dunham (1962) are used planning of well completions for improving their ef-
for classifying carbonates and interpreting depositional fectiveness. Fracture propagation and geometry are
environments. influenced by variables such as the frequency and
Various methods and algorithms have been pro- orientation of interfaces, which include natural frac-
posed to characterize and classify organic-rich mud- tures, lamination, and bedding (Suarez-Rivera et al.,
rocks. Kale et al. (2010) studied 800 core plugs from 2006, 2013). To account for elastic properties in a
four wells to identify and group similar facies using well-log–based rock classification, Saneifar et al. (2014)
a quantitative core-based classification. Statistical included a brittleness index as an additional factor of
methods such as principal component analysis and rock quality, along with organic richness, kerogen
clustering algorithms were used to group similar rocks porosity, and quartz and illite volumes. This inte-
based on porosity, mineralogy, TOC, and capillary grated study used a predetermined number of geo-
pressure measurements. Although, quantitative core- logical facies in the Haynesville Shale as an input to
based classifications are useful, the data acquisition the classification. The results of the classification were
requires a substantial amount of laboratory time and then crossvalidated with the thin-section petrography
cost. In addition, core samples are not available from images for each class. Aderibigbe et al. (2016) in-
most wells. Therefore, well-log–based classifica- corporated stress profiles in a production-oriented
tions are favored by the industry, providing complete rock classification applied to the Wolfcamp Forma-
vertical coverage and availability in many wells at tion. The completion intervals were selected based on
lower costs. In an example of a well-log–based study, the final results of the integrated rock classification.
Popielski et al. (2012) conducted a classifica- Previous publications have demonstrated the signifi-
tion using k-means clustering and factor analysis cant impact of reliable rock classification on enhance-
(i.e., used in reduction of the dimension of input ment of formation evaluation and subsequently an
data), to group rocks with similar well-log re- improved selection of completion intervals (Suarez-
sponses, and well-log–derived estimates. Furthermore, Rivera et al., 2011; Petriello et al., 2013; Aderibigbe
Petriello et al. (2013) introduced a well-log–based et al., 2016). However, the proposed workflow
multivariate classification scheme that employed a included geomechanical properties as input for
heterogeneous rock analysis algorithm. Similar rock classification.
patterns were identified and grouped along the zone In this paper, a core-based geochemical and
of interest. Petrophysical and geochemical measure- geological facies classification in the lower Eagle
ments were conducted on more than 200 core sam- Ford (LEF), one of two subsurface divisions in Eagle
ples, for which gas-filled porosity, pressure decay Ford Formation (Donovan et al., 2016), was initially
permeability, and TOC were the discriminant fac- conducted to determine the number of petrophysical
tors between rock classes in the Haynesville Shale. rock types. Then, by calibrating with the avail-
Following an increasing interest in production able core data, petrophysical, compositional, and
from organic-rich mudrocks in the recent years, nu- geomechanical properties of the rocks were de-
merous studies have been published on the lithofacies rived, solely using well logs in the model well. The

AMIN ET AL. 1359


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
well-log–based properties were used to conduct a definition of the classification methods based on geo-
petrophysical and geomechanical rock classification. logical (lithology), geochemical, petrophysical, and
The use of well logs to estimate in situ stress profiles geomechanical properties. To identify the reservoir-
as an input to the geomechanical rock classification is quality rock classes, the results of the petrophysical,
the uniqueness of this study and is an improvement geological, and geochemical rock classifications are
over previous approaches, which commonly compared integrated. The classification outcome of this inte-
geomechanical properties in a qualitative manner. The gration is used for recommending target intervals for
introduced rock classification method is practical in the lateral well placement and improvement of hydraulic
field and is applicable to wells for which core data are fracture design. Finally, both well-log–based rock
not available. In this paper, a second well was used classification methods are applied to a new test well
to conduct both petrophysical and geomechanical rock to recommend the best candidate zones for com-
classification, and the results were compared with the pletions and investigate possible factors affecting well
first well, including production performance. performance. Figure 1 illustrates the aforementioned
workflow.

METHOD Laboratory Measurements

This section covers a list of procedures and data types Five types of laboratory measurements were acquired in
used in this study. Initially, the laboratory measure- the first well (well 1). Two types of XRF measurements
ments are described, which are the data generated for were used in this study. The first is generally referred to
this study. Then described are the following proce- as handheld or portable energy-dispersive XRF (ED-
dures: well-log–based petrophysical evaluation, depth- XRF) because the device itself is mobile and can detect
by-depth assessment of geomechanical properties, and elements between Mg and U on the periodic table. The

Figure 1. A workflow illustrating the methods applied for the geologic, geochemical, petrophysical, and geomechanical classification
schemes included in this study. 2-D = two-dimensional; XRF = x-ray fluorescence.

1360 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
ED-XRF measurements were performed at 353 sam- an average of one sample per 5 m, in the LEF interval.
pling points. The ED-XRF data were used to charac- Correlations of element to mineral and mineral to
terize the log-scale elemental composition. The second mineral are acquired as additional constraints in the
type of XRF is called micro-XRF. For this type of XRF, multimineral analysis to decrease nonuniqueness of
a tabletop Horiba XGT7000 micro-XRF analyzer, with the inversion results and to improve the reliability of
a spatial resolution of 10 mm per pixel, was used to the estimated properties. Finally, the well-log–based
make elemental maps (~0.5 mm2). For organic geo- estimates of mineral concentrations are compared
chemistry, 175 LECO TOC measurements were made against core XRD measurements to validate the re-
by GeoMark and 28 of those samples were processed liability of the multimineral model. Water saturation
for Rock-Eval. Also, 10 XRD measurements were (Sw) in the LEF is estimated using the modified Si-
made on core from well 1 (courtesy of W.D. Von mandoux model (Bardon and Pied, 1969). The con-
Gonten Laboratories). From the core well 2, three types stant parameters in the model are obtained through
of lab data were available: TOC, XRD, and slabbed calibration against core measurements. The same con-
core photography. This study also made use of the stant parameters are used for Sw assessment in wells in
preexisting data from a set of 24 thin sections made which core data are not available.
for other studies on the core from well 1.
Assessment of Geomechanical Properties
Well-Log–Based Petrophysical Evaluation
In the absence of acoustic logs, a self-consistent ap-
proximation (SCA) model is used to calculate depth-
The LEF boundaries are determined based on gamma
by-depth elastic properties such as bulk and shear
ray and U concentration from spectral gamma-ray
moduli (Berryman, 1995). The inputs to the SCA
logs. A significant increase in total gamma ray and U
equations are (1) log-derived volumetric concentration
content defines the boundary, when crossing from
of each constituent, (2) elastic moduli specific to each
the LEF to the upper Eagle Ford (UEF). The UEF
component, and (3) the geometric shape factor asso-
is the upper division of the Eagle Ford Formation
ciated with that component. Assumptions behind the
(Donovan et al., 2016). Furthermore, a unique pyrite-
SCA theorem are that the constituents of the mix-
rich marker bed is present at the transition to LEF
ture are immiscible and in an isotropic elastic media
from the UEF (Lock et al., 2010). At the base of the
(Berryman, 1995). The possible ranges for elastic moduli
LEF, the boundary with the Buda Limestone is well
of the mineral constituents were obtained from
established because of the changes in basic well-log
previously published literature (Prasad et al., 2002;
responses such as gamma ray, neutron porosity, and
Vanorio et al., 2003; Mavko et al., 2009; Sone and
density logs (i.e., Donovan and Staerker, 2010; Donovan
Zoback, 2013).
et al., 2012). The TOC is estimated using a correla-
The SCA equations (Berryman, 1995), which are
tion between TOC (wt. %) from core measurements
simultaneously solved for composite bulk and shear
and bulk density from well logs. Density-based esti-
moduli, are described by
mates of TOC have been widely used in the industry
through either predefined correlations derived for N

�x
 
certain formations or core-based–derived correla- i Ki - KpSC P pi = 0 (1)
i=1
tions (Schmoker, 1979). Assuming a constant kero-
and
gen density of 1.2 g/cm3, volumetric concentration of
kerogen is then estimated as a function of depth. N

�x
 
The estimated volume of kerogen is an input to the i mi - mpSC Qpi = 0 (2)
i=1
inversion-based multimineral analysis workflow.
Triple-combo and ECS logs are jointly interpreted where xi is the volumetric concentration of compo-
to estimate mineral concentrations and petrophysical nent i; N is the total number of constituents; K*SC
properties, such as porosity and fluid saturations, as a and m*SC are the bulk and shear moduli of the
function of depth. In this process, the determination host rock; Ki and mi are the bulk and shear moduli
of mineralogy is based on XRD measurements, with of the component i; and P*i and Q*i are factors

AMIN ET AL. 1361


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
1362

on 10 February 2023
by Ovidiu Gheorghe Pinca
Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation
Figure 2. The location of the two wells evaluated in this paper. They are located approximately 32 km apart in the Maverick Basin at the Eagle Ford Shale play, north of Edwards shelf
margin (modified from Energy Information Administration [eia], 2014). The GOR is reported in SCF/STB.
on 10 February 2023
by Ovidiu Gheorghe Pinca
Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
Figure 3. Lower Eagle Ford (LEF) field example: conventional well logs; estimates of petrophysical, compositional and elastic properties; and the results of geochemical, geological,
and petrophysical classifications. Tracks from left to right include track 1: depth; track 2: high-resolution gamma ray (HGR), caliper (HCAL); track 3: array induction resistivity logs

AMIN ET AL.
(AT10–AT90); track 4: neutron porosity (NPHI) (in water-filled limestone units) and bulk density (RHOB); track 5: estimates of volumetric concentrations of minerals; track 6: estimates
of total porosity (POR); track 7: estimates of water saturation (Sw); track 8: well-log–based estimates of total organic carbon (TOC) (derived from bulk density log) compared to core
measurements; track 9: estimates of Young’s modulus (YM); track 10: estimates of Poisson’s ratio (PR); track 11: outcome of the geochemical rock classification (RC); track 12: outcome
of the geological RC; track 13: outcome of the well-log–based petrophysical RC; track 14: estimates of minimum horizontal stress gradient (MHS-grad); and track 15: outcome of the RC
based on stress profiles (SPRC). GAPI = API gamma-ray unit; UEF = upper Eagle Ford.

1363
1364

on 10 February 2023
by Ovidiu Gheorghe Pinca
Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation
Figure 4. Well 2, lower Eagle Ford (LEF) field example: conventional well logs; estimates of petrophysical, compositional, and elastic properties; and the results of the petrophysical
classification. Tracks from left to right include track 1: depth; track 2: stratigraphic zones; track 3: high-resolution gamma ray (HGR); track 4: array induction resistivity logs (AT10–AT90);
track 5: neutron porosity (NPHI) (in water-filled limestone units) and bulk density (RHOB); track 6: estimates of volumetric concentrations of minerals; track 7: estimates of total porosity
(POR); track 8: well-log–based estimates of water saturation (Sw) compared to core measurements; track 9: well-log–based estimates of total organic carbon (TOC) (derived from bulk
density log) compared to core measurements; track 10: estimates of Young’s modulus (YM); track 11: estimates of Poisson’s ratio (PR); track 12: estimates of minimum horizontal stress
gradient (MHS-grad); track 13: outcome of the well-log–based petrophysical rock classification based on estimates of POR, Sw, TOC, volumetric concentration of clay, YM, and PR; and
track 14: outcome of the rock classification (RC) based on stress profiles (SPRC). GAPI = API gamma-ray unit; UEF = upper Eagle Ford.
dependent on the shape assigned to each inclusion. according to the carbonate classification by Dunham
Shape factors for each component are assigned (1962), (2) degree of lamination, and (3) type of the
based on an observation of two-dimensional micro- existing fossils.
XRF elemental maps that show the distribution of
elements. At the microscale, individual grains (e.g.,
Geochemical Rock Classification
calcite, quartz, zircon) are differentiable based on
their elemental composition and thus for each min-
To classify rocks based on inorganic geochemical data
eral the typical grain shape and size could be mea-
from the LEF, a sample rate of one measurement for
sured. Finally, Young’s modulus (E) and Poisson’s
every 15 cm of core for XRF measurements were
ratio (v) are calculated via
collected. Through the analysis of major and trace
9KpSC mpSC element concentrations, the formation is divided into
E= (3)
3KpSC + mpSC different chemostratigraphic units. The geochemical
and classification provides information on vertical varia-
tions in paleoproductivity and redox conditions in
3Kp - 2mpSC depositional environments, as well as postdepositional
n =  SCp  (4)
2 3KSC + mpSC processes such as diagenesis, organic maturation, and
hydrocarbon generation. After classification, the
Furthermore, for planning well completion design
Rock-Eval data are reviewed by class to provide ad-
and hydraulic fracture design, depth-by-depth mini-
ditional information of source-rock properties for
mum horizontal stress (MHS) gradient as a function
each class, such as hydrogen index (HI), oxygen in-
of depth, modified after the Eaton’s equation (Eaton,
dex, and thermal maturity.
1969), is estimated via
n  
shmin = sv - ab Pp + ab Pp (5) Petrophysical Rock Classification
1-n
where shmin is the MHS, sv is the overburden stress, After estimating petrophysical, compositional, and
Pp is the pore pressure, v is Poisson’s ratio, and ab is geomechanical properties, a well-log–based rock
Biot’s constant. Bulk density and resistivity logs were classification was conducted, adopting a hierarchi-
used to predict the overburden stress and pore cal clustering algorithm to group similar rocks by
pressure in this case. Biot’s constant at each depth is minimizing the total intraclass variance in each step
estimated via (Ward, 1963). The input (number of rock classes)
K was determined based on the total number of lith-
ab = 1 - (6) ofacies observed from the combination of core and
Km
geochemical studies.
where K is the bulk modulus of the rock includ-
ing both mineral and fluid phases, and Km is the
zero-porosity (i.e., only mineral phase) bulk mod- Geomechanical Rock Classification
ulus. The bulk and shear moduli are estimated
using the SCA method (equations 1, 2). The LEF rocks are classified as low, medium, or high
stress based on the well-log–based estimates of MHS
at each depth.
Geological Rock Classification

Variation in depositional environment moving up-


ward from the base of Eagle Ford is noticeable in the FIELD EXAMPLE: EAGLE FORD FORMATION
texture of the recovered cores. These differences
are identified in the rock texture based on full core The described method was applied to the LEF. This
photographs in addition to thin-section images. The section includes an overview on the geology of the
names of the lithofacies are based on (1) the ratio of Eagle Ford Formation and the results of the described
grain to matrix (percentages of fossils in the matrix) rock classification schemes in two wells (well 1 and

AMIN ET AL. 1365


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
well 2) located in the oil window of the Eagle Ford or bentonitic claystones with very high clay content.
Formation. The packstone–grainstone facies are limestone beds
up to 20 cm thick and composed of more than 75%
calcite. Minerals found in the Eagle Ford Formation
Geologic Background of the Eagle Ford mainly consist of calcite, dolomite, quartz, pla-
Formation gioclase, mixed-layer clays, and pyrite (Jennings
and Antia, 2013). Additionally, minor amounts of
The Eagle Ford Formation, located mostly within
K-feldspars, siderite, and apatite are reported. Clay
Texas, is one of the largest North American shale
minerals typically have high percentages of illite
plays in terms of liquid hydrocarbon production. As
mixed layers, kaolinite, and minor chlorite content.
of 2014, according to Stoneburner (2017), produc-
The wells used in this study are located in the oil
tion was reported as 1.3 million bbl/day of oil and 4.9
window of the Eagle Ford shale play in the Maverick
BCF/day of natural gas.
Basin, just north of Edwards reef margin. Figure 2
The deposition of the Eagle Ford sediments
illustrates the geographic locations of the two wells.
represented a long-term flooding (mid-Cenomanian
In terms of mineralogy, calcite, quartz, illite mixed
through the end of Turonian) of the Comanche Reef
layers, and kaolinite are dominant minerals in the
platform throughout most of Texas during the mid-
Cretaceous. The basins on the Comanche platform depth intervals of interest.
and its margins initially experienced some restriction Rock-Eval pyrolysis measurements show high
vertical variation in organic properties such as TOC,
and episodic anoxia (Wehner et al., 2015). The Eagle
Ford sediments primarily consist of alternating HI, and thermal maturity. Type II oil-prone kerogen
mudstone or marlstone with limestone, along with is present in well 1, with the temperature at which
organic carbon content up to 10 wt. % TOC (Sun the maximum rate of hydrocarbon generation
et al., 2016), and occasional bentonitic ash beds. occurs during pyrolysis (Tmax) values ranging from
The LEF appears to have been deposited during 430°C to 450°C. For the petrophysical model, an
basin restriction, storm deposition, and episodic averaged vitrinite reflectance estimate (VRE) (data
anoxia; these factors contributed to the ultimate provided by GeoMark Research based on Rock-
incorporation of organic matter (Wehner et al., Eval) value of 0.81% was used to estimate a
2015). The variations of depositional condition kerogen density of 1.2 g/cm3, following the ap-
make the Eagle Ford Formation highly heteroge- proach of Alfred and Vernik (2013) and Jagadi-
neous in the vertical dimension (at the centimeter san et al. (2017). A similar vitrinite reflectance value
and decimeter thickness scales). Some of this ver- of 0.76% is reported in an experimental study on
tical heterogeneity can be predicted by sequence isolated kerogen samples from the Eagle Ford For-
stratigraphy (Donovan and Staerker, 2010; Donovan mation (Yang et al., 2016). Although using Tmax
et al., 2012; Donovan, 2016; Wehner et al., 2017). values to estimate VRE is known to be problematic
Gardner et al. (2013) investigated the lateral var- (e.g., Zumberge et al., 2016), no additional data were
iation of Eagle Ford strata. They reported that the available to constrain the Tmax-based VRE and the
scale of horizontal variation is minor at distances goal was an estimate of kerogen density.
of 5 km. Factors that influence vertical variation
include sea level at time of deposition, sediment
source, organic productivity, storm frequency, and Petrophysical Evaluation
dominant diagenetic mechanisms.
The Eagle Ford marlstones and limestones can Petrophysical and compositional properties were es-
be generally classified into one of Dunham’s classes, timated in the LEF section of the two wells. Tracks
commonly packstones and grainstones, with the 2–8 of Figures 3 and 4 illustrate the well logs and
majority of grains being silt size. It does contain other well-log–based estimates of petrophysical and com-
lithologies such as ash beds composed of bentonites positional properties, whereas Figure 5 shows ex-
(sometimes tonsteins). The ash beds, commonly less amples of core x-ray elemental distribution images in
than 5 cm thick, are classified as massive bioturbated core samples from LEF that were used to calculate

1366 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
Figure 5. Mineral grains identified using two-dimensional x-ray elemental distribution maps from the lower Eagle Ford core samples.

shape factors used in the SCA approach for geo- assessment of acoustic compressional and shear
mechanical properties. slowness. Although larger error was observed in
Based on the analysis of x-ray images, quartz and estimates of shear-wave slowness, the approxima-
pyrite appear to be more spherical shaped. An aspect tion remains reliable for most of the depth interval.
ratio (a) of 1 is assigned to these minerals. However, Because of the limited number of x-ray images, a
for clay minerals in which the structure consists of single model was used for the LEF interval. How-
nanometer-scale layers, an a of 0.01 was assumed. ever, the availability of several high-resolution
Figure 5 shows the identified calcite grains to be core images can potentially enable the assign-
compacted and interbedded with clays in an ellipti- ment of zonal shape factors based on rock facies
cal form at specific zones and in another place more classification, resulting in a more robust and reliable
spherically shaped. Therefore, two different shape model. After application of the calibrated SCA model
factors were assigned to calcite constituents (0.1 for
elliptical and 1 for spherical grains). Table 1 sum- Table 1. Input Parameters to the Self-Consistent Approximation
marizes the input constants that minimized the dif- Model for Geomechanical Properties
ference between estimated acoustic-wave slowness
Rock Constituents K, GPa m, GPa a
and those from acoustic well logs in the wells in
which acoustic measurements are available (e.g., well Calcite (spherical) 75 40 1
2). The same input parameters are used as input to Calcite (elliptical) 75 40 0.1
the SCA model in wells in which acoustic logs are not Quartz 38 44 1
available to estimate elastic properties. Pyrite 147 132 1
Figure 6 shows the estimates of petrophysical Illite mixed layers 8 6 0.01
and compositional properties, as well as the comparison Kaolinite 6 2 0.01
between the estimates of compressional- and shear- Kerogen 5 3 0.1
wave slowness from SCA and acoustic well logs in Abbreviations: a = aspect ratio; m = shear modulus of the rock; K = bulk modulus
well 2. The SCA results provide a higher-resolution of the rock.

AMIN ET AL. 1367


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
Figure 6. Well 2, lower Eagle Ford field example: conventional well logs, estimates of volumetric concentration of minerals, and acoustic
well logs. Tracks from left to right include track 1: depth; track 2: high-resolution gamma ray (HGR), caliper (HCAL); track 3: array
induction resistivity logs (AT10–AT90); track 4: neutron porosity (NPHI) (in water-filled limestone units), bulk density (RHOB); track 5:
estimates of volumetric concentrations of minerals; track 6: self-consistent approximation (SCA)–based estimates of compressional-wave
slowness compared to the acoustic well-log measurement (DTCO); and track 7: SCA-based estimates of shear-wave slowness compared to
the acoustic well-log measurement (DTSH). DTC = delta-T compressional; DTS = delta-T shear; GAPI = API gamma-ray unit.

1368 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
on 10 February 2023
by Ovidiu Gheorghe Pinca
Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
AMIN ET AL.
Figure 7. Well 1, lower Eagle Ford (LEF) field example: conventional well logs and vertical distribution of major and trace elements in the LEF. Tracks from left to right include track 1:
depth; track 2: high-resolution gamma ray (HGR), caliper (HCAL); track 3: array induction resistivity logs (AT10–AT90); track 4: neutron porosity (NPHI) (in water-filled limestone units),
bulk density (RHOB); track 5: chemostratigraphy of LEF; tracks 6–15: calcium (Ca), silicon (Si), aluminum (Al), potassium (K), iron (Fe), sulfur (S), uranium (U), molybdenum (Mo),
vanadium (V), and nickel (Ni) weight concentrations obtained from core x-ray fluorescence measurements; and track 16: total organic carbon (TOC) weight concentration (LECO TOC
measurements). GAPI = API gamma-ray unit; RC = rock classification.

1369
to wells 1 and 2, values of Young’s modulus and Pois- Fe, S) and trace elements (U, V, Ni, Mo) and TOC
son’s ratio were estimated as a function of depth. Tracks contents. Figure 7 displays the core XRF measure-
9 and 10 of Figure 3 and tracks 10 and 11 of Figure 4 ments obtained in the LEF as a function of depth.
illustrate the final estimates of Young’s modulus and Furthermore, Figure 8 displays a matrix scatterplot
Poisson’s ratio in well 1 and well 2, respectively. of the geochemical data. Through analysis of the
results, distinguishing geochemical characteristics
among the classes are highlighted: the lowest unit,
Chemostratigraphy in the LEF LEF5, is at the base of LEF, on top of the Buda
Limestone, and is chemostratigraphically distinct.
The LEF was divided into five geochemical classes The U content is the highest in this interval, and the
(LEF1 to LEF5, in ascending order from the bottom Si/Al ratio is very low (~1.5–3). In terms of organic
upward) based on variations in major (Ca, Si, Al, K, geochemistry, TOC is moderate to high (4–5 wt. %).

Figure 8. Matrix scatterplot showing the correlation between the concentration of major elements (Ca, Al, Si, Fe, S), trace elements (Mo,
U, Ni, V), and total organic carbon (TOC) in the lower Eagle Ford, obtained from core x-ray fluorescence measurements. Although some
elements show poor correlation or exhibit inverse correlation, the following correlated element pairs include Mo–U, Mo–TOC, Fe–S, Fe–Al,
and Si–Al. The trace elements Mo, U, Ni, and V are reported in parts per million; the rest are in weight percent.

1370 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
The next unit, LEF4, is distinguished by its lower classifications are defined by inferring the dominant
TOC content (2–4 wt. %). The LEF3 is similar to oxidant used for microbial respiration in the water
LEF4, except it contains higher TOC (>4 wt. %) and column. Such classifications are effectively based on
lower Ca content. Transitioning to LEF2, there is a how chemically reducing the environment was at the
significant increase in concentration of trace elements time of deposition. These classifications, or zones, are
such as V, Ni, Mo, and U. At the very top of the LEF, oxic, nitrogenous, manganous, ferruginous, sulfidic,
LEF1 has a linearly decreasing Ca content and, like- and methanic (Canfield and Thamdrup, 2009). Con-
wise, increasing Al and Si content. The V concentra- sistent with previous classifications, sulfidic conditions
tion remains similar to that in LEF2, and a significant in the water column are referred to as euxinic. In
decrease in Ni concentration is observed. Table 2 lists practice, paleoredox conditions corresponding to oxic,
the distribution of the elemental concentrations (i.e., ferruginous, and sulfidic zones can be inferred from
S, U, V, Ni), as well as TOC of the different classes, the relative concentrations of the trace metals U, Mo,
that were used as the basis of discrimination among V, and Ni. For the purposes of this study, oxic, ni-
the geochemical classes. Track 11 of Figure 3 shows trogenous, and manganous conditions are all classified
the chemostratigraphic units in well 1. The key pet-
as oxic, and sulfidic and methanic are classified as
rophysically relevant observations from the aforemen-
sulfidic or euxinic. Both U and Mo can accumulate in
tioned chemostratigraphic analysis are summarized
sediments under various anoxic conditions, but U
below.
tends to accumulate most efficiently under ferruginous
Although both the top and the bottom of LEF
conditions, whereas Mo reduction (and accumulation)
contain clays, the clay type distribution is not the
is not triggered until sulfate reduction begins (sulfi-
same. Kaolinite is significant in the lowermost unit,
dic or euxinic conditions). However, although V
LEF5, as presented by the very low Si/Al ratio (<2)
and Ni are reduced under anoxic conditions (Fe
and low K (0.2–0.3 wt. %). These characteristics
indicate that illite is diluted by the presence of a and sulfate reducing) in the ocean, they are more
clay mineral that is not potassic yet has low Si efficiently incorporated into the sediments by
content. bonding with specific organic molecules under
The best reservoir rock classes, LEF1 and LEF2, ferruginous and sulfidic conditions. According to
appear to be associated with intervals of extreme Lewan (1984), V and Ni concentrations, and their
enrichment of specific trace metals, including V and relative proportions, can convey information about
Ni. These two elements, along with U and Mo, are kerogen type and presence of trapped hydrocar-
considered to be useful indicators for productivity bons. Figure 9 shows a class-by-class distribution of
and preservation of organic matter, as well as for the Rock-Eval parameters, including the amount of
distinguishing among different types of anoxia (Wilde free hydrocarbons present in the sample (S1), HI, and
et al., 2004; Tribovillard et al., 2006). Although pa- Tmax. The S1 and the amount of hydrocarbons
leoredox conditions have been defined classically in generated during pyrolysis (S 2 ) values from
terms of inferred water column oxygen concentra- Rock-Eval measurements confirm that hydrocarbon
tions (i.e., oxic, suboxic, and anoxic), more consistent content is high when V and Ni are high.

Table 2. Vertical Distribution of Core X-Ray Fluorescence Measurements of Elemental Weight Concentrations Arranged by Geochemical
Rock Class in Well 1

Geochemical Rock Classification S, wt. % U, ppm V, ppm Ni, ppm TOC, wt. %
LEF1 1.4 – 0.2 7.0 – 1.3 404 – 102 64 – 24 4.3 – 0.6
LEF2 2.3 – 0.6 7.5 – 1.3 391 – 112 126 – 65 5.0 – 0.8
LEF3 2.5 – 0.7 6.2 – 1.7 176 – 61 81 – 37 4.4 – 0.9
LEF4 2.2 – 0.5 6.9 – 1.5 207 – 97 70 – 33 3.0 – 0.5
LEF5 2.6 – 0.6 12 – 2.0 192 – 62 75 – 34 4.1 – 0.4

Abbreviations: LEF = lower Eagle Ford; TOC = total organic carbon.

AMIN ET AL. 1371


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
The concentrations of trace elements appeared
to correspond more closely to hydrocarbon content
compared to TOC. Although the TOC varies be-
tween geological rock classes, it does not directly cor-
relate with the volume of in situ hydrocarbons.

Lithofacies Classification

Through the analysis of thin sections and core images


obtained from 44 m of drill core, four lithofacies in
the LEF interval were identified. Moving upward
from the base of the Eagle Ford, a transition from a
massive and weakly laminated texture to a repetitive
alternating pair of lithofacies that dominates the
majority of the LEF section was observed. The pair
consists of two main subfacies: foraminiferal mud-
stone and packstones. These facies alternate at different
thickness scales from submillimeter up to decimeter.
Furthermore, mixtures of these facies are common
depending on the scale of observation. Typically, this
pair of facies has minimal or no bioturbation and
transitions between facies are sharp. The foraminif-
eral mudstones consist of bioclastic grains (10–100
mm), carbonate mud, and clays. The bioclastic grains
are mostly foraminiferal tests but can include other
skeletal remains of ostracods, diatoms, and bivalves.
Perhaps with the exception of bivalves, the bioclastic
grains in the LEF are planktonic with no or very few
benthic ones having been reported to date (Lowery
et al., 2014). The packstone–grainstones are essentially
composed of planktonic foraminifera but without
significant clays or other siliciclastics. They tend to
be very light gray in color. Diagenetic packstone–
grainstones look much like the regular packstone–
grainstones but contain many diagenetic features,
including recrystallization, calcite cement, concre-
tions, and dolomitization. Figure 10 displays the core
Figure 9. Summary of the Rock-Eval pyrolysis measure-
photographs that represent different lithofacies.
ments (A) free hydrocarbons present in the sample (S1), (B)
The four identified lithofacies in the LEF include hydrogen index, and (C) temperature at which the maximum
massive argillaceous mudstone, laminated argillaceous rate of hydrocarbon generation occurs during pyrolysis
foraminiferal mudstone, bedded foraminiferal mudstone (Tmax ). The results are presented separately in the five
and nodular limestone, and bedded foraminiferal identified geochemical classes. LEF = lower Eagle Ford.
wackestone and limestone, which are described as
follows.
Eagle Ford. Low counts of foraminifera are iden-
1. Massive argillaceous mudstone: mudstone facies tified in these mudstones, and clay content is high
with little to no discernable bedding structures compared to other lithofacies, consisting of illite
or internal structures are present at the base of mixed layers and kaolinite.

1372 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
2. Laminated argillaceous foraminiferal mudstone: nodular or have thickening–thinning (carbonate
this facies has a medium gray color and millimeter- nodules described by Dawson, 2000). Commonly,
scale laminations with mudstones and limestone internal structures of the nodular limestones are
laminations composed of foraminifera. Commonly, obliterated by pore-filling cementation and recrys-
the proportion of mudstone to foraminifera-rich tallization. The transition between mudstones and
limestone changes in these facies. Reported min- limestones is sharp in this facies. Moreover, because
eralogy indicates carbonate content as low as the bounding surfaces of the nodular laminations are
45%, with total clay content up to 30% in weight, commonly curving, it is possible that in some cases,
consisting of illite and kaolinite. the nodular limestones are laterally discontinuous.
3. Bedded foraminiferal mudstone and nodular lime- 4. Bedded foraminiferal wackestone and limestone:
stone: dark-gray to black illite-rich mudstones with the wackestone is dark gray to black because of very
faint laminations, where the limestone beds are high organic content ranging from 4 to 6 wt. %.

Figure 10. Examples of core images associated with the lithofacies in the lower Eagle Ford, from well 1. Red lines represent sharp
contrast across rock types within lithofacies.

AMIN ET AL. 1373


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
Wackestones contain thin foraminiferal lags and or grainstone, wackestone, and mudstone facies. These
isolated bivalves. The limestone beds with thickness images show a variation in proportions of bioclastic
of a few inches up to 0.3 m are foraminiferal grains to mud. Foraminifers are the most abundant
packstone and grainstones transitioning at fossils in the LEF and are composed of calcite. In some
sharp boundaries with the wackestones. The cases, these fossils were impacted by various de-
abundance of bioclastic grains in this lithotype grees of dissolution. Furthermore, the dark-gray to
is high in the LEF unlike the deeper mudstone black color of the matrix mud is caused by a com-
facies. bination of high organic content and clay minerals
present in these rocks. Based on Dunham’s carbo-
Figure 11 illustrates different examples of the thin- nate classification, the grain-to-matrix ratio is more
section images associated with foraminiferal packstone than 90% for packstones and grainstones. This ratio in

Figure 11. Thin-section examples of the identified facies in the lower Eagle Ford (LEF). (A) Foraminiferal packstone or grainstone;
associated with limestone nodules at the top of LEF. (B) Skeletal packstone; an inoceramid shell is evident in this image. (C) Foraminiferal
wackestone; calcite-filled foraminifera are noticeable in the image. (D) Foraminiferal mudstone; dissolved foraminifera are noticeable in
this image. (E) Laminated foraminiferal mudstone; submillimeter foraminiferal lamina is evident in the image.

1374 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
wackestones is between 10% and 90% and less than
10% for mudstones.

Petrophysical Rock Classification

Inputs to the well-log–based classification include


porosity, S w , TOC, volumetric concentration of
clay, Young’s modulus, and Poisson’s ratio. Track
13 of Figure 3 and track 11 of Figure 4, show the
petrophysical classification results in well 1 and well 2,
respectively. Furthermore, Figure 12 shows crossplots
of the estimates of petrophysical and compositional
properties for both wells. Tables 3 and 4 include sta-
tistics by class of the well-log–based classification
results in well 1 and well 2, respectively.
The petrophysical classes were ranked through
an integrated rock characterization approach. To
fully characterize the petrophysical classes, the
results of the previously described geological and
geochemical classifications were interpreted, with
the results of petrophysical classification presented in
this section. The characteristics of each petrophysical
rock class (PRC) is described as follows.

• The PRC1 occurs interbedded throughout the


rock classes except the massive argillaceous mud-
stones. Geologically, this class is a carbonate with
low porosity and occurs interbedded with other
PRCs, with low hydrocarbon content (i.e., less than
0.3 wt. %).
• The PRC2 corresponds to the massive argilla-
ceous mudstones. This class is notable for kaolinite
content up to 15% (vol. %). The TOC is high (2–3
wt. %), and the rock contains small volumes of free
hydrocarbon.
• The PRC3 is associated with the laminated ar-
gillaceous foraminiferal mudstones. It appears to
be a mudstone with low porosity and high TOC
(~2–3 wt. %).
• The PRC4 is interbedded with PRC1 in the
bedded foraminiferal mudstone and nodular
limestone. It is a porous (~8.5% porosity) mud-
stone with very high TOC content (3–5 wt. %) and
the lowest estimated Young’s modulus among the Figure 12. Well-log–based estimates of (A) total organic
carbon (TOC) versus porosity, (B) TOC versus volumetric con-
other rock classes. The Sw is the highest in this rock
centration of clay (Vclay), and (C) water saturation (Sw) versus
class. porosity in well 1 and well 2. Different colors represent different
• The PRC5 represents the organic-rich forami- petrophysical rock classes (PRC).
niferal wackestone facies. Petrophysically, it is

AMIN ET AL. 1375


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
Table 3. Well 1: Statistics by Class of the Well-Log–Based Rock Classification Results in the Lower Eagle Ford

PRC Porosity, % TOC, wt. % Vclay, % Sw, % YM PR


PRC1 3.5 – 0.6 3.8 – 0.8 12 – 4 17 – 7 39 – 6 0.27 – 0.01
PRC2 5.5 – 0.8 2.8 – 1.1 29 – 4 20 – 3 26 – 4 0.29 – 0.01
PRC3 3.9 – 0.3 2.7 – 0.5 24 – 4 16 – 3 34 – 2 0.27 – 0.01
PRC4 8.5 – 1.1 4.8 – 0.5 14 – 2 28 – 4 22 – 2 0.33 – 0.01
PRC5 7.0 – 1.3 5.0 – 0.7 12 – 4 16 – 3 28 – 3 0.30 – 0.01
Abbreviations: PR = Poisson’s ratio; PRC = petrophysical rock classification; Sw = water saturation; TOC = total organic carbon; Vclay = volume of clay; YM = Young’s modulus.

similar to PRC4, with slightly higher hydrocarbon create effective fracture networks as pathways for
content and lower Sw estimates compared to PRC4. hydrocarbon transport. In tectonically relaxed areas,
hydraulically induced fractures propagate in the vertical
In the determination of reservoir-quality rock classes, direction, perpendicular to the MHS (Zhou et al., 2008).
factors such as storage capacity, organic richness, Furthermore, the required fracture initiation and pro-
thermal maturity of kerogen, and free hydrocarbon pagation pressure depends on the magnitude of the
content in the rocks are considered. Figure 13 shows least principal stress (Hubbert and Willis, 1972).
Rock-Eval measurements of S1 versus TOC in the The MHS gradient in well 1 and well 2 were
LEF, separated by petrophysical class. It is evident estimated using well logs and cross-validated with
from the data that high TOC is not a direct indicator the output from the available fracture closure pressure
of high free hydrocarbon content. For instance, measurements obtained from multistage hydraulic
at TOC concentrations more than 4 wt. %, PRC5 fracture data in well 2. Track 14 of Figure 3 and track
and PRC4 contain significantly higher free hydro- 12 of Figure 4, illustrate the estimated MHS gradient
carbons compared to PRC3 and PRC2. Through in well 1 and well 2, respectively. Track 15 of Figure 3
integration of petrophysical and geochemical data, and track 14 of Figure 4 show the results of geo-
PRC5 and PRC4 were determined as the best
mechanical rock classifications based on stress
reservoir-quality rock classes in the LEF. The PRC5 and
profiles (SPRC), including high MHS (SPRC1),
PRC4 are organic-rich, oil-prone, and porous mud-
medium MHS (SPRC2), and low MHS (SPRC3)
rocks that contain substantial quantities of in situ
levels. The results indicate that the MHS gradient
hydrocarbon.
varies considerably when moving across geologic fa-
cies. The MHS gradient ranges between 8 and 19
Geomechanical Rock Classification Based kPa/m (0.35–0.85 psi/ft), with an average MHS
on Stress Profiles gradient of 13 kPa/m (0.58 psi/ft) in well 1 and 14
kPa/m (0.62 psi/ft) in well 2 in the LEF. Table 5
After identification of the best reservoir-quality rock summarizes the statistics by class of the MHS gradi-
classes in the formation, it is essential to conduct a ent in both wells. The results of the classification
geomechanical evaluation to ensure the ability to based on MHS indicate higher proportions of the

Table 4. Well 2: Statistics by Class of the Well-Log–Based Rock Classification Results in the Lower Eagle Ford

PRC Porosity, % TOC, wt. % Vclay, % Sw, % YM PR

PRC1 4.6 – 0.8 2.1 – 0.8 9– 3 28 – 7 36 – 6 0.28 – 0.01


PRC2 6.6 – 1.1 2.7 – 0.7 20 – 3 28 – 6 26 – 4 0.30 – 0.01
PRC3 4.8 – 0.7 2.4 – 0.6 17 – 2 30 – 3 28 – 5 0.30 – 0.01
PRC4 8.1 – 0.9 3.5 – 0.4 20 – 3 39 – 6 23 – 3 0.32 – 0.01
PRC5 7.0 – 1.0 4.5 – 0.7 17 – 3 18 – 4 25 – 3 0.30 – 0.01

Abbreviations: PR = Poisson’s ratio; PRC = petrophysical rock classification; Sw = water saturation; TOC = total organic carbon; Vclay = volume of clay; YM = Young’s modulus.

1376 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
hydraulic fracturing is based on an integrated analysis
of the conducted rock classifications in the pilot sec-
tion of well 1 and well 2. This integration considers
properties such as organic richness, storage capacity,
fluid saturation, clay volume, and MHS. Additionally,
for well 1, geological and geochemical characteristics
of the selected target interval for decision-making
were included.
Based on the results of the integrated rock clas-
sification reported in Table 6, in well 1, X180–X192 m
(XX430–XX470 ft) was selected as the best target
interval for completions. Geologically, this interval
represents the foraminiferal wackestone facies where
the organic content can reach up to 6 wt. %. Fur-
thermore, the geochemical classification results indi-
Figure 13. Well 1: Rock-Eval pyrolysis measurements of free
hydrocarbons present in the sample (S1) versus total organic cated significant enrichment of trace elements (e.g., V
carbon (TOC) in the lower Eagle Ford. Different colors represent and Ni), which correlates with higher hydrocarbon
different petrophysical rock classes (PRC). contents. This target interval contains the best
reservoir-quality class (PRC5) combined with PRC1,
low-stress rock class (i.e., SPRC3) in well 1 (55% of which represent the thinly bedded limestone units.
gross thickness) compared to well 2 (34% of gross However, because of their limited thickness, these
thickness). Figure 14 shows the histogram compari- limestone beds are not expected to act as a barrier to
son of MHS gradient in the LEF for well 1 and well 2. the induced hydraulic fractures. Geomechanically,
The green bars represent SPRC3 where the MHS the selected target zone is a low-MHS rock, which
gradient is lower than 12.4 kPa/m (0.55 psi/ft), and makes it a good candidate for fracture initiation and
red bars represent SPRC1 where the MHS gradient is propagation. A 1500-m horizontal lateral at X186 m
higher than 14.7 kPa/m (0.65 psi/ft). Because rocks (XX450 ft) was landed in well 1. Production history
of SPRC3 can fracture at lower pressure (Figure 14),
it is considered a better completion quality class
compared to SPRC2 and SPRC1.

Integrated Completion-Based
Recommendation

Final recommendation of the best target interval for


horizontal well placement followed by multistage

Table 5. Statistics by Class of the Minimum Horizontal Stress


Gradient in Wells 1 and 2 in the Lower Eagle Ford

MHS Gradient, kPa/m (psi/ft)

Rock Class Well 1 Well 2


SPRC1 16.5 (0.73) 16.1 (0.71)
SPRC2 14.0 (0.62) 14.0 (0.62)
Figure 14. Comparison of histograms of the minimum hori-
SPRC3 10.9 (0.48) 12.4 (0.55) zontal stress (MHS) gradient for wells 1 and 2 in the lower Eagle
Abbreviations: MHS = minimum horizontal stress; SPRC1 = rock class with high Ford. Different colors represent different rock classes. SPRC1 =
MHS profile; SPRC2 = rock class with medium MHS profile; SPRC3 = rock class rock class with high MHS profile; SPRC2 = rock class with me-
with low MHS profile. dium MHS profile; SPRC3 = rock class with low MHS profile.

AMIN ET AL. 1377


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
Table 6. Well 1: Quantitative Analysis of the Petrophysical, Geochemical, and Geomechanical Characteristics of Each Petrophysical Rock
Class and Its Representative Lithofacies

Petrophysical Rock
Class PRC1 PRC2 PRC3 PRC4 PRC5
Massive Argillaceous Laminated Argillaceous Foraminiferal Foraminiferal
Lithofacies Limestone Mudstone Foraminiferal Mudstone Mudstone Wackestone
Calcite, vol. % 60 –4 47 – 4 56 – 3 49 – 6 54 – 7
Quartz, vol. % 15 –3 10 – 3 10 – 2 18 – 3 16 – 4
Clays, vol. % 12 –4 29 – 4 24 – 3 14 – 3 12 – 4
Porosity, % 3.5– 0.7 5.5 – 0.8 3.9 – 0.3 8.4 – 1.1 7.0 – 1.3
Sw, % 18 –7 20 – 3 15 – 3 29 – 4 17 – 3
TOC, wt. % — 2.8 2.7 – 0.4 4.8 – 0.5 5.0 – 0.7
S1, mgHC/g — 2.3 2.4 – 0.5 6.0 – 1.3 7.6 – 2.0
S2, mgHC/g — 2.7 2.7 – 0.7 4.6 – 1.4 6.0 – 1.0
Tmax, °C — 447 448 – 6 441 – 7 443 – 8
HI — 63 66 – 7 131 – 13 127 – 20
MHS-grad, psi/ft 0.50 – 0.1 0.56 – 0.05 0.47 – 0.03 0.76 – 0.05 0.63 – 0.09
Mean and standard deviation values are reported for each class.
Abbreviations: — = not applicable; HI = hydrogen index; MHS-grad = minimum horizontal stress gradient; PRC = petrophysical rock classification; S1 = free hydrocarbons
present in the sample; S2 = amount of hydrocarbons generated during pyrolysis; Sw = water saturation; Tmax = temperature at which the maximum rate of hydrocarbon
generation occurs during pyrolysis; TOC = total organic carbon.

shows a 90-day production of greater than 50,000 BOE hydrocarbons and a water production of 1000
BOE. bbl in the first 90 days after completions. In com-
In well 2, X330–X338 m (XX925–XX950 ft) parison, well 2 produced 43,000 BOE hydrocarbons
was selected as the best target interval for comple- and 10,000 bbl of water over the same time interval.
tions. This target is a low-MHS interval adjacent to In wells 1 and 2, very similar petrophysical prop-
PRC5 rocks, which represents the best PRC. In both erties such as organic richness, storage capacity, fluid
wells, the upper part of the LEF appears to be saturations, and total volumetric concentration of clay
ductile, with MHS estimates of greater than 15.8 were observed in the reservoir-quality rock classes.
kPa/m (greater than 0.75 psi/ft). A horizontal lateral However, the geomechanical analysis showed a different
with a length of more than 1800 m was drilled and distribution of stress profiles among these wells.
hydraulically fractured in well 2. In most stages, Overall, well 1 contains higher proportions of
the fracture closure pressure gradient of 0.68 psi/ft higher-quality reservoir rock and a better vertical
(15.4 kPa/m) was observed. It is possible that the continuity of low-MHS rock units, as well as a
upper 9 m of the LEF in these wells acts as a slightly lower average MHS in the LEF, compared
fracture barrier and restricts the propagation up- to well 2. Furthermore, thin-bedded limestone
ward out of the target formation. Production his- intervals in well 1 and thicker limestone layers in
tory of well 2 shows a 90-day production of greater well 2 were observed, which can potentially act as
than 40,000 BOE, which is approximately 20% barriers to hydraulic fractures. This difference in
less than well 1. stratigraphic distribution of rock classes and MHS
can impact the vertical propagation of hydrauli-
cally induced fractures and, consequently, the total
Comparison of Well Productivity stimulated reservoir volume. Although the pres-
ence of natural fractures was not investigated because
Figure 15 illustrates the cumulative 90-day hydro- of the lack of image logs, the different geomechanical
carbon and water production for well 1 and well 2. characteristics among these wells could potentially
Well 1 produced a cumulative volume of 54,000 explain some of the well production difference.

1378 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
of low-MHS rock class, produced an excess of 11,000
BOE (20 vol. %) in the first 90 days after comple-
tions. The results demonstrated that the difference in
production between the two wells could not be ex-
plained by only relying on the estimated petrophysical
properties of the formation. Geomechanical properties
such as in situ stresses and presence of natural fractures
are important factors in well productivity and must
be considered in rock classification for completion
decisions.

REFERENCES CITED

Figure 15. Cumulative 90-day hydrocarbon and water pro- Aderibigbe, A., C. Chen Valdes, and Z. Heidari, 2016, Inte-
duction in wells 1 and 2. Green and blue bars represent grated rock classification in the Wolfcamp Shale based on
hydrocarbon and water production, respectively. reservoir quality and anisotropic stress profile estimated from
well logs: Interpretation, v. 4, no. 2, p. SF1–SF18, doi:
10.1190/INT-2015-0138.1.
CONCLUSIONS Alfred, D., and L. Vernik, 2013, A new petrophysical model
for organic shales: Petrophysics, v. 54, no. 3, p. 240–247.
Results demonstrated that a reliable integrated rock Archie, G. E., 1952, Classification of carbonate reservoir rocks
classification can significantly improve unconventional and petrophysical considerations: AAPG Bulletin, v. 36,
reservoir productivity forecast. Recommendation of no. 2, p. 278–298, doi:10.1306/3D9343F7-16B1-11D7-
8645000102C1865D.
target intervals for horizontal well placement and Bardon, C., and B. Pied, 1969, Formation water saturation in
multistage hydraulic fracturing based on an integrated shaly sands: Society of Petrophysicists and Well Log
analysis of rock classification results was then pre- Analysts 10th Annual Logging Symposium, Houston,
sented. The correlation between PRCs and the Texas, May 25–28, 1969, SPWLA-1969-Z, 19 p.
Berryman, J. G., 1995, Mixture theories for rock properties,
identified lithofacies showed distinct petrophysical
in T. J. Ahrens, ed., Rock physics and phase relations: A
properties in each geological rock class, and this result handbook of physical constants: Washington, DC, American
was used as a validation for the reliability of the Geophysical Union, v. 3, p. 205–228, doi:10.1029/RF003.
petrophysical classification in well 1. Additionally, Bhattacharya, S., T. R. Carr, and M. Pal, 2016, Comparison of
high-resolution geochemical measurements on the supervised and unsupervised approaches for mudstone
lithofacies classification: Case studies from the Bakken
core were used to build accurate mineral models, and Mahantago-Marcellus Shale, USA: Journal of Natural
differentiate clay types, and provide insights of the Gas Science and Engineering, v. 33, p. 1119–1133, doi:
chemical conditions of the depositional environment 10.1016/j.jngse.2016.04.055.
related to generation and preservation of organic Brumsack, H. J., 2006, The trace metal content of recent
matter. Results showed that the concentration of organic carbon-rich sediments: Implications for Creta-
ceous black shale formation: Palaeogeography, Palae-
certain elements such as Ni and V in the LEF cor- oclimatology, Palaeoecology, v. 232, no. 2–4, p. 344–361,
related directly with in situ hydrocarbon content doi:10.1016/j.palaeo.2005.05.011.
obtained from Rock-Eval measurements. Fur- Canfield, D. E., and B. Thamdrup, 2009, Towards a consis-
thermore, it was observed that reservoir quality is tent classification scheme for geochemical environ-
ments, or, why we wish the term ‘suboxic’ would go
significantly reduced in the bottom interval of LEF,
away: Geobiology, v. 7, no. 4, p. 385–392, doi:10.1111
where kaolinitic clays are present. /j.1472-4669.2009.00214.x.
The comparison between well 1 and well 2, which Dawson, W. C., 2000, Shale microfacies: Eagle Ford Group
are petrophysically similar, proves the importance of (Cenomanian-Turonian) north-central Texas outcrops
geomechanical properties in the assessment of well and subsurface equivalents: GCAGS Transactions, v. 50,
p. 607–622.
productivity. Lithological differences between these Donovan, A. D., 2016, Making outcrops relevant for an un-
wells are possibly causing the observed variation in ge- conventional source rock play: An example from the
omechanical properties. Well 1, with higher proportions Eagle Ford Group of Texas, in M. Bowman, H. R. Smyth,

AMIN ET AL. 1379


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
S. R. Passey, J. P. P. Hirst, and C. J. Jordan, eds., The Lewan, M. D., 1984, Factors controlling the proportionality
value of outcrop studies in reducing subsurface uncer- of vanadium to nickel in crude oils: Geochimica et
tainty and risk in hydrocarbon exploration and produc- Cosmochimica Acta, v. 48, no. 11, p. 2231–2238, doi:
tion: Geological Society, London, Special Publications 10.1016/0016-7037(84)90219-9.
2016, v. 436, doi:10.1144/SP436.12. Lock, B. E., L. Peschier, and N. Whitcomb, 2010, The Eagle
Donovan, A. D., and T. S. Staerker, 2010, Sequence stra- Ford (Boquillas Formation) of Val Verde County,
tigraphy of the Eagle Ford (Boquillas) Formation in the Texas—A window on the South Texas Play: GCAGS
subsurface of South Texas and outcrops of West Texas: Transactions, v. 60, p. 419–434.
GCAGS Transactions, v. 60, p. 861–899. Lowery, C. M., M. J. Corbett, R. M. Leckie, D. Watkins,
Donovan, A. D., T. S. Staerker, R. Gardner, M. C. Pope, A. M. Romero, and A. Pramudito, 2014, Foraminiferal
A. Pramudito, and M. Wehner, 2016, Findings from the and nannofossil paleoecology and paleoceanography of
Eagle Ford outcrops of West Texas and implications to the Cenomanian–Turonian Eagle Ford Shale of southern
the subsurface of South Texas, in J. A. Breyer, ed., The Texas: Palaeogeography, Palaeoclimatology, Palaeoecology,
Eagle Ford Shale: A renaissance in U.S. oil produc- v. 413, p. 49–65, doi:10.1016/j.palaeo.2014.07.025.
tion: AAPG Memoir 110, p. 301–336, doi:10.1306 Mavko, G., T. Mukerji, and J. Dvorkin, 2009, The rock
/13541967M1101083. physics handbook: Tools for seismic analysis of porous
Donovan, A. D., T. S. Staerker, A. Pramudito, W. Li, media: New York, Cambridge University Press, 511 p.
M. J. Corbett, C. M. Lowery, A. M. Romero, and Petriello, J., S. Marino, R. Suarez-Rivera, D. A. Handwerger,
R. D. Gardner, 2012, The Eagle Ford outcrops of S. Herring, W. Woodruff, and K. Stevens, 2013, Inte-
west Texas: A laboratory for understanding hetero- gration of quantitative rock classification with core-based
geneities within unconventional mudstone reservoirs: geologic studies: Improved regional-scale modeling and
GCAGS Journal, v. 1, p. 162–185. efficient exploration of tight shale plays: Society of Pe-
Dunham, R. J., 1962, Classification of carbonate rocks ac- troleum Engineers Unconventional Resources Conference
cording to depositional textures, in W. E. Ham, ed., Clas-
and Exhibition-Asia Pacific, Brisbane, Queensland, Aus-
sification of carbonate rocks—A symposium: AAPG
tralia, November 11–13, 2013, SPE-167048-MS, 13 p.
Memoir 1, p. 108–121.
Popielski, A. C., Z. Heidari, and C. Torres-Verdı́n, 2012, Logs
Eaton, B. A., 1969, Fracture gradient prediction and its ap-
in hydrocarbon-bearing shale: Society of Petroleum En-
plication to oilfield operations: Journal of Petroleum
gineers Annual Technical Conference and Exhibition,
Technology, v. 21, no. 10, p. 1353–1360, doi:10.2118
San Antonio, Texas, October 8–10, 2012, SPE-159255-
/2163-PA.
MS, 19 p.
Energy Information Administration, 2014, EIA updates Eagle
Prasad, M., M. Kopycinska, U. Rabe, and W. Arnold, 2002,
Ford maps to provide greater geologic detail, accessed
Measurement of Young’s modulus of clay minerals
January 19, 2021, https://www.eia.gov/todayinenergy
using atomic force acoustic microscopy: Geophysical
/detail.php?id=19651#.
Folk, R. L., 1962, Spectral subdivision of limestone types, Research Letters, v. 29, no. 8, p. 13-1–13-4, doi:10.1029
in W. E. Ham, ed., Classification of carbonate rocks—A /2001GL014054.
symposium: AAPG Memoir 1, p. 62–84. Saneifar, M., A. Aranibar, and Z. Heidari, 2014, Rock
Gardner, R. D., M. C. Pope, M. P. Wehner, and A. D. Donovan, classification in the Haynesville shale based on pet-
2013, Comparative stratigraphy of the Eagle Ford Group rophysical and elastic properties estimated from well
strata in Lozier Canyon and Antonio Creek, Terrell logs: Interpretation, v. 3, no. 1, p. SA65–SA75, doi:
County: GCAGS Journal, v. 2, p. 42–52. 10.1190/INT-2013-0198.1.
Hubbert, M. K., and D. G. Willis, 1972, Mechanics of hy- Schmoker, J. W., 1979, Determination of organic content of
draulic fracturing, in T. D. Cook, ed., Underground waste Appalachian Devonian shales from formation-density
management and environmental implications: AAPG logs: AAPG Bulletin, v. 63, no. 9, p. 1504–1509, doi:
Memoir 18, p. 239–257. 10.1306/2F9185D1-16CE-11D7-8645000102C1865D.
Jagadisan, A., A. Yang, and Z. Heidari, 2017, Experimental Sone, H., and M. D. Zoback, 2013, Mechanical properties
quantification of the impact of thermal maturity on ker- of shale-gas reservoir rocks—Part 1: Static and dynamic
ogen density: Petrophysics, v. 58, no. 6, p. 603–612. elastic properties and anisotropy: Geophysics, v. 78, no. 5,
Jennings, D. S., and J. Antia, 2013, Petrographic character- p. D381–D392, doi:10.1190/geo2013-0050.1.
ization of the Eagle Ford Shale, South Texas: Mineralogy, Stoneburner, R. K., 2017, The Eagle Ford shale field in the
common constituents, and distribution of nanometer- Gulf Coast basin of south Texas, U.S.A.: A “perfect”
scale pore types, in W. Camp, E. Diaz, and B. Wawak, unconventional giant oil field, in R. K. Merrill and
eds., Electron microscopy of shale hydrocarbon reser- C. A. Sternbach, eds., Giant fields of the decade 2000–
voirs: AAPG Memoir 102, p. 101–114, doi:10 2010: AAPG Memoir 113, p. 121–140, doi:10.1306
.1306/13391708M1023586. /13572003M1133682.
Kale, S., C. Rai, and C. Sondergeld, 2010, Rock typing in gas Suarez-Rivera, R., J. Burghardt, E. Edelman, and S. Stanchits,
shales: Society of Petroleum Engineers Annual Technical 2013, Geomechanics considerations for hydraulic frac-
Conference and Exhibition, Florence, Italy, September ture productivity: American Rock Mechanics Association
19–22, 2010, SPE-134539-MS, 20 p. 47th US Rock Mechanics/Geomechanics Symposium,

1380 Integrated Completion-Oriented Rock Classification in the Eagle Ford Formation


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023
San Francisco, California, June 23–26, 2013, ARMA-13- Wehner, M. P., R. D. Gardner, M. C. Pope, and A. D. Donovan,
666, 9 p. 2017, The Eagle Ford Group returns to Big Bend National
Suarez-Rivera, R., B. Conner, J. Kieschnick, and S. Green, Park, Brewster County: GCAGS Journal, v. 6, p. 161–176.
2006, Hydraulic fracturing experiments help under- Wehner, M. P., M. M. Tice, M. C. Pope, R. Gardner,
standing fracture branching on tight gas shales: American A. D. Donovan, and T. S. Staerker, 2015, Anoxic,
Rock Mechanics Association 41st US Symposium on storm dominated inner carbonate ramp deposition of
Rock Mechanics, Golden, Colorado, June 17–21, 2006, Lower Eagle Ford Formation, West Texas: Society of
ARMA/USRMS-06-1130, 7 p. Petroleum Engineers/AAPG/Society of Exploration Geo-
Suarez-Rivera, R., C. Deenadayalu, M. Chertov, R. N. Hartanto, physicists Third Unconventional Resources Technology
P. Gathogo, and R. Kunjir, 2011, Improving horizontal Conference, San Antonio, Texas, July 22–25, 2015,
completions on heterogeneous tight-shales: Society of Pe- URTEC-2154667-MS, 16 p.
troleum Engineers Canadian Unconventional Resources Wilde, P., T. W. Lyons, and M. S. Quinby-Hunt, 2004,
Conference, Calgary, Alberta, Canada, November 15–17, Organic carbon proxies in black shales: Molybdenum:
2011, SPE-146998-MS, 21 p. Chemical Geology, v. 206, no. 3–4, p. 167–176, doi:
Sun, X., T. Zhang, Y. Sun, K. T. Milliken, and D. Sun, 2016, 10.1016/j.chemgeo.2003.12.005.
Wright, A. M., D. Spain, and K. Ratcliffe, 2010, Application
Geochemical evidence of organic matter source input
of inorganic whole rock geochemistry to shale resource
and depositional environments in the lower and upper
plays: Society of Petroleum Engineers Canadian Un-
Eagle Ford Formation, south Texas: Organic Geochem-
conventional Resources and International Petroleum
istry, v. 98, p. 66–81, doi:10.1016/j.orggeochem.2016.05.018.
Conference, Calgary, Alberta, Canada, October 19–21,
Tribovillard, N., T. J. Algeo, T. Lyons, and A. Riboulleau,
2010, SPE-137946-MS, 18 p.
2006, Trace metals as paleoredox and paleoproductivity
Yang, A., G. Firdaus, and Z. Heidari, 2016, Electrical resistivity
proxies: An update: Chemical Geology, v. 232, no. 1–2, and chemical properties of kerogen isolated from organic-rich
p. 12–32, doi:10.1016/j.chemgeo.2006.02.012. mudrocks: Geophysics, v. 81, no. 6, p. D643–D655, doi:
Vanorio, T., M. Prasad, and A. Nur, 2003, Elastic proper- 10.1190/geo2016-0071.1.
ties of dry clay mineral aggregates, suspensions and Zhou, J., M. Chen, Y. Jin, and G. Q. Zhang, 2008, Analysis of
sandstones: Geophysical Journal International, v. 155, no. 1, fracture propagation behavior and fracture geometry using
p. 319–326, doi:10.1046/j.1365-246X.2003.02046.x. a tri-axial fracturing system in naturally fractured reser-
Wang, G., and T. R. Carr, 2013, Organic-rich Marcellus Shale voirs: International Journal of Rock Mechanics and Min-
lithofacies modeling and distribution pattern analysis in ing Sciences, v. 45, no. 7, p. 1143–1152, doi:10.1016
the Appalachian Basin: AAPG Bulletin, v. 97, no. 12, /j.ijrmms.2008.01.001.
p. 2173–2205, doi:10.1306/05141312135. Zumberge, J., H. Illich, and L. Waite, 2016, Petroleum geo-
Ward, J. H. Jr., 1963, Hierarchical grouping to optimize an chemistry of the Cenomanian–Turonian Eagle Ford oils
objective function: Journal of the American Statistical of south Texas, in J. A. Breyer, ed., The Eagle Ford Shale:
Association, v. 58, no. 301, p. 236–244, doi:10.1080 A renaissance in U.S. oil production: AAPG Memoir 110,
/01621459.1963.10500845. p. 135–165, doi:10.1306/13541960M110449.

AMIN ET AL. 1381


Downloaded from http://pubs.geoscienceworld.org/aapgbull/article-pdf/105/7/1357/5347821/bltn16520.pdf
by Ovidiu Gheorghe Pinca
on 10 February 2023

You might also like