Fault zone processes in mechanically layered mudrock and chalk

Journal of Structural Geology 97 (2017) 118e143

Contents lists available at ScienceDirect

Journal of Structural Geology journal homepage: www.elsevier.com/locate/jsg

Fault zone processes in mechanically layered mudrock and chalk David A. Ferrill a, *, Mark A. Evans b, Ronald N. McGinnis a, Alan P. Morris a, Kevin J. Smart a, Sarah S. Wigginton a, Kirk D.H. Gulliver a, Daniel Lehrmann c, Erich de Zoeten a, Zach Sickmann a Southwest Research Institute®, 6220 Culebra Road, San Antonio, TX 78238, USA Department of Geological Sciences, Central Connecticut State University, 1615 Stanley Street, New Britain, CT 06050, USA c Geoscience Department, Trinity University, One Trinity Place, San Antonio, TX 78212, USA a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 December 2016 Received in revised form 17 February 2017 Accepted 22 February 2017 Available online 24 February 2017

A 1.5 km long natural cliff outcrop of nearly horizontal Eagle Ford Formation in south Texas exposes northwest and southeast dipping normal faults with displacements of 0.01e7 m cutting mudrock, chalk, limestone, and volcanic ash. These faults provide analogs for both natural and hydraulically-induced deformation in the productive Eagle Ford Formation e a major unconventional oil and gas reservoir in south Texas, U.S.A. e and other mechanically layered hydrocarbon reservoirs. Fault dips are steep to vertical through chalk and limestone beds, and moderate through mudrock and clay-rich ash, resulting in refracted fault profiles. Steeply dipping fault segments contain rhombohedral calcite veins that cross the fault zone obliquely, parallel to shear segments in mudrock. The vertical dimensions of the calcite veins correspond to the thickness of offset competent beds with which they are contiguous, and the slip parallel dimension is proportional to fault displacement. Failure surface characteristics, including mixed tensile and shear segments, indicate hybrid failure in chalk and limestone, whereas shear failure predominates in mudrock and ash beds e these changes in failure mode contribute to variation in fault dip. Slip on the shear segments caused dilation of the steeper hybrid segments. Tabular sheets of calcite grew by repeated fault slip, dilation, and cementation. Fluid inclusion and stable isotope geochemistry analyses of fault zone cements indicate episodic reactivation at 1.4e4.2 km depths. The results of these analyses document a dramatic bed-scale lithologic control on fault zone architecture that is directly relevant to the development of porosity and permeability anisotropy along faults. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Normal fault Fault zone Dilation Eagle Ford Formation Calcite Fault refraction Failure mode Hybrid failure Mechanical stratigraphy

1. Introduction Understanding natural deformation processes in fine-grained sedimentary strata has become increasingly important with the growing development of shale and self-sourced reservoirs for oil and gas production. The Cenomanian to Turonian Eagle Ford Formation (deposited between about 94 and 88 Ma) in south Texas (USA) has long been recognized as oil and gas source rock, and is a major self-sourced (unconventional) oil and gas “shale” reservoir (Fig. 1; Haymond, 1991; Robinson, 1997; Martin et al., 2011; Bodziak et al., 2014). The Eagle Ford Formation also is an important aquitard that separates the overlying Austin Chalk from the underlying Buda and Edwards Limestones, and is considered part of the upper confining unit for the Edwards Aquifer system that provides water

* Corresponding author. E-mail address: [email protected] (D.A. Ferrill). http://dx.doi.org/10.1016/j.jsg.2017.02.013 0191-8141/© 2017 Elsevier Ltd. All rights reserved.

for much of south central Texas (Livingston et al., 1936; Maclay and Small, 1983; Maclay, 1989; Ferrill et al., 2004). Although often referred to as shale, the Eagle Ford Formation throughout the subsurface oil and gas play of south Texas is a heterolithic carbonate-rich unit composed of thicker mudrock with relatively thin interbeds of chalk, limestone, and volcanic ash. At the regional scale, the productive Eagle Ford Formation through much of the play is essentially a gentle homocline dipping south or southeast, and productive well depths vary from as shallow as approximately 1219 m, dominated by liquid production, to 4267 m, dominated by dry gas production (U. S. Energy Information Administration, 2014). Despite the apparent simplicity of this play, the geologic setting in the productive Eagle Ford trend includes numerous normal faults that resulted from Cretaceous and Tertiary regional extension around the margin of the Gulf of Mexico basin (Treadgold et al., 2010; Ferrill et al., 2014b; McGinnis et al., 2016). Faulting in the Eagle Ford trend includes the Balcones fault system, which is at the up-dip limit of the Gulf Coast extensional system and is coincident

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

119

Fig. 1. Structural and tectonic synthesis map with inset showing details of study area, with red dot indication location of the Sycamore Bluffs exposure. Inset on the bottom right is the regional paleogeography of North America showing flooded continental margin and Western Interior Seaway at the time of Eagle Ford Formation deposition indicated by light blue color (modified from Blakey, 2011). Boquillas Formation is the lateral equivalent of the Eagle Ford Formation.

with much of the Eagle Ford Formation outcrop belt (Ferrill et al., 2014a). The Eagle Ford Formation is a low-permeability sourcerock reservoir within which induced hydraulic fracturing is used to generate and enhance fractures through which oil and gas can be produced (Donovan and Staerker, 2010; Hentz and Ruppel, 2010; Basu et al., 2012; Bodziak et al., 2014; Busetti et al., 2014; Gale et al., 2014; Smart et al., 2014). Consequently, natural fractures can have a significant influence on oil and gas production within the Eagle Ford Formation (Treadgold et al., 2010; Basu et al., 2012). Induced hydraulic fracturing is the result of the interaction between a fluid driven pressure increase in the reservoir and the in situ stress field, mechanical properties of the strata (mechanical stratigraphy), and pre-existing natural deformation (Ferrill et al., 2014a). Mechanical stratigraphy represents the mechanical properties of the rock, thickness of the mechanical layers, and frictional properties of the boundaries between mechanical layers within a stratigraphic section (Groshong, 2006; Ferrill and Morris, 2008; Ferrill et al., 2017). Variations in mineralogy, depositional texture, porosity, and degree of cementation influence the rock strength (tensile and compressive strength, friction angle, cohesion) and stiffness (Young's modulus) which in turn govern how a rock layer or multilayer responds to natural tectonic deformation (e.g., tensile versus hybrid versus shear; e.g., Hancock, 1985; Reches and Lockner, 1994; Sibson, 1998, 2000; Ramsey and Chester, 2004;

Engelder et al., 2009; Ferrill et al., 2012a, 2014b; Petrie et al., 2014; McGinnis et al., 2015; Giorgetti et al., 2016) as well as induced hydraulic fracturing (Gale et al., 2014; Smart et al., 2014). Mechanical stratigraphy has been found to influence fault nucleation and growth (Ferrill and Morris, 2003, 2008; Roche et al., 2012a, 2012b, 2013; Kettermann and Urai, 2015; Giorgetti et al., 2016; Kettermann et al., 2016), fault zone processes (Jamison, 1979; Young, 1982), and fault geometry and network characteristics (Ferrill and Morris, 2003; Morris et al., 2009b), and is considered key for correctly interpreting geologic structure (McGinnis et al., 2016). Analysis of microseismic data (Busetti et al., 2014) and geomechanical modeling (Smart et al., 2014) of induced hydraulic fracturing indicates that a substantial amount, and in some cases the bulk, of induced failure is in shear (faulting) or hybrid mode rather than tensile (extension fracture) mode, and also indicates that pre-existing faults may reactivate during induced hydraulic fracturing (Smart et al., 2014). In this paper, we analyze normal faults in heterolithic mudrock, chalk, limestone, and volcanic ash in the Eagle Ford Formation in south Texas within a 1.5 km long natural cliff exposure. These faults provide analogs for both natural and hydraulically-induced deformation in the productive Eagle Ford Formation e a major unconventional oil and gas reservoir in south Texas, U.S.A. e and other mechanically layered hydrocarbon reservoirs. The faults are

120

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

southeast and northwest dipping and are consistent with a relatively simple deformation history. Detailed study of three faults with displacements between 0.01 m and 7 m e cutting the same heterolithic stratigraphic section within 100 m of each other e provides the opportunity to evaluate failure characteristics and understand the progressive early development of fault zones in mechanically layered fine-grained strata. The results of these analyses document a dramatic bed-scale lithologic and mechanical stratigraphic control on fault zone architecture that is directly relevant to the development of porosity and permeability anisotropy along faults. 2. Geologic setting 2.1. Stratigraphy The Late Cretaceous Buda Limestone, Eagle Ford Formation, and Austin Chalk were deposited sequentially in the transition zone between the Gulf of Mexico basin and the southern part of the western interior seaway (Fig. 1 bottom right inset; Phelps et al., 2013). These three formations were deposited during a major turning point in deposition marking the end of long-lived shallowmarine, rimmed carbonate platforms including the Sligo and Edwards (also referred to as Stuart City; Phelps et al., 2013 and references therein) reef systems in south Texas (Fig. 1) that dominated Early Cretaceous sedimentation of the northern Gulf of Mexico and represent the beginning of deep-marine siliciclastic and shelf chalk deposition that dominated the Gulf of Mexico and Western Interior Seaway during the Late Cretaceous (Goldhammer and Johnson, 2001; Scott, 2010; Blakey, 2011). At the regional scale, the Eagle Ford and Austin Chalk Formations can be viewed as a thick succession of pelagic sediments dominated by mudrock, and planktonic foraminifer and nannoplankton chalk that blanketed the entire northern and northwestern shelf of the Gulf of Mexico during the maximum flooding of the first-order Cretaceous sea level cycle (Hovorka and Nance, 1994; Goldhammer and Johnson, 2001; Donovan and Staerker, 2010). The Eagle Ford Formation has been described as both an organic-rich shale and as a calcareous mudrock. Detailed descriptions, however, reveal it to be a complex assemblage of lithofacies including calcareous mudrock, chalk, limestone, and volcanic ash (Lock and Peschier, 2006; Hentz and Ruppel, 2010; bourg et al., 2016). Climate Scott, 2010; Lock et al., 2010; Fre related cyclic alternation between chalk and mudrock beds is common through the middle Eagle Ford Formation (Eldrett et al., 2015). These bed-scale compositional and textural variations in the Eagle Ford produced contrasting mechanical layers that in outcrop are observed to fundamentally influence joint and fault development in the Eagle Ford Formation (Ferrill et al., 2014a; McGinnis et al., 2017). 2.2. Structural and tectonic setting The structural and tectonic history of the Eagle Ford Formation in south Texas includes both extensional and contractional deformation (Fig. 1; Ferrill et al., 2014b). The Gulf of Mexico coastal plain has experienced protracted Jurassic through present extensional deformation. Some extensional faults root downward and flatten (sole) into Jurassic salt and lose displacement upward into the Cretaceous section (Treadgold et al., 2010; Ferrill et al., 2016a; McGinnis et al., 2016). Stratigraphic growth, antithetic dip of layering, and resulting wedge shaped stratigraphic packages associated with these salt-rooted faults suggest synsedimentary growth faulting within the Cretaceous section. This salt-detached synsedimentary faulting, which represents the earliest regional

deformation of the Eagle Ford Formation, was restricted to the area initially underlain by the Jurassic salt, including areas along and basinward (southeastward) from the “up-dip salt limit” illustrated in Fig. 1 (Ewing, 2003, 2010), and has not been recognized in the outcrop belt of the Eagle Ford. The long-lived San Marcos Arch (Halbouty, 1966; Young, 1986; Salvador, 1987; Laubach and Jackson, 1990; Ewing, 1991, 2009; Hentz and Ruppel, 2010; Dooley et al., 2013; Hudec et al., 2013; Phelps et al., 2013; Bodziak et al., 2014) appears to have influenced distribution of salt, based on the wrapping of the up-dip limit of salt around the arch, which may reflect the pattern of Jurassic salt deposition. Furthermore, thinning and changes in lithofacies within the Eagle Ford Formation (Bodziak et al., 2014) on and across the arch indicate that it was a positive bathymetric feature during the Cretaceous. Other large-scale extensional faults are interpreted to have formed within the Cretaceous stratigraphic sequence to accommodate Tertiary extension, nucleating in brittle, competent layers and causing fault tip folding in less competent formations (Ferrill and Morris, 2008; Morris et al., 2009a, 2009b; Ferrill et al., 2012b). These faults generally do not connect downward to salt, and many are strata-bound and do not penetrate upward to the surface, instead vertically terminating (“tipping”) within the stratigraphic section (e.g., Treadgold et al., 2010; Ferrill et al., 2016a; McGinnis et al., 2016). The Balcones fault system (Fig. 1) e an extensional fault system along the Balcones escarpment that defines the margin between the Edwards Plateau and the Gulf Coastal Plains province e marks the surface expression of the up-dip limit of normal faulting related to the Tertiary extension along the northwestern margin of the Gulf of Mexico Basin (Collins and Hovorka, 1997; Ferrill et al., 2004, 2011; Ferrill and Morris, 2008). The Balcones fault system extends from Del Rio, Texas in the west, to near Dallas, Texas in the north, changing trend by approximately 70
through the central portion between the cities of Austin, San Antonio, and Uvalde (Foley, 1926; Weeks, 1945; Collins and Hovorka, 1997). The fault system in central and western portions of Texas accommodates southeast directed extension on primarily southeast-dipping normal faults, although antithetic (northwest-dipping) normal faults also are present (Collins and Hovorka, 1997; Ferrill et al., 2004). The Sycamore Bluffs exposure is positioned at the western end of the Balcones fault system, and we interpret the faulting in the exposure to be related to this extensional deformation. Laramide-age (~80e40 Ma) shortening deformed the Cretaceous section in south and west Texas, producing contractional folds, thrust faults, intermontane basins, tectonic stylolites, joints, and veins (e.g., Fowler, 1956; Maxwell et al., 1967; Moustafa, 1988; Erdlac, 1990, 1994; Lehman, 1991; Turner et al., 2011; Ferrill et al., 2016b). The Laramide structural front (Fig. 1) follows the northwest-southeast trending Santiago Mountains in the Big Bend region of west Texas (Turner et al., 2011), steps eastward to include the Devils River uplift (Webster, 1980), and in south Texas includes the northwest-southeast trending Zavala syncline and Chittim anticline adjacent to the Rio Grande (Fowler, 1956; Rose, 1984; Scott, 2010; Boon and Bacon, 2014). The Chittim anticline is located above a Triassic-aged graben, or half-graben, filled by Jurassic and Cretaceous sedimentary strata and is considered to be an inversion structure (Scott, 2010). In addition to these macrostructures, it is likely that Laramide contractional meso- and microscale faulting and folding has variably influenced the Eagle Ford Formation and has reactivated (e.g., continued propagation or inversion) these pre-existing structures across south Texas to the San Marcos Arch, as discussed by Ferrill et al. (2014a). The Sycamore Bluffs exposure (Fig. 2) is located (i) approximately 35 km northwest and up-plunge along trend of the Chittim anticline (Fig. 1), (ii) along the southern margin of the Devils River

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

121

Fig. 2. (a) Outcrop photograph of Sycamore Bluffs exposure. Annotation shows the contact between the Eagle Ford Formation and the overlying Austin Chalk, and numbered metering positions indicate stratigraphic heights of marked beds in measured stratigraphic section shown in Fig. 3. View is to the NW. (b) Section of Sycamore Bluffs photomosaic showing locations of Textbook, Big Indigo, and Spanish Goat faults. The white line represents the horizon at stratigraphic height of 5.6 m. Cliff is approximately 30 m high in center of photograph.

uplift, and (iii) near the westward termination of the Balcones fault system. The Eagle Ford strata are horizontal to gently southeast dipping at the study location, similar to the gentle southeast dipping homocline of the region of Eagle Ford production, making this a relevant analog for gaining insights into natural deformation features encountered in wells away from macroscale structures.

3. Methodology 3.1. Lithostratigraphy and mechanical stratigraphic characterization The lithostratigraphic character of the exposure was measured, described, and sampled in the field using standard field tools

122

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

including Jacobs staff, measuring tape, and hand lens. A companion mechanical rebound data set was collected using an N-type Schmidt hammer (Aydin and Basu, 2005) to measure the presentday in situ elastic rebound (hereafter “rebound”) properties of rock and characterize the relative competence (or stiffness) of the layers that comprise the mechanical stratigraphy (Ferrill and Morris, 2008). For this rebound analysis, we followed the approach described by Morris et al. (2009b) and Ferrill et al. (2011,

2012a, 2012b). As described in those studies (i) measurements were made on subvertical rock surfaces to eliminate the need to correct the readings for variations induced by gravity, (ii) each sample consisted of at least 10 rebound measurements within an area of approximately 25 cm2, and (iii) we present the results in terms of rebound value R, which has been correlated to unconfined compressive strength and Young's modulus through laboratory testing (e.g., Katz et al., 2000; Aydin and Basu, 2005). We use the

Fig. 3. Lithostratigraphic and rebound profile for Eagle Ford Formation and Austin Chalk at Sycamore Bluffs.

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

term “competence” to express the ability of a rock to resist inelastic (permanent) deformation (Ferrill and Morris, 2008). Rock properties are expected to have been modified somewhat because of unloading, continued diagenesis, and weathering. Notwithstanding the limitations of Schmidt hammer rebound values (see for example, Morris et al., 2009b), we interpret the R-value profiles as reflecting some aspects of relative competence differences between measured rock layers, although we note that weathering commonly reduces strength of clay-rich beds more than carbonates. 3.2. Fault characterization An initial survey was made to document faults throughout the length of the exposure with fault locations (UTM coordinates) measured using a hand-held GPS device. Fault characteristics were measured in the field using Brunton compass and measuring tape. Strike and dip of failure surfaces were measured for fault segments through different lithologic beds. Slickenlines and bedding offsets were used to measure rake and displacement parallel to slip directions. This information was collected for 13 normal faults and 1 thrust fault to provide constraints on fault orientation and displacement and analysis of failure angle with respect to host lithology. The normal faults generally dip northwest or southeast consistent with northwest-southeast directed extension, which we interpret to be related to Gulf of Mexico coastal plain extension consistent with a lateral continuation of the extension represented by the Balcones fault zone. The single thrust fault dips southsouthwest, has 62 cm dip slip displacement, and accommodates north-northeast directed shortening, consistent with fold orientations west of the study area in Mexico (Fig. 1), which we interpret to have been produced by Laramide contractional deformation in the region. Three extensional faults were selected for detailed analysis presented in this paper to elucidate the evolution of normal fault zones that reflect displacements of several cm (Textbook fault), to several 10s of cm (Spanish Goat fault), and finally several meters (Big Indigo fault). For these faults, a total of 141 measurements were made, representing both the hanging wall and footwall, to trace the faults through multiple lithologic layers and document stages of fault zone development. In the case of the Big Indigo fault, irregularity of the outcrop surface makes precise measurements of fault offsets impractical, and the upper half of the exposure is a vertical cliff that is not safely accessible for detailed orientation and displacement measurements. Therefore, for the main trace of the Big Indigo fault we used a spatial scanning system (Trimble VX™ Spatial Station) to survey the fault trace and measure the three dimensional positions of offset marker beds. These offset marker positions were used in conjunction with measured fault and slickenline orientations to determine displacements along the main trace of the fault, including throw and heave components. 3.3. Fluid inclusion and stable isotope geochemistry analysis 3.3.1. Fluid inclusion microthermometry Fluid inclusion microthermometry is a well-established approach to constraining the temperature of vein formation, and with additional assumptions forms the basis for interpreting depth at the time of vein formation (e.g., Evans et al., 2012). To constrain the fluid history of calcite cement in normal fault zones, fluid inclusion and stable isotope geochemistry analyses were performed on fault-related vein samples. In addition to the samples of veins, companion host-rock samples were collected at each site from a distance of at least 1 m from the vein sample. For petrographic analysis and for fluid inclusion microthermometry,

123

doublyepolished thick sections were made of vein mineral samples (see Evans and Battles, 1999 for techniques used). Prior to collecting microthermometric data, prepared polished fluid inclusion plates were scanned for usable inclusions and these were mapped as fluid inclusion assemblages (Goldstein and Reynolds, 1994) in a manner similar to that described by Touret (2001). Although we attempted to characterize the assemblages as to their origin (primary, pseudosecondary, secondary), because of the high density of inclusions and the general lack of growth features, most inclusions are classified as undifferentiated. Where possible, fluid inclusion microthermometry was done using two-phase inclusions. Conventional heating and freezing microthermometric analyses were performed using a modified U.S. Geological Survey-type heating-freezing stage manufactured by FLUID Inc. and a Linkham stage. The stages were calibrated at 0
C (ice bath), 374.1
C (critical point of water), and 56.6
C (CO2 triple point; the latter two standards supplied by SYNFLINC, Inc.). Isochores for H2O-NaCl inclusions were determined using the computer program FLUIDS (Bakker, 2003; Bakker and Brown, 2003). Salinities for the H2O-NaCl inclusions were calculated using the equation presented by Bodnar (1993). In addition, ultraviolet (UV) fluorescence photomicroscopy was performed on all of the samples in order to assess the presence of hydrocarbons and approximate American Petroleum Institute (API) gravity of hydrocarbon-bearing inclusions (e.g., Bourdet and Eadington, 2012). The polished vein thick sections were examined under UV illumination provided by a high-pressure mercury lamp and a Lecia™ A filter cube that is characterized by a UV excitation band of 340e380 nm. A 420 nm epifluorescence barrier filter allows only the longer wavelength fluorescence to reach the observer. In order to objectively measure the fluorescence colors of the hydrocarbon inclusions, JPEG images were captured and their spectral responses were determined by the process described in Evans et al. (2014). The color of the fluorescence reflects the API gravity of the trapped hydrocarbon fluids, with lower gravity fluids fluorescing yellows and oranges, and higher gravity fluids fluorescing blues and whites. 3.3.2. Stable isotope analysis Oxygen and carbon stable isotope analyses were performed on selected samples of calcite vein filling to assess whether calcite cements were sourced from local versus externally derived aqueous fluids. In veins with multiple stages of mineralization, each stage of calcite was sampled separately where possible. In cases where multiple stages of calcite were present and physical separation of the stages was not possible, the calcite samples were run as bulk samples for the vein. Oxygen and carbon isotope composition of vein calcite (33 samples) and host rock (6 samples) were measured at the Yale University Earth System Center for Stable Isotopic Studies. 3.4. Resolved stress analysis Analysis of resolved stresses acting on fault and fracture surfaces can be used to assess the occurrence or likelihood for slip or dilation in crustal stress fields independent of information for friction or cohesion. Slip tendency, Ts, is the ratio of resolved shear (t) and normal stress (sn) acting on a surface (Morris et al., 1996):

Ts ¼ ðt=sn Þ

(1)

For faults and fractures, slip is likely to occur on a surface when the resolved shear stress on that surface equals or exceeds the frictional resistance to sliding. Whether or not a surface experiences slip depends upon its cohesive strength, if any, and the coefficient of static friction (m). The coefficient of static friction is the

124

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

value of Ts at which slip occurs on a cohesionless surface and is often referred to as the fault “strength” in earthquake focal mechanism analysis (e.g., Harmsen, 1994). According to Byerlee (1978), the coefficient of static friction for most crustal conditions is 0.6e0.85, so faults with Ts  0.6 have slip tendency that equals or exceeds the commonly assumed coefficient of friction of 0.6 and therefore can be considered poised for slip or critically stressed (e.g., Zoback, 2007). Data from laboratory experiments involving phyllosilicates (e.g., Morrow et al., 1992; Moore and Lockner, 2004; Ikari et al., 2007; Haines et al., 2013) show that coefficient of static friction can be much lower (also see recent compilation by Ferrill et al., 2017), and values of <0.2 have been determined for some clay-rich fault rocks. Dilation of faults or fractures is largely controlled by the resolved normal stress, which is a function of lithostatic and

tectonic stresses and fluid pressure. Dilation tendency, Td, is defined as follows (Ferrill et al., 1999):

Td ¼ ðs1  sn Þ=ðs1  s3 Þ

(2)

where s1 is the maximum principal stress magnitude and s3 is the minimum principal stress magnitude. The normal stress experienced by a fracture depends on the magnitude and direction of the principal stresses relative to the fracture plane. The ability of a fault or fracture to dilate and transmit fluid is directly related to its aperture, which in turn is a function of the effective normal stress acting upon it. Dilation tendency can be computed for all fault surface orientations within a known or hypothesized stress field to assess relative likelihood of dilation and associated fluid flow and related potential for mineralization or dissolution.

Fig. 4. Interpreted field photograph of Textbook fault (a) showing measurement locations along the footwall and hanging wall portions of the fault, annotation of shear (white) and dilational (red) fault segments, and a displacement profile (e). Close-up photos show early stages of calcite growth along dilational portions of the fault (b, c, and d). Calcite sample locations are marked by black boxes and corresponding sample letters e B is a host sample, and A (in subset c), is a vein sample e with prefix SN1152-004 (see Table 1). Orange multitool shown in b, c, and d is 8.1 cm long.

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

4. Lithostratigraphy and mechanical stratigraphy The measured section for the Sycamore Bluffs exposure starts at the lowest accessible stratigraphic level in the exposure (0.0 m in measured section) and includes 31.35 m of Eagle Ford Formation, and 2.0 m of overlying Austin Chalk (Fig. 3). The lower part of the exposed section, from the base of the outcrop to a stratigraphic height of 15 m, includes the faults characterized in detail in this study and exhibits the highly cyclic alternation between organicrich calcareous mudrock and chalk beds typical of the hydrocarbon productive portion of the Eagle Ford Formation. The basal 15 m of the Sycamore Bluff outcrop section corresponds to the interval between 58.35 and 73.35 m depth in the nearby Iona Core described by Eldrett et al. (2014, 2015). Other than two 1- to 1.5 m-thick slumped intervals that are thickened by soft-sediment deformation structures, the succession in the lower exposures at Sycamore Bluff is almost entirely composed of a cyclic alternation between laminated calcareous mudrock containing abundant nannoplankton and planktonic foraminifers and massive to laminated calcareous micropackstone (chalk) (Fig. 3). A few beds in the field were classified as marl, and are essentially a facies intermediate in terrigenous clay content between the mudrock and chalk. Mudrock intervals range from 0.04 to 0.8 m thick with an average thickness of 0.19 m and comprise 53% of the vertical section thickness. Calcareous micropackstone (chalk) beds range from 0.05 to 0.8 m thick and

125

comprise 47% of the vertical section thickness (Fig. 3). The lower and upper contacts of chalk beds are gradational with adjacent mudrock intervals (Fig. 3). The chalk may consist of laterally continuous beds or may be highly nodular and discontinuous. In thin section, the calcareous mudrock and the chalk are texturally and compositionally very similar. Both are dominated by planktonic foraminifers and calcareous nannoplankton but the mudrock contains greater terrigenous clay content. The upper 10 m of the Eagle Ford Formation contains a greater proportion of limestone (Fig. 3). The gradational contacts between beds, subtle variation in clay and carbonate content, prevalence of horizontal lamination, and lack of evidence for traction currents all point to an extremely low energy mode of sedimentation as “pelagic rain” below the storm wave base. Sedimentary structures and grainstone texture in the limestone in the upper Eagle Ford Formation indicates winnowing by bottom currents. The rebound profile shows distinctive mechanical layering throughout the section (Fig. 3). Average rebound (R) values measured in the measured section range from 5 to 43. Volcanic ash beds have the lowest average rebound values consistently of 5, mudrock beds generally have average rebounds of 5e15, chalk beds generally have average rebounds of 15e40, and limestone beds generally have average rebounds of 15e50, including the highest bed-average rebounds of 50 and 53.

Fig. 5. Interpreted field photograph of Spanish Goat fault (a), showing measurement locations along the footwall and hanging wall portions of the fault, annotation of shear (white) and dilational (red) fault segments, and a displacement profile. Notice that overlap of the upper (brown) and lower (green) segments of the fault define a breached relay ramp e note fault connection at branch point. Close-up photos show slickensided surfaces on exposed calcite sheets (b) and calcite sheet growth beginning to overlap along the lower segment (c). Calcite sample locations are marked by black boxes and corresponding sample letters e C is a host sample, and A, B, and D are veins samples e with prefix SN1152-007 (see Table 1). Yellow field book shown in (a) is 12.2 cm wide and 19.1 cm tall. Orange multitool in (c) is 8.1 cm long.

126

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

5. Fault characterization 5.1. Faults in Sycamore Bluffs exposure Faults in the exposure dip to the SE in the regional down-to-the Gulf of Mexico direction, as well as to the NW antithetic to the regional bedding-dip direction, and collectively accommodate SE directed extension. The general orientation of layering at the NW end of the exposure is horizontal, and layering exposed in the SE half of the exposure dips very gently (<5
) to the SE, however, fault related folding has produced local layer dips (synthetic to the associated fault) as steep as 23
. A bed-parallel survey of a bed at 5.6 m stratigraphic height across the entire length of the exposure revealed a total of 13 normal faults and 1 reverse fault (all within a distance of 850 m) with maximum displacements ranging from <1 cm to 7 m.

In this analysis, three faults were selected for detailed analysis that are within a lateral distance of 100 m (Fig. 2b) and represent displacements on the scale of centimeters (Textbook fault), 10's of centimeters (Spanish Goat fault), and meters (Big Indigo fault). These faults were all accessible, and each cut a portion of the Eagle Ford from 5.6 to 7.1 m in the measured stratigraphic section so that differences in structural style could be evaluated in the context of fault displacement, within a consistent lithologic package and large-scale structural context. 5.2. Textbook fault The Textbook fault in the exposure cuts 7.2 m of section, from a lower tip at stratigraphic height of 1.1 m to an upper tip at stratigraphic height of 8.3 m (Fig. 4a). The fault strike range is 179e230
, with WNW dips of 25e90
. The fault profile is completely exposed

Fig. 6. (a) Outcrop photograph of the Big Indigo fault from Sycamore Creek (bottom) to top of cliff where Tertiary gravel rests on angular unconformity that truncates fault and Eagle Ford and Austin Chalk beds. Cliff is 30 m tall. Numbered lines correspond to stratigraphic heights in measured stratigraphic section shown in Fig. 3. (b) Upper portion of the fault is filled with smeared clay-rich volcanic ash derived from ash beds at stratigraphic heights of 22 m and 25 m (see Fig. 3). (c) Displacement profile for the fault, including throw and heave components, based on total station surveying and spatial scanning using Trimble VX™ Spatial Station.

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

(i.e., no covered intervals) and offset beds, slickenlines, and width of clear or white calcite fill provide a measureable record of the direction, sense, and magnitude of displacement along the fault from tip-to-tip. Normal-oblique (rake of 108e150
) fault-parallel displacement ranges from 0 to 15 cm. Seventeen dilational segments were identified, represented by calcite-cemented relatively steep fault segments bounded by less steeply dipping shear segments that are narrow and generally barren of calcite cement. Steep dilational fault segments tend to be associated with chalk, limestone, and calcareous mudrock beds, and shear segments generally correspond to more clay-rich mudrock beds. Most of the dilational segments are completely cemented with calcite, with margins adjacent to host beds typically represented by multiple veins separated by wall rock slivers and central portions cemented with blocky calcite. One prominent exception is incompletely cemented, and retains open porosity between euhedral terminations of prismatic calcite crystals that have grown inward from dilational fault zone boundaries (Fig. 4b). Transitions from dilational to shear segments range from smooth to abrupt, generally corresponding to subtle versus distinct lithologic contrasts (Fig. 4c and d). The fault is continuous through the section, changing dip from bed-to-bed, and lacking segments, branches, or splays with blind terminations (tips) aside from the overall top and bottom terminations. The displacement profile shows the points of zero displacement at the fault tips, and maximum displacement measured in slickenline direction of 15 cm. Along most of the fault, heave is greater than throw due to the plunge of the slip vectors being <45
. The Textbook fault has two distinct displacement maxima at stratigraphic heights of 3.5 m and 5.85 m, and 74% of the fault in the exposure has displacement <10 cm. Vertical changes in displacement tend to span multiple lithological layers, as illustrated in Fig. 4, rather than simply corresponding to lithologic units. Although fault zone character varies dramatically as a function of lithology, the fault displacement e measured in the displacement direction represented by slickenside lineations on the shear segments or opening directions on the hybrid segments e does not show a simple correspondence to lithology. Because the displacement vector does not change dramatically from bed-to-bed, the heave and throw do not show distinctive bed-to-bed changes. However, because the fault failure surface dip does change dramatically from bed-to-bed, the role of dilation is much greater in chalk and limestone beds where fault dips are steeper, compared to the general lack of dilation on shear segments in mudrocks where fault failure surface dips are more gentle and parallel to the displacement vector. 5.3. Spanish Goat fault The Spanish Goat fault was analyzed in a continuous exposure through 4.3 m of section (through the stratigraphic height range of 4.6 me8.9 m) (Fig. 5). The fault strike is generally N, with azimuths of 028e326
, and dips east to east-south-east at 36e90
. The fault is continuously exposed (i.e., no covered intervals) in the analyzed area. As with the Textbook fault, offset beds, slickenlines, and width of calcite fill provide a measureable record of the direction, sense, and magnitude of displacement along the fault from tip-to-tip. Two now-linked fault segments are separated by a breached extensional relay structure. Normal dip-slip displacement, along slickenline rakes of 74 to 108
, ranges from 19.5 to 52 cm on the fault system, although displacement on the lower tip of the upper fault segment and upper tip of the lower fault segment diminish to zero at the tips. Displacement, heave, and throw profiles are plotted for the upper and lower segments and summed for the two segments combined (Fig. 5d). The displacement profile illustrates distinct local maxima at stratigraphic heights of 5.1, 5.9, and 6.7 m, each associated with a

127

chalk bed, and a broad maximum from 7.3 to 8.6 m within mudrock and a mudrock-dominated slumped interval. The two vertically overlapping segments include fourteen dilational steps, each associated with a carbonate bed (chalk, limestone, calcareous mudrock). The dilational steps have hosted calcite mineralization, which localized at steeply dipping fault sections that continue upward and downward into narrow, barren, less steeply dipping shear segments. This character is essentially the same as the style observed on the Textbook fault where steep dilational fault segments tend to be associated with chalk and calcareous mudrock beds, and shear segments generally correspond to more clay-rich mudrock beds. The dilational segments have hosted calcite cementation throughout the displacement history of the fault such that the dilational jogs are completely cemented forming calcite sheets that connect corresponding hanging wall cutoffs of offset carbonate with clear or white calcite cement. As with the Textbook fault, transitions from dilational to shear segments range from smooth to abrupt, corresponding to gradual versus distinct lithologic transitions. As noted above, the fault consists of two segments, each of which is for the most part continuous through the section, changing dip from bed-to-bed but lacking segments, branches, or splays with blind terminations. The lower tip of the upper segment, however, includes several overlapping, curved en-echelon segments within mudrock with calcite cement filling steep dilational portions at and near segment intersections. The upper tip of the lower fault segments curves to a near vertical dip in the mudrock at the point of intersection with the upper fault segment, and this curved portion is a shear fracture with no evidence of dilation. 5.4. Big Indigo fault The Big Indigo fault is one of three faults that cut the entire exposed height of the Sycamore Bluffs exposure. Cutting the full exposed thickness of the Eagle Ford Formation (~30 m), the fault juxtaposes the Austin Chalk against the upper Eagle Ford in the upper few meters of the exposure (Fig. 6). The Big Indigo fault has a strike range of 192e228
, dips northwest at 38e81
, and accommodates a maximum of 7.2 m of dominantly normal dip-slip displacement (Fig. 6c), with measured slickenline rakes ranging from 80 to 140
. Throughout most of the height of the exposure, the Big Indigo fault consists of a single main displacement zone. However, the lower 3 m of the exposure contains a main surface (accommodating 5.5 m of displacement), and a linked segment (accommodating 0.5 m of displacement) that is approximately parallel to the main segment in the lower 2 m of the exposure, above which it curves to intersect the main fault zone at a high angle (~70
; Figs. 6a and 7a, d). We interpret this to be a breached relay structure linking two initially vertically separated and noncoplanar fault segments. This northwest (linked) segment of the Big Indigo fault is made up of thin shear segments that transition to steeper dilational segments through competent chalk beds, similar to the style of the Spanish Goat and Textbook faults. Dilational fault jogs are cemented with massive calcite to form localized calcite sheets that appear rhombohedral in cross section and that connect corresponding offset carbonate beds across the fault (Fig. 7d). The main strand of the Big Indigo fault has minimum measured displacement of 5.5 m both at the bottom of the exposure (stratigraphic height 5.5 m) and in the upper Eagle Ford (stratigraphic height 24 m), and a displacement maximum of 7.2 m at stratigraphic height of 15 m (Fig. 7c). In the middle and lower portions of the exposure, the main (southeast) segment of the Big Indigo fault is represented by multiple slickensided surfaces in a thin laminated fault zone 1 to several cm thick (e.g., Fig. 7b), and by thick zones (up to 0.6 m thick) consisting of multiple stacked calcite sheets that are

128

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

individually 1 to >10 cm thick and 10's of cm to >1 m long in the slip direction, and composed of massive prismatic to blocky and coarsely crystalline (1 cm to several cm) white to grey calcite (Fig. 7c and d). These calcite sheets are separated by clay-coated and slickensided slip surfaces, and generally connect to a chalk or limestone bed on either the footwall or hanging wall of the fault. Where the calcite sheets connect to a chalk or limestone bed, the massive calcite vein transitions to the host bed across a laminated zone of calcite veins filled with more finely crystalline (<1 mm) calcite cement and thin carbonate host bed laminae. Synthetic layer dip (e.g., Fig. 7b) also is locally observed adjacent to the fault on either the hanging wall or footwall side of the fault. The surrounding rock mass contains six other smaller (<1 m) displacement faults, and each of these shows fault refraction and dilational jogs associated with carbonate beds that was described earlier for the Textbook and Spanish Goat faults (e.g., Fig. 7e). Five of these minor faults are synthetic to the Big Indigo fault, and one minor fault is antithetic. 5.5. Fault orientation summary and interpretation Stereonet plots of fault poles and slickenlines along with histograms of dip magnitude for the Textbook, Spanish Goat, and Big Indigo faults illustrate the orientation ranges for the three faults (Fig. 8). As discussed above, dip variation is primarily controlled by the lithologic variation in the section, with steep dips associated with limestone or chalk beds, and lower dips generally associated with failure surfaces through mudrock beds. Faults formed as refracted failure surfaces, steeper in more competent carbonate rich beds, and with lower dip angles in the less competent mudrock beds. Shear segments also change dip within relatively consistent lithologic packages associated with relay structures along the Spanish Goat and Big Indigo faults, producing fault segments that curve toward intersections with adjacent overlapping faults. These relay-related fault dip changes include portions that dip more steeply and more gently than typical for these lithologic units. As displacement accumulated on the faults, the slip occurred parallel to the less steep shear surfaces through mudrock, and dilation occurred associated with steep segments. Repeated dilation and calcite cementation of steep segments progressively led to the formation of calcite sheets, separated by thin clay-coated shear surfaces. These shear surfaces lengthened in the displacement direction, and the shear surfaces and the intervening calcite sheets became the dominant fabric elements within the fault zones. The calcite sheets and shear surfaces still connect offset beds across the Textbook and Spanish Goat faults, as well as the NW-linked segment of the Big Indigo fault. In contrast, the calcite sheets are not continuous across most of the Big Indigo fault. The longest calcite sheets along the Big Indigo fault can be traced along the fault slip direction for >1 m. Beyond this displacement, calcite sheets apparently became disconnected across the fault zone as slip progressively localized along a narrow fault core. 6. Fluid history and conditions 6.1. Fluid inclusion microthermometry results Analysis here concentrates on vein calcite samples of the (i)

129

Textbook fault (Fig. 4), (ii) Spanish Goat fault (Fig. 5), and Big Indigo fault and associated smaller faults (Fig. 9) collected from the sample locations indicated in the referenced figures. Vein samples are all from dilational segments along faults, and include a variety of textures representing calcite precipitated after multiple fault slip and dilation events. Varying degrees of calcite twinning e including untwinned calcite and weakly, moderately, and highly twinned calcite (Fig. 10, Table 1) e indicate variable amounts of crystalplastic deformation of the calcite after cementation, during progressive deformation in these fault zones. 6.1.1. Aqueous inclusions Aqueous inclusions were found to be relatively common in veins associated with the Big Indigo fault, but were considerably less common in the Spanish Goat and Textbook faults. In many cases, these inclusions occur in assemblages that do not contain vapor bubbles (Fig. 11a) or in assemblages in which some inclusions have vapor bubbles and others do not (Fig. 11b). Only half of the samples taken had aqueous inclusions that could be used for microthermometry, and all of those were in the Big Indigo fault veins. In addition, some stages of calcite within the fault veins exhibit strong deformation twinning (Table 1) and aqueous inclusions were not present. The two-phase aqueous inclusions that are used for microthermometry are typically <5 to over 20 mm in size and very irregularly shaped, while some inclusions are up to 30 mm (Fig. 11b). The vapor bubble occupies less than 5% of the inclusion volume. Because of the paucity of inclusions with a vapor bubble and the fact that they commonly stretched upon heating, only 5 to 15 repeatable measurements could be made for each sample. Ice melting temperatures (Tm) were reproducible within 0.2
C and homogenization values (Th) within 3
C. Crushing tests in a nonpolar solvent did not show methane evolving from these types of inclusions. Ice melting (eutectic) was extremely difficult to observe due to the small size of the inclusions. Where observed, it is above ~ 20.8
C, indicating NaCl-rich fluids (Roedder, 1984). Upon heating, the two-phase aqueous inclusions homogenized to the liquid phase in two broad groups at (i) 62e76
C and (ii) 80e100
C (Table 1, Fig. 12). The higher temperature group is typically in fluid inclusion planes that may be interpreted as secondary. During cooling runs, the vapor bubbles of many aqueous inclusions completely collapsed during freezing. The bubbles commonly did not reappear until temperatures above 0
C and up to 8
C, resulting in final ice melting in the absence of a vapor phase. This metastable behavior results from low to negative pressures within the inclusion, and means that the final ice melting temperature cannot be used for determining salinity (Roedder, 1984). For those inclusions where ice melted in the presence of a vapor bubble, the Tm values ranged from 2.5 to 0.0
C, indicating salinities of 0.0e4.2 wt % NaCl equivalent, with the salinities greater than 0% in inclusions with Th > 83
C (Table 1, Fig. 12). 6.1.2. Aqueous fluid trapping conditions Estimates of fluid trapping conditions can be made via fluid inclusion microthermometry. Fluid trapping occurred under less than lithostatic fluid pressure conditions based on the absence of bed-parallel veins, and likely near hydrostatic conditions due to the common lack of vapor bubble formation. Similarly, there is no

Fig. 7. Interpreted field photograph (a) of the Big Indigo fault showing measurement locations along the footwall and hanging wall portions of the fault, annotation of shear (white) and dilational (red) fault segments, and multiple reference beds to illustrate displacement along the fault. Yellow field book visible in (a) is 12.2 cm wide and 19.1 cm tall. (b) Closeup image showing beds with synthetic dip that likely record early fault propagation. (c) Fault zone locally contains four overlapping sheets or layers of blocky calcite fill attached to dilational fault segments along the adjacent wall rock. (d) Detail of NW segment with rhombic massive calcite veins in dilational jogs along fault. Rock hammer visible in (c) and (d) is 30.5 cm long. (e) Detail of small displacement refracted fault with complex calcite veins in dilational segments linked to more gently dipping shear segments. Orange multitool is 8.1 cm long. Rock hammer visible in (c) and (d) is 30.5 cm long.

130

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Fig. 8. Lower-hemisphere equal-area stereographic projections of fault measurements (colored great circles and poles indicate fault measurements, and black dots indicate slip directions), and fault dip histograms and for fault orientation measurements from (a) Textbook, Spanish Goat and Big Indigo faults combined, (b) Textbook fault, (c) Spanish Goat fault, and (d) Big Indigo faults.

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

131

Fig. 9. (a) Sample locations within and associated with the Big Indigo fault zone. All samples are SN1152-003. Dashed black outlines indicate locations of close-up photos in Fig. 11. Samples E, H, R, and S are host samples, and the rest are vein samples for which data are summarized in Table 1. (b) Close-up of main fault zone of the Big Indigo fault zone showing the three distinct mineral zones and the sites of thin section and stable isotope samples. (c) Close-up of lower exposed portion of the main splay of the Big Indigo fault zone showing the three distinct mineral zones and the sites of thin section and stable isotope samples. All samples are SN1152-003. Scale bars are 10 cm.

132

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Fig. 10. Examples of vein-fill morphologies. All photomicrographs were taken through partially crossed polars. Scale bars are 1 cm. (a) Early blocky calcite with multiple opening events followed by later blocky calcite (sample SN1152-007D). (b) Part of fault vein showing multiple opening events (sample SN1152-007B). (c) Moderately-twinned blocky calcite (sample SN1152-003F). (d) Coarsely crystalline weakly to moderately-twinned blocky calcite (sample SN1152-003L). (e) Moderately-twinned blocky calcite followed by untwined blocky calcite (sample SN1152-03N).

direct knowledge of the geothermal gradient during fluid trapping. The current geothermal gradient for the region is approximately 26
C km1 (Kron and Stix, 1982). However, a higher gradient may have been in place during the initiation of the Rio Grande rift and Basin and Range to the west. Therefore, a range of geothermal gradients (25
C km1 and 35
C km1) is used to model the trapping conditions, along with an estimated surface temperature of 20
C. The estimated trapping conditions are illustrated in Fig. 13. The main, low-temperature fluid trapping event in the Big Indigo fault is estimated to have occurred at pressure corrected temperatures of 70e89
C and pressures of 14e29 MPa (1.4e2.9 km depth). The high-temperature fluid trapping in the Big Indigo fault is estimated to have occurred at pressure corrected temperatures of 90e122
C and pressures of 20e42 MPa (2.0e4.2 km depth). 6.1.3. Hydrocarbon inclusions Dark brown to black single-phase inclusions interpreted to be liquid hydrocarbon, degraded hydrocarbons, and bitumen are found in all samples from all three faults (Table 1, Fig. 11c). Most inclusions did not fluoresce, indicating that they have lost their aromatic components (Alpern et al., 1993). Many other inclusions, however, fluoresce a weak pale blue (Fig. 11d), indicating that they typically contain liquid hydrocarbon with ~38 to >50 API gravity oil (light oil to condensate) (Fig. 14). 6.2. Stable isotope results Calcite vein d18O and d13C values (Figs. 15 and 16) show a tight clustering for samples associated with the Big Indigo fault, with most d18O values between 9.3 and 8.6‰ V-PDB and d13C values between 1.3 and 1.8‰ V-PDB. The only exceptions are a nearby small subsidiary fault (sample SN1152-003U) with slightly lower d18O and d13C values and a late fault vein within the main Big Indigo fault (sample SN1152-003N) with significantly higher d18O and d13C values (Table 1, Figs. 15 and 16). The d18O and d13C values for the Spanish Goat and Textbook fault veins generally are close to those

of the Big Indigo, but have a larger range. All vein isotope values are significantly different from their associated host rocks (Table 1, Figs. 15 and 16), indicating that the isotopic composition of the fluids (and vein calcite from the fluids) was influenced by an open fluid system. The d13C values of calcite veins fall within the range of values typical of carbonate rocks of similar age (Fig. 15; Veizer et al., 1999), yet the d18O values are significantly different. Similarly, the host rock compositions do not correspond to values expected for Late Cretaceous carbonates (Fig. 15; Veizer et al., 1999). Plots of vein isotopic composition compared with the host rock composition (Fig. 16) also support an open fluid system. However, a general trend of increasing fluid connectivity with increasing fault displacement is indicated by the d13C values. The small displacement Textbook fault has a d13C value that is the same as the host rock, while those of the larger displacement Spanish Goat fault are greater than the host rock, and the large displacement Big Indigo fault has the largest difference in d13C values from the host rock, particularly the late fault vein within the main Big Indigo fault (sample SN1152-003N).

6.3. Equilibrium fluid calculation By combining the fluid inclusion and isotopic data, the oxygen isotopic composition of the fluid that precipitated the vein minerals may be determined, thereby providing information on the nature of the fluids (such as meteoric versus formation water). The oxygen isotopic composition of calcite depends on the isotopic composition of the fluid from which it precipitates and the isotopic fractionation between the fluid and the solid. If precipitation in equilibrium with the fluids is assumed, it would depend upon the temperature. If the temperature of the fluid is known from the fluid inclusions, the isotopic composition of the fluid can be calculated using the equation of Friedman and O'Neil (1977) for calcite-water fractionation. Calculated fluid compositions from the Big Indigo fault exhibit a very narrow d18O range, from þ1.4‰ to þ4.0‰ V-

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

133

Table 1 Fault vein description and fluid inclusion microthermometric data. Sample

Vein Description Vein Orientation (strike/dip)

Textbook Fault SN1152-004A 190/38

Spanish Goat Fault SN1152-007A 015/45

SN1152-007B

N/A

SN1152-007D

355/40

Big Indigo Fault SN1152-003A 247/58

SN1152-003B

247/58

SN1152-003C

247/58

SN1152-003D1 247/58

SN1152-003D2

Mineral for Stable Isotope Analysis

Vein Calcite

Host Rocka

d13C

d13C

d18O

d18O

Range ThA (oC)

N ThA Range Tm (
C)

N Tm Hydrocarbon Inclusion Fluorescence

(V-PDB) (V-PDB)

(V-PDB) (V-PDB)

‰

‰

‰

‰

Calcite ~4.0 cm thick. Multiple stages of alternating weakly to moderately-twinned, blocky calcite and host rock lithons reflecting veinnormal dilation. At least 10 opening events.

1.320

9.071

1.256

4.963

e

e

e

e

pale blue

~2.4 cm thick. Early 0.5 cm thick layer of fine-grained blocky calcite exhibiting fault parallel extension and numerous opening events. Later moderately-twinned blocky calcite. ~8.0 cm section of fault vein with at least 80 separate opening events of 0.01 e0.3 cm of weakly-twinned to untwinned blocky calcite. ~2.0 cm thick. Early moderately-twinned blocky calcite exhibiting multiple opening events followed by moderatelytwinned blocky calcite. Early calcite has a corroded margin.

Calcite

1.260

9.406

0.701

5.253

e

e

e

e

none

Calcite

1.193

8.571

0.701

5.253

e

e

e

e

none

Early Calcite 1.427 Late Calcite 1.127

7.297 8.434

0.701 0.701

5.253 5.253

e e

e e

e e

e e

none pale blue

~5.0 cm thick vein at the margin of the 36 cm wide fault vein structure. Moderately to highlytwinned blocky calcite. ~6.0 cm thick vein within the 36 cm wide fault vein structure, 5 cm from fault margin. Moderatelytwinned blocky calcite followed by weaklytwinned blocky calcite. Early calcite has a corroded margin. From 25 cm thick vein within 36 cm wide fault vein structure. Sample taken 20 cm from hanging wall fault margin. Highlytwinned blocky calcite followed by moderatelytwinned blocky calcite. Early calcite has a corroded margin. From 25 cm thick vein within 36 cm wide fault vein structure. Sample taken near hanging wall fault margin. Moderatelytwinned blocky calcite with strong growth zoning. From 25 cm thick vein within 36 cm wide fault vein structure. Sample taken 7 cm from hanging wall fault margin. Moderately-twinned blocky calcite.

Calcite

1.516

9.328

0.724

5.011

70.9 to 76.6

5

0.0

5

none

Early Calcite 1.570 Late Calcite 1.364

8.705 8.949

0.724 0.724

5.011 5.011

e e 65.7 to 74.6, 9 80.0

e 0.0

e 9

pale blue pale blue

Early Calcite 1.456 Late Calcite 1.417

8.935 9.350

0.724 0.724

5.011 5.011

e 69.9 to 70.2

e 0.0

e 8

none pale blue

Calcite

1.609

9.069

0.724

5.011

80.5 to 88.6, 11 122.1

11

pale blue

Calcite

1.485

8.924

0.724

5.011

e

e

pale blue

e 8

e

e

(continued on next page)

134

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Table 1 (continued ) Sample

Vein Description Vein Orientation (strike/dip)

SN1152-003D3

SN1152-003F1

190/42

SN1152-003F2

SN1152-003F3

SN1152-003G

227/73

SN1152-003I

215/43

SN1152-003J

222/85

SN1152-003K

230/40

SN1152-003L

208/65

SN1152-003M

224/56

SN1152-003N

224/57

SN1152-003O

224/58

From 25 cm thick vein within 36 cm wide fault vein structure. Sample taken 12 cm from hanging wall fault margin. Highlytwinned blocky calcite followed by moderatelytwinned blocky calcite. Early calcite has a corroded margin. ~10 cm thick. Sample taken from 2 cm from hanging wall margin. Moderatelytwinned blocky calcite. Sample taken from 5 cm from hanging wall margin. Moderately-twinned blocky calcite. Sample taken from 1 cm from footwall margin. Moderately-twinned blocky calcite. ~1.5 cm thick sigmoidal vein along fault. Multiple stages of alternating very weakly to untwinned, blocky calcite and host rock lithons reflecting veinnormal dilation. At least 12 opening events. ~2.0 cm thick. Multiple stages of alternating very weakly to untwinned, blocky calcite and host rock lithons reflecting veinnormal dilation. At least 30 opening events. ~1 cm thick bed extension vein. Moderately-twinned to weakly-twinned blocky calcite. ~1.8 cm thick. Multiple stages of alternating very weakly to untwinned, blocky calcite and host rock lithons reflecting veinnormal dilation. At least 35 opening events. From 27 cm thick vein within a ~60 cm thick vein structure. Sample taken 5 cm from footwall margin of vein. Weakly to moderately-twinned blocky calcite. From 27 cm thick vein within a ~60 cm thick vein structure. Sample taken 20 cm from footwall margin of vein. Early highlytwinned blocky calcite followed by weakly to moderately-twinned blocky calcite. ~2 cm thick fault vein within a ~60 cm thick vein structure. Early moderately-twinned finegrained blocky calcite followed by very weaklytwinned to untwinned blocky calcite.

Host Rocka

Mineral for Stable Isotope Analysis

Vein Calcite (V-PDB) (V-PDB)

(V-PDB) (V-PDB)

‰

‰

‰

‰

Calcite

1.457

8.900

0.724

Calcite

1.420

8.722

Calcite

1.495

Calcite

Range ThA (oC)

N ThA Range Tm (
C)

N Tm Hydrocarbon Inclusion Fluorescence

5.011

e

e

e

pale blue

0.169

4.674

68.0 to 72.0, 14 80.0 to 83.5

14

pale blue

8.936

0.169

4.674

71.4 to 78.0, 16 84.5 to 87.4

1.0 to 0.0

16

pale blue

1.617

8.615

0.169

4.674

e

e

e

e

pale blue

Calcite

1.356

8.889

0.169

4.674

64.5 to 72.5

10

0.0

10

none

Calcite Calcite

1.454 1.448

8.858 8.869

0.283 0.283

5.429 5.429

74.8 e

7 e

e e

7 e

none none

Calcite

1.381

8.903

0.283

5.429

58.8 to 69.7, 9 91.5 to 99.9

1.8e2.5 9

none

Calcite

1.168

8.198

0.283

5.429

e

e

e

none

Calcite

1.506

8.864

0.283

5.429

54.6 to 67.5, 13 92.5 to 99.0

13

pale blue

Early Calcite 1.654 Late Calcite 1.481

8.870 9.015

0.283 0.283

5.429 5.429

e 67.3 to 74.6

e 6

pale blue none

Early Calcite 6.200

5.574

0.283

5.429

7

pale blue

Late Calcite

2.131

8.050

0.283

5.429

72.0 to 74.5, 7 84.3 to 85.7 e e

e

e

pale blue

Calcite

1.699

9.146

0.283

5.429

7

0.0

7

none

13

d C

18

d O

d13C

d18O

e

e 6

e

e e

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

135

Table 1 (continued ) Sample

SN1152-003P

Vein Description Vein Orientation (strike/dip)

208/90

SN1152-003Q1 175/50

SN1152-003Q2

SN1152-003Q3

SN1152-003T

SN1152-003U

236/16

025/62

From 30 cm thick brecciated vein within a ~60 cm thick vein structure. Sample taken 20 cm from hanging wall margin of vein. Weakly to moderately-twinned blocky calcite. ~2.0 cm thick bed extension vein. Multiple stages of alternating very weakly to untwinned, blocky calcite and host rock lithons reflecting vein-normal dilation. At least 9 opening events. 11.0 cm thick vein. Sample taken 3 cm from hanging wall fault margin. Highlytwinned to moderatelytwinned blocky calcite. Sample taken 7 cm from hanging wall fault margin. Highly-twinned to moderately-twinned blocky calcite. Sample taken 10 cm from hanging wall fault margin. Highly-twinned to moderately-twinned blocky calcite. ~3 cm thick. Early moderately-twinned finegrained blocky calcite with multiple opening events followed by weaklytwinned blocky calcite. ~2.0 cm thick. Weaklytwinned blocky calcite.

Mineral for Stable Isotope Analysis

Host Rocka

Vein Calcite 13

d C

18

d O

d13C

d18O

(V-PDB) (V-PDB)

(V-PDB) (V-PDB)

‰

‰

‰

Range ThA (oC)

N ThA Range Tm (
C)

N Tm Hydrocarbon Inclusion Fluorescence

‰ 69.6, 90.0 to 99.9

Calcite

1.593

8.794

0.283

5.429

e

e

0.0

e

none

Calcite

1.437

8.731

0.283

5.429

70.0, 81.1

7

e

7

none

Calcite

e

e

0.283

5.429

72.0 to 75.5, 7 85.3 to 93.3

0.0e1.2

7

pale blue

Calcite

e

e

0.283

5.429

70.4 to 72.0

7

0.3 to 0.0

7

none

Early Calcite 1.791

9.348

0.366

4.707

8

0.0e0.0

8

pale blue

Late Calcite

1.617

9.274

0.366

4.707

51.8, 73.9, 95.5 e

e

e

e

pale blue

Calcite

1.140

10.087 0.366

4.707

e

e

e

e

none

ThA ¼ homogenization of aqueous inclusions, Tm ¼ ice melting temperature, N ¼ number of observations. Hydrocarbon inclusion fluorescence is evaluated from dark (black) to brown-tinted single-phase. a The corresponding host rock sample(s) for each fault is denoted as follows: Textbook Fault ¼ SN1152-004B (see Fig. 4), Spanish Goat Fault ¼ SN1152-007C (see Fig. 5), Big Indigo Fault ¼ SN1152-003E (for samples A-D), SN1152-003H (for samples F-G), SN1152-003R (for samples I-Q), and SN1152-003S (for samples T-U) (see Fig. 9).

SMOW using a 25
C km1 geothermal gradient and þ0.8‰ to þ3.4‰ V-SMOW using a 35
C km1. Based on this analysis, the source fluids for the vein cement is formation water. 6.4. Summary of fluid history and conditions Fluid inclusion and stable isotope analysis of the three faults that are the focus of this study provides insight into the fluid conditions and connectivity during faulting. The smaller displacement Textbook and Spanish Goat faults contain only mature liquid and degraded hydrocarbon inclusions that were trapped during multiple fault reactivation events. The d13C values of the calcite along these faults suggest that they formed with little to no fluid connectivity with external fluids. The larger displacement Big Indigo fault, on the other hand, not only has significantly thicker fault zone mineralization (36e60 cm thick) e measured normal to the fault zone e than the smaller displacement faults (4e8 cm thick), the d13C values of the calcite indicate an open fluid system where the fluids may have been derived from an external fluid reservoir. The similarity of most of the isotope values suggest that the fluid source remained the same during faulting, except for the single SN1152-003N sample with very high d13C (þ6.2) that suggests

influx into the Big Indigo fault of microbial or thermally degraded hydrocarbons. Based on equilibrium water calculations, the calcite along the Big Indigo fault was precipitated from formation waters. However, the low-temperature, low-salinity inclusions suggest that there might have been meteoric mixing to dilute the fluids. Fluid trapping along the Big Indigo fault likely occurred at depths of 1.4e2.9 km, and possibly up to 4.2 km if the only control on fluid temperature is depth and geothermal gradient. However, it is possible that the higher-temperature, higher salinity fluids were sourced from deeper, warmer fluid reservoirs, and that there was little to no change in burial depth between trapping events. 7. Fault developmental sequence Comparing observations from the three faults discussed here leads to the following interpretations. Faults appear to have nucleated with lithologically controlled failure modes, including hybrid or opening mode failure through the most competent chalk and limestone beds, and shear failure in less competent mudrock beds. Failure surfaces cutting the chalk and limestone beds are generally very steep, dipping 70e90
, implying failure angles <20


136

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Fig. 11. Fluid inclusion photomicrographs. (a) Aqueous inclusions without a vapor bubble. (b) Fluid inclusion assemblage with two-phase aqueous inclusions (with arrows) and some without a vapor bubble. (c) Single-phase hydrocarbon inclusions that appear dark brown to black through the microscope appear grey to black in photo. (d) Photograph of same area shown in (c) under UV fluorescence e inclusions that show a weak pale blue fluorescence through the microscope appear grey in image (examples are indicated by white arrows). Scale bar in all images is 50 mm.

Fig. 12. Summary plot of homogenization temperature (Th) versus water ice melting temperature (Tm) for aqueous inclusion in the Big Indigo fault vein calcite. Black squares are data for inclusions with both Th and Tm values. Histogram is a summary of all Th values with N being the number of measurements. Note the general increase in inclusion salinity with increasing Th.

with respect to interpreted maximum principal stress direction at failure angles (c.f., Ramsey and Chester, 2004). These failure surfaces also often exhibit composite shear and opening segments, and show oblique opening (e.g., Fig. 4d) e these characteristics are all consistent with hybrid failure (c.f., Ferrill et al., 2012b, 2014b). Faults that cut multiple layers refract through the heterolithic multilayer, as described by Ferrill et al. (2014b); also Peacock (2002); Wilkins and Gross (2002); Ferrill and Morris (2003); Ferrill et al. (2012a); Petrie et al. (2014); Giorgetti et al. (2016). Although there are examples where vertical segmentation is observed, similar to the structural style and evolution described by Micarelli and Benedicto (2008) or Micarelli et al. (2005), relatively simple refracted fault architecture with steep dilational fault

segments smoothly transitioning upward and downward into more gently dipping shear segments appears to dominate early fault development. The lack of evidence of a pre-existing opening mode or tensile fracture network in the form of a systematic joint or vein network, or evidence of offset joint or vein tips associated with the faults, further supports propagation of refracted faults rather than shear linkage and reactivation of pre-existing opening mode fractures. In a few cases, fault segmentation, anomalous layer dips, and anomalous fault dip changes appear to result from displacement transfer and curved propagation of fault segments that are associated with segment linkage at branch lines. The majority of dip changes appear to be continuous failure surfaces from bed to bed, consistent with propagation of a fault along a refracted fault profile,

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Fig. 13. Pressure versus temperature diagram showing fluid trapping conditions for fluid inclusions in the Big Indigo fault. Black isochores and field are for the main, lowtemperature trapping event. Grey isochores and field are for the later, possibly highertemperature trapping events. Lithostatic (solid lines) and hydrostatic (dashed lines) theromobaric gradients for 25
C km1 and 35
C km1 also are shown.

Fig. 14. Fluorescence colors of liquid hydrocarbon inclusions from Big Indigo and Textbook faults plotted on a CIE-1931 diagram (Commission Internationale de  l'Eclairage, 1971; 1986; e.g., McLimans, 1987; Conliffe et al., 2010). Increasing hydrocarbon maturity is toward the lower left corner. Numbers represent approximate API gravity. Squares represent data from the Big Indigo fault and the triangle represents data from the Textbook fault.

137

Fig. 15. Plot of d13C composition versus d18O composition of calcite veins. Filled shapes are vein compositions and unfilled shapes are host rock compositions. Squares represent data from the Big Indigo fault, diamonds are from the Spanish Goat fault, and the triangle represents the Textbook fault. Also shown is the range of isotopic compositions for seawater carbonate from the Late Cretaceous (Veizer et al., 1999).

et al., 1997; Lee and Wiltschko, 2000). Repeated reactivation and cementation continued following the same overall pattern to displacements on the order of 10's of cm or more, forming calcite sheets connecting competent beds across the fault zone to displacements of 10's of cm or more (Fig. 17 e Stage 2). Displacements sufficient to juxtapose separate competent chalk or limestone beds are represented by two stacked calcite sheets (Fig. 17 e Stage 3). Larger displacement faults where several competent chalk or limestone beds pass a reference chalk or limestone beds across the fault exhibit three (Fig. 17 e Stage 4) or more calcite sheets stacked by this process, with the massive prismatic to blocky calcite increasingly common with increasing fault displacement. The association of massive prismatic to blocky calcite within only the central portions of the larger displacement faults, and that postdates early crack-seal fault zone textures, indicates that later slip events may have involved centimeters of slip, opening large voids along faults, whereas early slip increments were typically sub-millimeter displacements. Larger single-event slip magnitudes would generally be expected to occur only on larger faults (c.f., Wells and Coppersmith, 1994; Ferrill et al., 2008) and such events are likely to connect permeability over larger distances laterally and vertically, progressively opening up fluid communication. This interpretation is compatible with the observed stable isotope geochemistry data for the Big Indigo fault and indications of an “external” fluid source. 8. Slip and dilation tendency analysis

rather than linkage of originally isolated, unconnected segments. Fault displacement occurred by slip parallel to more gently dipping shear segments and oblique dilation of steep segments. This fault displacement opened fracture porosity, primarily along the steep fault segments, and calcite precipitated in this fracture porosity and healed the fracture to form calcite veins (Fig. 17 e Stage 1). Fault reactivation tended to localize failure at the vein/wall rock interface, which resulted in detached thin veneers of host wall rock that were incorporated into the veins as they grew. This progressive reactivation and cementation produced an antitaxial “crack-seal” texture (younger vein material at the wall-rock margins of the vein), similar to observed faults in the Austin Chalk (Lee

We interpret these faults to have formed in a normal-faulting stress regime consistent with southeast directed extension around the margin of the Gulf of Mexico basin. Results of stress inversion for the normal faulting along Sycamore Creek (2 km away) by Ferrill et al. (2014b), using the approach of McFarland et al. (2011, 2012; also see Morris et al., 2016) showed relative magnitudes for s1: s2: s3 of 1.0: 0.57: 0.35. As discussed in Ferrill et al. (2014b), if s1 was defined by s1 ¼ sv  44 MPa or higher, consistent with 2 km burial, then s2  25 MPa and s3  15 MPa, which we consider reasonable for stress conditions during activity of the faults in the present study. Here we apply this stress

138

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

der Pluijm, 2012; Haines et al., 2013, 2014). The most obvious low slip tendency portion of a fault in the analyses of the three faults is along the upper part of the NW segment of the Big Indigo fault (Fig. 18c) where the shear portions of the NW segment have dips between 0
and 20
. Numerous nearvertical low slip tendency segments are evident along all three faults. 8.2. Dilation tendency Dilation tendency analysis of the Textbook, Spanish Goat, and Big Indigo faults using the interpreted normal faulting regime stress tensor shows that the steeply dipping fault segments, which are primarily in chalk and limestone layers, tend to have the highest dilation tendencies. Fault segments with moderate to gentle dips have moderate to low dilation tendencies. Associated with the high dilation tendency fault segments are veins that are cemented with calcite, and these steep segments are the wall-rock contacts for calcite sheets that have grown down the fault between offset chalk or limestone bed segments. As with the slip tendency analysis, the most obvious low dilation tendency portion of a fault in the analyses of the three faults is along the upper part of the NW segment of the Big Indigo fault (Fig. 18c) where the fault dips between 0
and 20
. This shallowly dipping portion of the fault is not well oriented for slip in the interpreted normal faulting regime because it is nearly perpendicular to the maximum principal stress and has low resolved shear stress and high resolved normal stress. However, we interpret that the fault formed in a perturbed stress field associated with the displacement transfer between the two overlapping fault segments. 9. Discussion and implications

Fig. 16. (a) d13C versus d13C composition of calcite veins. (b) Plot of whole rock d18O versus d18O composition of calcite veins. Squares represent data from the Big Indigo fault, diamonds represent data from the Spanish Goat fault, and the triangle represents data from the Textbook fault.

condition to analyze the slip tendency and dilation tendency of the Textbook, Spanish Goat, and Big Indigo faults. Because these faults likely initiated and were at least initially active in local stress fields that were ideally oriented for slip on those faults, we assume that the intermediate principal stress was parallel to the strike of the faults, and is therefore perpendicular to the interpreted profile section, which contains the maximum principal stress (vertical) and the minimum principal stress (horizontal). 8.1. Slip tendency Slip tendency analysis of the Textbook, Spanish Goat, and Big Indigo faults using the interpreted normal faulting regime stress tensor shows that the moderately dipping fault segments, which are primarily in mudrock layers, tend to have the highest slip tendencies (Fig. 18). These segments generally lack calcite cement and are very narrow zones in the outcrop. Low slip tendency surfaces tend either to have very low dips (approaching horizontal) or very steep dips (approaching vertical). Segments with very low dip are situated in mudrock, where the value of slip tendency at which slip occurs is lower due to the lower coefficient of friction of clay-rich mudrock compared with chalk and limestone (Haines and van

This study has demonstrated that natural faults in the mechanically layered Eagle Ford Formation are controlled by a combination of shear and hybrid fracturing rather than shear fracturing alone. Based upon the fluid inclusion and stable isotope data, we infer that this faulting initiated and continued at depths of 1.4e4.2 km. This depth range corresponds to the depth range through which horizontal production well drilling and associated hydraulic fracturing are being conducted within the Eagle Ford Formation. Throughout much of the productive Eagle Ford play, lateral well placement targets the organic rich pelagic section that we referred to as the middle Eagle Ford in this study and which is the general interval in which the current analysis is focused. Induced hydraulic fracturing is often monitored by recording microseismic events that provide direct indication of shear and hybrid failure (Busetti et al., 2014) or reactivation of faults in the stimulation zone. The optimal orientation of lateral drilling throughout much of the Eagle Ford play in south Texas is generally considered to be NW or SE in the interpreted present-day minimum principal stress direction. Potential motivations for this drilling direction are to (i) encounter natural fractures that are likely to be permeable pathways for movement of hydrocarbons to the wellbore, and that may also provide some (minor) storage of hydrocarbons, and (ii) take advantage of the stress field to optimize the generation of new fractures during induced hydraulic fracturing. We suggest that the observations presented in this paper are relevant to understanding the character of natural small displacement faults. As described before by Ferrill et al. (2014b) and McGinnis et al. (2017), opening mode fractures tend to have very limited heights within the Eagle Ford Formation, are best developed in carbonate rich chalk or limestone beds, and commonly

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

139

Fig. 17. Illustration of the progressive development of layered calcite cement in fault zone with analog photos from all three faults. These photos were previously featured in Figs. 4c, 7e and 5b and 7c (top to bottom).

140

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Fig. 18. Slip (left) and dilation (right) tendency analysis of (a) Textbook (see Fig. 4 for context), (b) Spanish Goat (see Fig. 5 for context), and (c) Big Indigo faults (see Fig. 7 for context).

terminate into vertically adjacent mudrock beds. Faults e even those with small displacements e are able to propagate across multiple beds, providing significant vertical connectivity within the Eagle Ford strata. Furthermore, the Sycamore Bluff faults, having formed in similar rock and under similar in situ conditions to those of the productive Eagle Ford hydrocarbon play, serve as useful analogs for understanding the newly formed induced fractures created during hydraulic fracturing operations. Faults like these that propagate or connect across multiple mechanical layers link complex natural and induced bed-restricted fracture networks, generate permeability conduits, and lead to the fracture network complexity desired by producers. These results are broadly relevant to other mechanically layered unconventional hydrocarbon reservoirs (e.g., Gale et al., 2014). Natural or induced refracted faults can produce high permeability

zones that are self-propping, cut multiple mechanical layers, and interconnect otherwise vertically isolated fracture networks within the heterolithic mechanical multilayer (e.g., Ferrill et al., 2014b). These fault zone characteristics are in stark contrast to faults in hydrocarbon producing conventional clastic reservoir sections that are often characterized by permeability reducing cataclastic fault zone deformation and shale- or clay-smear (e.g., Antonellini and Aydin, 1995; Yielding et al., 1997, 2010). Whereas fault-related cataclasis and shale- or clay-smear tend to reduce permeability and interconnectivity within conventional reservoirs, faults tend to increased permeability in fine-grained source rocks that have very low permeability. A recent study of cap-rock seals for conventional reservoirs by Petrie et al. (2014) found that mechanical layering in fine-grained, heterolithic sequences strongly influenced faulting and fracturing processes. They identified the occurrence of

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

“extensional shear” failure (hybrid failure in our terminology) in outcrop examples, and explored the depth-dependent occurrence of extensional shear versus extensional (opening mode) failure. The various studies documenting the vast array of faulting processes in hydrocarbon reservoir and seal strata, as well as microscale deformation mechanisms in experiments and nature (e.g., Ferrill and Groshong, 1993), demonstrate the importance of considering mechanical stratigraphy and deformation conditions in prediction of deformation mechanisms and the related influences on permeability and fluid flow. 10. Conclusions Analysis of normal faults in heterolithic mudrock, chalk, limestone, and volcanic ash in the Eagle Ford Formation (south Texas) at Sycamore Bluff, southeast of Del Rio, Texas, indicates bed-scale mechanical stratigraphic control on failure modes and fault zone architecture associated with normal faults. A 1.5 km long natural cliff of the Eagle Ford Formation exposes southeast and northwest dipping normal faults, with displacements of 0.01e7 m. Fault dips are steep to vertical through chalk and limestone beds, and moderate through mudrock and clay-rich ash. Failure modes included hybrid failure in chalk and limestone and shear failure in mudrock and ash beds, resulting in refracted fault profiles. Slip on shear segments caused dilation of the steeper hybrid segments e these steep fault segments were cemented with calcite. With repeated fault slip, dilation, and cementation, sheets of calcite formed to link offset competent beds obliquely across the fault zones. These sheets of calcite extend meters along the larger faults representing numerous slip events, with vertical thicknesses proportional to thickness of the offset competent beds. Fluid inclusion and stable isotope geochemistry analyses indicate fault zone cementation during episodic reactivation at 1.4e4.2 km depths. These observations constrain fault zone mechanics and fluid flow associated with natural faulting with implications for induced fracturing in unconventional hydrocarbon reservoirs. Acknowledgements We thank Janice Moody and Heath Grigg for allowing us research access to the Rancho Rio Grande. Financial support for this work was provided by Southwest Research Institute's Eagle Ford joint industry project (SwRI project number 17264), funded by Anadarko Petroleum Corporation, BHP Billiton, Chesapeake Energy Corporation, ConocoPhillips, Eagle Ford TX LP, EP Energy, Hess Corporation, Marathon Oil Corporation, Murphy Exploration and Production Company, Newfield Exploration Company, Pioneer Natural Resources, and Shell. We thank the staff of these sponsor companies for the interaction and constructive feedback. We thank Carolina Giorgetti and Michael Kettermann for their thorough reviews and constructive comments on the manuscript, and Editor Joao Hippertt for his efficient handling of the manuscript. References Alpern, B., Lemos de Sousa, M.J., Pinheiro, H.J., Zhu, X., 1993. Detection and evaluation of hydrocarbons in source rocks by fluorescence microscopy. Org. Geochem. 20, 789e795. Antonellini, M., Aydin, A., 1995. Effect of faulting on fluid flow in porous sandstones: geometry and spatial distribution. AAPG Bull. 79, 642e671. Aydin, A., Basu, A., 2005. The Schmidt hammer in rock material characterization. Eng. Geol. 81, 1e14. Bakker, R.J., 2003. Package FLUIDS. Computer programs for analysis of fluid inclusion data and from modelling bulk fluid properties. Chem. Geol. 194, 3e23. http://dx.doi.org/10.1016/S0009-2541(02)00268-1. Bakker, R.J., Brown, P.E., 2003. Computer modeling in fluid inclusion research. In: Samson, I., Anderson, A., Marshall, D. (Eds.), Fluid Inclusions: Analysis and Interpretation. Mineralogical Association of Canada, Ottawa, pp. 175e212. Short

141

Course Series 32. Basu, N., Barzola, G., Bello, H., Clarke, P., Viloria, O., 2012. Eagle Ford Reservoir Characterization from Multisource Data Integration. American Association of Petroleum Geologists. Search and Discovery Article #80234. Blakey, R., 2011. Paleogeography of North America (Colorado Plateau Geosystems. Arizona, USA). Bodnar, R.J., 1993. Revised equation and table for determining the freezing point depression of H2OeNaCl solutions. Geochim. Cosmochim. Acta 57, 683e684. Bodziak, R., Clemons, K., Stephens, A., Meek, R., 2014. The role of seismic attributes in understanding the ‘frac-able’ limits and reservoir performance in shale reservoirs: an Example from the Eagle Ford Shale, South Texas, USA. AAPG Bull. 98, 2217e2235. Boon, X.Y.R., Bacon, R., 2014. Attribute-based fracture inference in a faulted carbonate play. Interpretation 2, SA77eSA82. Bourdet, J., Eadington, P., 2012. Fluorescence and Infrared Spectroscopy of Inclusion Oil. CISRO Technical Report EP129625. Busetti, S., Jiao, W., Reches, Z., 2014. Geomechanics of hydraulic fracturing microseismicity: Part 1. Shear, hybrid, and tensile events. AAPG Bull. 98, 2439e2457. Byerlee, J., 1978. Friction of rocks. Pure Appl. Geophys. 116, 615e626. Collins, E.W., Hovorka, S.D., 1997. Structure Map of the San Antonio Segment of the Edwards Aquifer and Balcones Fault Zone, South-central Texas: Structural Framework of a Major Limestone Aquifer. The University of Texas at Austin, Bureau of Economic Geology, Kinney, Uvalde, Medina, Bexar, Comal, and Hays Counties. Miscellaneous Map No. 38, scale 1:250, 000, 2 sheets.  Commission Internationale de l'Eclairage, 1971. Colorimetry. Publication No. 15.  Commission Internationale de l’Eclairage, Paris.  Commission Internationale de l'Eclairage, 1986. Standard on Colorimetric Ob servers. CIE S002. Commission Internationale de l’Eclairage, Vienna. Conliffe, J., Blamey, N.F., Feely, M., Parnell, J., Ryder, A.G., 2010. Hydrocarbon migration in the Porcupine Basin, offshore Ireland: evidence from fluid inclusion studies. Pet. Geosci. 16, 67e76. Donovan, A.D., Staerker, T.S., 2010. Sequence stratigraphy of the Eagle Ford (Boquillas) formation in the subsurface of South Texas and outcrops of West Texas. Gulf Coast Assoc. Geol. Soc. Trans. 60, 861e899. Dooley, T.P., Jackson, M.P.A., Hudec, M.R., 2013. Coeval extension and shortening above and below salt canopies on an uplifted, continental margin: application to the northern Gulf of Mexico. AAPG Bull. 97, 1737e1764. Eldrett, J.S., Ma, C., Bergman, S.C., Ozkan, A., Minisini, D., Lutz, B., Jackett, S.-J., Macaulay, C., Kelly, A.E., 2015. Origin of limestoneemarlstone cycles: Astronomic forcing of organic-rich sedimentary rocks from the Cenomanian to early Coniacian of the Cretaceous Western Interior Seaway, USA. Earth Planet. Sci. Lett. 423, 98e113. Eldrett, J.S., Minisini, D., Bergman, S.C., 2014. Decoupling of the carbon cycle during ocean anoxic event 2. Geology 42, 567e570. tegui, R.S., 2009. Joint sets that enhance production Engelder, T., Lash, G.G., Uzca from Middle and Upper Devonian gas shales of the Appalachian Basin. AAPG Bull. 93, 857e889. Erdlac Jr., R.J., 1990. A Laramide-age push-up block: the structures and formation of the Terlingua-Solitario structural block, Big Bend region, Texas. Geol. Soc. Am. Bull. 102, 1065e1078. Erdlac Jr., R.J., 1994. Laramide paleostress trajectories from stylolites in the Big Bend region. In: Laroche, T.M., Viveiros, J.J. (Eds.), Structure and Tectonics of the Big Ben Area and Southern Permian Basin, Texas. West Texas Geological Society 1994 Field Trip Guidebook. West Texas Geological Society Publication 94e95, Midland Texas, 165-187. Evans, M.A., Battles, D.A., 1999. Fluid inclusion and stable isotope analysis of veins from the central Appalachian Valley and Ridge province: implications for regional syn-orogenic hydrologic structure and fluid migration. Geol. Soc. Am. Bull. 111, 1841e1860. Evans, M.A., Delisle, A., Leo, J., Lafonte, C.J., 2014. Deformation conditions for fracturing in the Middle Devonian sequence of the central Appalachians during the Late Paleozoic Alleghenian orogeny. AAPG Bull. 98, 2263e2299. Evans, M.A., Bebout, G.E., Brown, C.H., 2012. Changing fluid conditions during folding: an example from the central Appalachians. Tectonophysics 576e577, 99e115. Ewing, T.E., 1991. Chapter 3 e structural framework. In: Salvador, A. (Ed.), The Geology of North America Volume J., the Gulf of Mexico Basin. The Geological Society of America, pp. 31e52. Ewing, T.E., 2003. Review of the tectonic history of the lower Rio Grande Border Region, South Texas and Mexico, and implications for hydrocarbon exploration. Soc. Indep. Prof. Earth Sci. Newsl. 40, 16e21. Ewing, T.E., 2009. The Ups and downs of the Sabine Uplift and the Northern Gulf of Mexico Basin: Jurassic Basement Blocks, Cretaceous thermal Uplifts, and Cenozoic Flexure. Gulf Coast Assoc. Geol. Soc. Trans. 59, 253e269. Ewing, T.E., 2010. Pre-Pearsall geology and exploration plays in South Texas. Gulf Coast Assoc. Geol. Soc. Trans. 60, 241e260. Ferrill, D.A., Groshong Jr., R.H., 1993. Deformation conditions in the northern Subalpine Chain, France, estimated from deformation modes in coarse-grained limestone. J. Struct. Geol. 15, 995e1006. Ferrill, D.A., McGinnis, R.N., Morris, A.P., Smart, K.J., 2012a. Hybrid failure: field evidence and influence on fault refraction. J. Struct. Geol. 42, 140e150. Ferrill, D.A., McGinnis, R.N., Morris, A.P., Smart, K.J., Sickmann, Z.T., Bentz, M., Lehrmann, D., Evans, M.A., 2014b. Control of mechanical stratigraphy on bedrestricted jointing and normal faulting: Eagle Ford Formation, south-central Texas, U.S.A. AAPG Bull. 98, 2477e2506.

142

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143

Ferrill, D.A., Morris, A.P., 2003. Dilational normal faults. J. Struct. Geol. 25, 183e196. Ferrill, D.A., Morris, A.P., 2008. Fault zone deformation controlled by carbonate mechanical stratigraphy, Balcones fault system, Texas. AAPG Bull. 92, 359e380. Ferrill, D.A., Morris, A.P., Hennings, P.H., Haddad, D.E., 2014a. Faulting and fracturing in shale and self-sourced reservoirs: Introduction. AAPG Bull. 98, 2161e2164. Ferrill, D.A., Morris, A.P., McGinnis, R.N., 2012b. Extensional fault-propagation folding in mechanically layered rocks: the case against the frictional drag mechanism. Tectonophysics 576e577, 78e85. Ferrill, D.A., Morris, A.P., McGinnis, R.N., Smart, K.J., Ward, W.C., 2011. Fault zone deformation and displacement partitioning in mechanically layered carbonates: the Hidden Valley fault, central Texas. AAPG Bull. 95, 1383e1397. Ferrill, D.A., Morris, A.P., McGinnis, R.N., Smart, K.J., 2016a. Myths about normal faulting. In: Childs, C., Holdsworth, R.E., Jackson, C.A.-L., Manzocchi, T., Walsh, J.J., Yielding, G. (Eds.), The Geometry and Growth of Normal Faults. Geological Society, London. Special Publications 439. http://doi.org/10.1144/ SP439.12. Ferrill, D.A., Morris, A.P., McGinnis, R.N., Smart, K.J., Wigginton, S.S., Hill, N.J., 2017. Mechanical stratigraphy and normal faulting. J. Struct. Geol. 95, 160e170. Ferrill, D.A., Morris, A.P., Wigginton, S.S., Smart, K.J., McGinnis, R.N., Lehrmann, D., 2016b. Deciphering thrust fault nucleation and propagation and the importance of footwall synclines. J. Struct. Geol. 85, 1e11. Ferrill, D.A., Sims, D.W., Waiting, D.J., Morris, A.P., Franklin, N., Schultz, A.L., 2004. Structural framework of the Edwards Aquifer recharge zone in south-central Texas. Geol. Soc. Am. Bull. 116, 407e418. Ferrill, D.A., Smart, K.J., Necsoiu, M., 2008. Displacement-length scaling for singleevent fault ruptures: insights from Newberry Springs Fault Zone and implications for fault zone structure. In: Kurz, W., Imber, J., Wibberley, C.A.J., Collettini, C., Holdsworth, R.E. (Eds.), The Internal Structure of Fault Zones: Fluid Flow and Mechanical Properties. Geological Society, London, pp. 113e122. Special Publication 299. Ferrill, D.A., Winterle, J., Wittmeyer, G., Sims, D., Colton, S., Armstrong, A., Morris, A.P., 1999. Stressed rock strains groundwater at Yucca Mountain, Nevada. GSA Today 9 (5), 1e8. Foley, L.L., 1926. Mechanics of the Balcones and Mexia faulting. AAPG Bull. 10, 1261e1269. Fowler, P., 1956. Faults and folds of south-central Texas. Gulf Coast Assoc. Geol. Soc. Trans. 6, 37e42. bourg, G., Ruppel, S.C., Loucks, R.G., Lambert, J., 2016. Depositional controls on Fre sediment body architecture in the Eagle Ford/Boquillas system: insights from outcrops in west Texas, United States. AAPG Bull. 100, 657e682. Friedman, I., O'Neil, J.R., 1977. Compilation of stable isotope fractionation factors of geochemical interest. In: Fleischer, M. (Ed.), Data of Geochemistry Chapter KK. United States Geological Survey Professional Paper 440-KK. Gale, J.F.W., Laubach, S.E., Olson, J.E., Eichhubl, P., Fall, A., 2014. Natural fractures in shale: a review. AAPG Bull. 98, 2165e2216. Giorgetti, C., Collettini, C., Scuderi, M.M., Tesei, T., 2016. Fault geometry and mechanics of marly carbonate multilayers: an integrated field and laboratory study from the Northern Apennines, Italy. J. Struct. Geol. 93, 1e16. Goldhammer, R.K., Johnson, C.A., 2001. Middle Jurassic e Upper Cretaceous, paleogeographic evolution and sequence stratigraphic framework of the Northwest Gulf of Mexico Rim. In: Bartolini, C., Buffler, R.T., Cantu Chapa, A. (Eds.), The Western Gulf of Mexico Basins, Tectonics, Sedimentary Basins, and Petroleum Systems, pp. 45e81. AAPG Memoir 75. Goldstein, R.H., Reynolds, T.J., 1994. Systematics in Fluid Inclusions in Diagenetic Minerals. SEPM Short Course 31. Groshong Jr., R.H., 2006. 3-D Structural Geology. Springer-Verlag, Heidelberg. Halbouty, M.T., 1966. Stratigraphic-trap possibilities in Upper Jurassic rocks, San Marcos arch, Texas. AAPG Bull. 50, 3e24. Haines, S.H., Kaproth, B., Marone, C., Saffer, D., van der Pluijm, B., 2013. Shear zones in clay rich fault gouge: a laboratory study of fabric development and evolution. J. Struct. Geol. 51, 206e225. Haines, S., Marone, C., Saffer, D., 2014. Frictional properties of low-angle normal fault gouges and implications for low-angle normal fault slip. Earth Planet. Sci. Lett. 408, 57e65. Haines, S., van der Pluijm, B., 2012. Patterns of mineral transformations in clay gouge, with examples from low-angle normal fault rocks in the western USA. J. Struct. Geol. 43, 2e32. Hancock, P.L., 1985. Brittle microtectonics: principles and practice. J. Struct. Geol. 7, 437e457. Harmsen, S.C., 1994. The Little Skull Mountain, Nevada, earthquake of 29 June 1992: Aftershock focal mechanisms and tectonic stress field implications. Seismol. Soc. Am. Bull. 84, 1484e1505. Haymond, D., April 1991. The Austin Chalk e an overview. Houst. Geol. Soc. Bull. 27e34. Hentz, T.F., Ruppel, S.C., 2010. Regional lithostratigraphy of the Eagle Ford Shale: Maverick Basin to East Texas Basin. Gulf Coast Assoc. Geol. Soc. Trans. 60, 325e337. Hovorka, S.D., Nance, H.S., 1994. Dynamic depositional and early diagenetic processes in a deep-water shelf setting, Upper Cretaceous Austin Chalk, North Texas. Gulf Coast Assoc. Geol. Soc. Trans. 44, 269e276. Hudec, M.R., Jackson, M.P., Peel, F.J., 2013. Influence of deep Louann structure on the evolution of the northern Gulf of Mexico. AAPG Bull. 97, 1711e1735. Ikari, M.J., Saffer, D.M., Marone, C., 2007. Effect of hydration state on the frictional properties of montmorillonitebased fault gouge. J. Geophys. Res. 112, B06423. http://dx.doi.org/10.1029/2006JB004748.

Jamison, W.R., 1979. Laramide Deformation of the Wingate Sandstone, Colorado National Monument: a Study of Cataclastic Flow. Ph.D. Dissertation. Texas A&M University, College Station, Texas, U.S.A., 170pp. Katz, O., Reches, Z., Roegiers, J.-C., 2000. Evaluation of mechanical rock properties using a Schmidt Hammer. Int. J. Rock Mech. Min. Sci. 37, 723e728. Kettermann, M., Urai, J.L., 2015. Changes in structural style of normal faults due to failure mode transition: first results from excavated scale models. J. Struct. Geol. 74, 105e116. Kettermann, M., Von Hagke, C., Van Gent, H.W., Grützner, C., Urai, J.L., 2016. Dilatant normal faulting in jointed cohesive rocks: a physical model study. Solid Earth 7, 843e856. Kron, A., Stix, J., 1982. Geothermal Gradient Map of the United States Exclusive Alaska and Hawaii. National Oceanographic and Atmospheric Administration National Geophysics Data Center, Boulder Colorado scale 1:2,500,000. Laubach, S.E., Jackson, M.L., 1990. Origin of arches in the northwestern Gulf of Mexico basin. Geology 18, 595e598. Lee, Y.-J., Wiltschko, D.V., 2000. Fault controlled sequential vein dilation: competition between slip precipitation rates in the Austin Chalk, Texas. J. Struct. Geol. 22, 1247e1260. Lee, Y.-J., Wiltschko, D.V., Grossman, E.L., Morse, J.W., Lamb, W.M., 1997. Sequential vein growth with fault displacement: an example from the Austin Chalk Formation, Texas. J. Geophys. Res. 102, 22,611e22,628. Lehman, T.M., 1991. Sedimentation and tectonism in the Laramide Tornilla Basin of west Texas. Sediment. Geol. 75, 9e28. Livingston, P., Sayre, A.N., White, W.N., 1936. Water Resources of the Edwards Limestone in the San Antonio Area, Texas. United States Department of the Interior, pp. 59e113. Water-Supply Paper 773-B. Lock, B.E., Peschier, L., 2006. Boquillas (Eagle Ford) upper slope sediments, west Texas: outcrop analogs for potential shale reservoirs. Gulf Coast Assoc. Geol. Soc. Trans. 56, 491e508. Lock, B.E., Peschier, L., Whitcomb, N., 2010. The Eagle Ford (Boquillas formation) of Val Verde County, Texas e a window on the south Texas play. Gulf Coast Assoc. Geol. Soc. Trans. 60, 419e434. Maclay, R.W., Small, T.A., 1983. Hydrostratigraphic subdivisions and fault barriers of the Edwards Aquifer, south-central Texas, USA. J. Hydrol. 61, 127e146. Maclay, R.W., 1989. Edwards Aquifer in San Antonio: its hydrogeology and management. Bull. South Tex. Geol. Soc. 30 (4), 11e28. Martin, R., Baihly, J., Malpani, R., Lindsay, G., Atwood, L., 2011. Understanding production from Eagle Ford-Austin Chalk system. In: Society of Petroleum Engineers Annual Technical Conference and Exhibition, 30 October e 2 November. Society of Petroleum Engineers, Denver, Colorado. Maxwell, R.A., Lonsdale, J.T., Hazzard, R.T., Wilson, J.A., 1967. Geology of the Big Bend National Park, Brewster County, Texas. University of Texas at Austin. Bureau of Economic Geology Publication 6711. McFarland, J., Morris, A., Bichon, B., Riha, D., Ferrill, D., McGinnis, R.N., 2011. Geological Stress State Calibration and uncertainty analysis. Proc. Soc. Exp. Mech. Ser. 3, 557e570. McFarland, J., Morris, A.P., Ferrill, D.A., 2012. Stress inversion using slip tendency. Comput. Geosci 41, 40e46. McGinnis, R.N., Ferrill, D.A., Morris, A.P., Smart, K.J., 2016. Insight on mechanical stratigraphy and subsurface interpretation. In: Krantz, B., Ormand, C., Freeman, B. (Eds.), 3-D Structural Interpretation: Earth, Mind, and Machine, pp. 111e120. AAPG Memoir 111. McGinnis, R.N., Ferrill, D.A., Morris, A.P., Smart, K.J., Lehrmann, D., 2017. Mechanical stratigraphic controls on natural fracture spacing and penetration. J. Struct. Geol. 95, 160e170. McGinnis, R.N., Ferrill, D.A., Smart, K.J., Morris, A.P., Higuera-Diaz, C., Prawica, D., 2015. Pitfalls of using entrenched fracture relationships: fractures in bedded carbonates of the hidden Valley Fault Zone, Canyon Lake Gorge, Comal County, Texas. AAPG Bull. 99, 2221e2245. McLimans, R.K., 1987. The application of fluid inclusions to migration of oil and diagenesis in petroleum reservoirs. Appl. Geochem. 2, 585e603. Micarelli, L., Benedicto, A., 2008. Normal fault terminations in limestones from the SE-Basin (France): implications for fluid flow. In: Kurz, W., Imber, J., Wibberley, C.A.J., Holdsworth, R.E., Collettini, C. (Eds.), The Internal Structure of Fault Zones: Implications for Mechanical and Fluid-flow Properties. Geological Society, London, pp. 123e138. Special Publications 299. Micarelli, L., Benedicto, A., Invernizzi, C., Saint-Bezar, B., Michelot, J.L., Vergely, P., 2005. Influence of P/T conditions on the style of normal fault initiation and growth in limestones from the SE-Basin, France. J. Struct. Geol. 27, 1577e1598. Moore, D.E., Lockner, D.A., 2004. Crystallographic controls on the frictional behavior of dry and water-saturated sheet structure minerals. J. Geophys. Res. 109, B03401. http://dx.doi.org/10.1029/2003JB002582. Morris, A.P., Ferrill, D.A., Henderson, D.B., 1996. Slip tendency analysis and fault reactivation. Geology 24, 275e278. Morris, A.P., Ferrill, D.A., McGinnis, R.N., 2009a. Fault frequency and strain. Lithosphere 1, 105e109. Morris, A.P., Ferrill, D.A., McGinnis, R.N., 2009b. Mechanical stratigraphy and faulting in Cretaceous carbonates. AAPG Bull. 93, 1459e1470. Morris, A.P., Ferrill, D.A., McGinnis, R.N., 2016. Using fault displacement and slip tendency to estimate stress states. J. Struct. Geol. 83, 60e72. Morrow, C., Radney, B., Byerlee, J., 1992. Frictional strength and the effective pressure law of montmorillonite and illite clays. In: Evans, B., Wong, T.-F. (Eds.), Fault Mechanics and Transport Properties of Rocks, pp. 69e88. Moustafa, A.R., 1988. Structural Geology of Sierra del Carmen, Trans-Pecos Texas.

D.A. Ferrill et al. / Journal of Structural Geology 97 (2017) 118e143 Geologic Quadrangle Map 54. Bureau of Economic Geology: Austin, Texas, The University of Texas at Austin, scale 1:48,000. Peacock, D.C.P., 2002. Propagation, interaction and linkage in normal fault systems. Earth-Science Rev. 58, 121e142. Petrie, E.S., Evans, J.P., Bauer, S.J., 2014. Failure of cap-rock seals as determined from mechanical stratigraphy, stress history, and tensile-failure analysis of exhumed analogs. AAPG Bull. 98, 2365e2389. Phelps, R.M., Kerans, C., Loucks, R.G., Da Gama, R.O.B.P., Jeremiah, J., Hull, D., 2013. Oceanographic and eustatic control of carbonate platform evolution and sequence stratigraphy on the Cretaceous (ValanginianeCampanian) passive margin, northern Gulf of Mexico. Sedimentology 61, 461e496. Ramsey, J.M., Chester, F.M., 2004. Hybrid fracture and the transition from extension fracture to shear fracture. Nature 428, 63e66. Reches, Z., Lockner, D.A., 1994. Nucleation and growth of faults in brittle rock. J. Geophys. Res. 99, 18159e18173. Robinson, C.R., 1997. Hydrocarbon source rock variability within the Austin Chalk and Eagle Ford Shale (Upper Cretaceous), East Texas, U.S.A. Int. J. Coal Geol. 34, 287e305. Roche, V., Homberg, C., Rocher, M., 2012a. Architecture and growth of normal fault zones in multilayer systems: a 3D field analysis in the South-Eastern Basin, France. J. Struct. Geol. 37, 19e35. Roche, V., Homberg, C., Rocher, M., 2012b. Fault displacement profiles in multilayer systems: from fault restriction to fault propagation. Terra Nova 24, 499e504. Roche, V., Homberg, C., Rocher, M., 2013. Fault nucleation, restriction, and aspect ratio in layered sections: quantification of the strength and stiffness roles using numerical modeling. J. Geophys. Res. Solid Earth 118, 1e15. Roedder, E., 1984. Fluid Inclusions. Mineralogical Society of America Reviews in Mineralogy, p. 12. Rose, P.R., 1984. Oil and gas occurrence in Lower Cretaceous rocks, Maverick Basin area, southwest Texas. In: Smith, C.I. (Ed.), Stratigraphy and Structure of the Maverick Basin and Devils River Trend, Lower Cretaceous, Southwest Texas: a Field Guide and Related Papers. San Antonio Geological Society, pp. 80e93. Salvador, A., 1987. Late Triassic-Jurassic paleogeography and origin of Gulf of Mexico basin. AAPG Bull. 71, 419e451. Scott, R.W., 2010. Cretaceous stratigraphy, depositional systems and reservoir facies of the northern Gulf of Mexico. Gulf Coast Assoc. Geol. Soc. Trans. 60, 597e609. Sibson, R.H., 1998. Brittle failure mode plots for compressional and extensional tectonic regimes. J. Struct. Geol. 20, 655e660. Sibson, R.H., 2000. Fluid involvement in normal faulting. J. Geodyn. 29, 469e499.

143

Smart, K.J., Ofoegbu, G.I., Morris, A.P., McGinnis, R.N., Ferrill, D.A., 2014. Geomechanical modeling of hydraulic fracturing: why mechanical stratigraphy, stress state, and pre-existing structure matter. AAPG Bull. 98, 2237e2261. Touret, J.L.R., 2001. Fluids in Metamorphic rocks. Lithosphere 55, 1e25. Turner, K.J., Berry, M.E., Page, W.R., Lehman, T.M., Bohannon, R.G., Scott, R.B., Miggins, D.P., Budahn, J.R., Cooper, R.W., Drenth, B.J., Anderson, E.D., Williams, V.S., 2011. Geologic Map of Big Bend National Park. U.S. Geological Survey Scientific Investigations Map 3142, scale 1: 75,000. Treadgold, G., McLain, B., Sinclair, S., Nicklin, D., 2010. Eagle Ford Shale Prospecting with 3D Seismic Data within a Tectonic and Depositional System Framework. Bulletin of the South Texas Geological Society LI, pp. 19e28. U. S. Energy Information Administration, 2014. Eagle Ford Production Increasingly Targets Oil-rich Areas. http://www.eia.gov/todayinenergy/detail.cfm?id¼14951. Accessed August 31, 2016. Veizer, J., Ala, D., Azmy, K., Bruckschen, P., Buhl, D., Bruhn, F., Carden, G.A.F., Diener, A., Ebneth, S., Godderis, Y., Jasper, T., Korte, C., Pawellek, F., Podlaha, O., Strauss, H., 1999. 87Sr/86Sr, d 13C and d18O evolution of Phanerozoic seawater. Chem. Geol. 161, 59e88. Webster, R.E., 1980. Structural analysis of Devils River Upliftesouthern Val Verde Basin, southwest Texas. AAPG Bull. 64, 221e241. Weeks, A.W., 1945. Balcones, Luling, and Mexia fault zones in Texas. AAPG Bull. 29, 1733e1737. Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bull. Seismol. Soc. Am. 84, 974e1002. Wilkins, S.J., Gross, M.R., 2002. Normal fault growth in layered rocks at Split Mountain, Utah: influence of mechanical stratigraphy on dip linkage, fault restriction and fault scaling. J. Struct. Geol. 24, 1413e1429. Yielding, G., Freeman, B., Needham, D.T., 1997. Quantitative fault seal prediction. AAPG Bull. 81, 897e917. Yielding, G., Bretan, P., Freeman, B., 2010. Fault Seal Calibration. A brief review. Geol. Soc. London Spec. Pub. 347, pp. 243e255. Young, K., 1986. Cretaceous, marine inundations of the San Marcos platform, Texas. Cretac. Res. 7, 117e140. Young, S.R., 1982. Characterization of and Parameters Controlling Small Faults in Naturally Deformed, Porous Sandstones. M.S. Thesis. Texas A&M University, College Station, Texas, 118 pp. Zoback, M.D., 2007. Reservoir Geomechanics. Cambridge University Press, Cambridge, 449 pp.