The successes of George Mitchell and the team at Mitchell Energy & Development Corp. in migrating hydraulic fracturing technology from conventional reservoirs to mudrock for the extraction of natural gas from the Barnett Shale of the Fort Worth Basin brought about a permanent paradigm shift in the oil and gas industry around the world. Prior to the mid-2000s, organic-rich mudrocks at depth were viewed typically as the “oil kitchen,” or as seals, or as “frac” barriers. If any significant analytical work was done on shales, it was to determine seal capacity, and seldom were conventional cores intentionally taken in them. With the lack of industrial interest in mudrocks to that point, there had been little work done by the industry or academia to understand the reservoir attributes of shales (Jurgen Scheiber, personal communication, 2006).

This pivot to gas shale was facilitated by the then recent successes in the petroleum industry in hydraulically fracturing tight gas sandstones and, to a lesser extent, the extraction of coal-bed methane. Those successes provided the industrial impetus to move into shale and learn how to exploit source rocks as reservoirs, but also highlighted the need for measurements of the geomechanical properties of both reservoir units and their bounding layers. Laboratory measurements of rock strength properties on core samples are used to calibrate wireline sonic data to build brittleness profiles through the potential reservoir, and through any underlying or overlying bounding units that could slow or stop fracture propagation. The rock mechanics data are loaded into engineering simulator programs to test different fracture design programs for efficiency and prospects of success. These simulators allow engineers to examine the effects of changes in fluid type and viscosity, pump rates, hydraulic pressure, rate of injection, and proppant types and concentrations on vertical and lateral fracture growth and proppant concentrations, with the goal of maximizing fracture contact with the formation while avoiding zones with moveable water, or zones that have no commercial value.

Initial work on the Barnett and several other dry-gas-window mudrocks indicated a correlation between rock mechanical properties such as Young’s modulus and Poisson’s ratio (PR) and the relative proportion of “brittle” minerals such as quartz and carbonate as determined by x-ray diffraction (XRD) mineralogy. Several authors, and company engineers, developed algorithms to determine the proportion of brittle minerals against total clay volume and organic content such as:

These simple arithmetic constructs are referred to as “brittleness indices” (BIs), and different authors sometimes choose to incorporate different constituents in the calculations. These indices provided at least qualitative insight into changes in mechanical properties within and between formations in a given basin, and specific index values or ranges of values were used to proxy ranges of Young’s modulus and PRs. However, when these correlations were exported to other formations of different geological age, or into basins with different subsidence histories, engineers were often unable to predict the rock mechanical properties of the new formations using XRD mineralogy. For example, a BI–mechanical properties relationship in the Barnett or Marcellus shales was unable to predict the mechanical properties in the Devonian Woodford Shale of Oklahoma even in zones where their XRD mineralogy is nearly the same.

The question then is: Why is bulk mineralogy alone not the dominant control on mudrock mechanical properties? Our early investigations into mineralogy and rock mechanical properties showed only a crude relationship between BI and rock strength as characterized by Young’s modulus and PR (Figure 1). Although there is a general trend with increasing BI, there is also substantial scatter in the data. We explored other relationships such as BI vs. depth, and Young’s modulus and PR vs. depth (Figure 2) but found poor correlations. This work will investigate the control that initial composition and the relative timing of several diagenetic processes have on the development of brittleness in siliceous mudrocks and their mechanical properties.

The analysis of the relationship between brittleness and composition builds from the work of Pollastro (1993) and utilizes an integrated data set that has been collected over the past 12 years, designed to identify the parameters controlling reservoir quality and production properties in mudrocks. It includes, among other data, XRD bulk and clay mineralogy, total helium porosity, total organic carbon (TOC) content, vitrinite reflectance ratio or vitrinite reflectance ratio-equivalent data, and rock mechanics measurements (Young’s modulus, PR) derived from samples taken from conventional cores from mudrock formations in North America and is the property of Core Laboratories NV. The thermal maturation levels in these formations range from the biogenic gas window, through the thermogenic oil window, to the dry gas window. Each data point has a formation, age, location, and depth associated with it. The sampling process involved drilling horizontal plugs from the two-thirds section of 4-inch–diameter conventional cores from which ultra-thin section petrography, XRD mineralogy, and geochemical analyses could be conducted on the same layer. Vertical plugs for rock mechanics were cut adjacent to the XRD samples at specific depths where needed to calibrate the sonic logs and completions design simulators. The remaining core material was used for the measurement of porosity and other rock petrophysical parameters using a protocol developed by the Gas Research Institute (Guidry et al., 1995).

Analysis of these data will focus on the relationship between mineralogy and rock mechanical properties with depth and burial history. The database was filtered first for the subset of samples with associated, measured rock mechanical properties, and then subsequently for mineralogy, TOC, and porosity. Those data were then sorted by formation, age, and depth and form the foundation for the analyses in this chapter. The organic geochemistry data are used to develop a burial and thermal framework against which changes in XRD mineralogy and rock mechanical properties can be analyzed.

The focus of this chapter is on mudrocks (maximum grain size <62 microns) with less than 5 vol. % carbonate minerals (Figure 3). This filtering removes the effects of carbonate diagenesis/recrystallization and cementation on the development of brittleness and allows us to focus on siliceous mudrocks with compositions dominated by quartz and clay. We will also limit the discussion to marine, organic-bearing mudrocks.

Mineralogy herein is expressed in terms of XRD detectability, and the discussion will contrast “initial” or depositional composition, including both granular (detrital) quartz and biogenic silica, with “terminal” or the rock composition, following burial diagenesis. As mentioned above, the samples analyzed by XRD used in this work have companion ultra-thin sections, which allow for a visual comparison of the volume of detrital quartz grains against the total XRD quartz volume. This is especially useful in samples where XRD quartz volumes are substantially higher than the detrital quartz volume. Using these thin-section observations, we then infer that a portion of the XRD quartz is present as a cement.

The “initial” composition of fine-grained sediments is broadly controlled by energy levels and productivity of the water column. Organic-bearing sediment accumulates in suspension-dominated aqueous settings characterized by low current energy and can include hemipelagic and pelagic sediments. The requisite environment is typically far enough down the energy gradient that the hemipelagic input is minor or ephemeral and includes a water column capable of producing pelagic, hydrocarbon-rich organic material including microorganisms and fecal material (Sageman et al., 2003). In the oceans, this material is referred to as marine snow, and water temperature, light levels, turbidity, water chemistry, and bathymetry determine the composition and volume. Hemipelagic sediments include terrigenous mud and silt brought in by a variety of processes including hyperpycnal flow, gravity-driven density flows, and fallout of sediment brought to the basin by either hypopycnal flows or aeolian processes (Abouelresh and Slatt, 2011). Changes in the relative influences of these processes can reflect climate change, and changes in sea level (Arthur and Sageman, 2005). These changes are recorded in micro-, meso-, and macro-scale heterogeneity in sediment grade and composition, stratification, and organic content and can affect the initial composition, namely the relative proportion of biogenic silica in the form of opaline tests of various microorganisms, clay, and detrital quartz (May and Anderson, 2013).

The relationship between the “initial” and “terminal” composition of mudrocks has been treated by Lazar et al. (2015) and is shown in Figure 4. This ternary diagram broadly illustrates a proportional, volumetric relationship among the dominant constituents in mudrocks, namely clay, quartz, and carbonate minerals. They break out the sources of the terminal quartz in a second ternary plot, centered on the quartz apex, that partitions detrital grains, biogenic quartz, and detrital clays.

Three mudrock systems from North America have been selected to examine the development of brittleness and the factors controlling brittleness in mudrocks. Mudrocks of the Western Interior Cretaceous Seaway provide a laboratory to examine the effects of increasing burial depth (and increasing temperature) on clay-rich sediments with varying amounts of detrital quartz as noted in thin-section analysis. These shale systems have undergone a relatively simple burial history including, typically, a single episode of subsidence, burial, and heating, followed by subsequent uplift. Present day depths range from a few hundred meters to almost 3500 m (11,483 ft). The Devonian Marcellus Shale will be analyzed to examine the effects of greater age and higher temperatures on the diagenesis of a second argillaceous system that, like the Western Interior Cretaceous mudrocks, can contain abundant quartz silt based on thin-section observation. Finally, the Woodford Shale of Oklahoma allows us to examine the effects of diagenesis on a biogenic quartz-dominated system that is characterized by interbedded chert, and siliceous shale with variable detrital silt volumes.

Unlike sandstone diagenesis, mudrock diagenesis is complicated by the presence of organic matter (see Gautier et al., 1985). Nevertheless, the integrated data set used in this study suggests that there are three recognizable diagenetic processes influencing the development and timing of brittleness in siliceous mudrocks that are also operative in sandstones. These are mechanical compaction, the generation of quartz by of the conversion of smectite to illite, and the generation of quartz by the conversion of biogenic opal-A to quartz.

Mechanical compaction (induced by increasing overburden pressure) results in decreased particle spacing and an increasing number of particle contact points and can lead to an increase in pore pressure by reducing the permeability of the sediment/rock (Burst, 1976). Rutter et al. (2017) presented an informed overview on the effects of burial compaction and increasing temperature on porosity loss in shales, and the mechanical properties of shales based on composition. Their observations are broadly similar to those of the author; however, while they discussed the effects of bulk composition on the mechanical properties of shales, they did not mention the timing of development of diagenetic minerals or the nature of the various constituents. Both of these aspects influence the changes in brittleness of a given mudrock formation at different depths across a basin, as well as the differences between two mudrocks of similar bulk composition.

Mud at the time of deposition on the sea floor is characterized by upward of 75–80% water-filled porosity and a grain proportion of 20–25% (Baldwin, 1971; Baldwin and Butler, 1985). An exponential burial compaction curve in Figure 5 is drawn using data digitized from Baldwin (1971) and indicates a loss of 25% of the depositional porosity within the first 50 m (164 ft) of burial. Much of this early compaction is accomplished through pore-water expulsion and reduction in particle spacing. Compaction caused by dewatering has probably terminated by 300 m (984 ft), and porosity may be as low as one-half of the original volume. Further compaction is accomplished by mechanical deformation of ductile grains (clays and organic matter) (Burst, 1976), and by 3000 m (9843 ft) porosity is down to between 10 and 15%, with a reciprocal grain proportion of 85–90%. Also, as the grain proportion increases, so does bulk density.

The illitization of smectite has been treated at length by several authors (see for example Eberl and Hower, 1976; Hower et al., 1976; Boles and Franks, 1979), and the reader may refer to that body of work for details of the chemistry and kinetics. Relevant to this discussion is that the conversion of smectite to illite is driven by temperature and pressure and, to a lesser degree, by time (Pollastro, 1993) and releases large volumes of mobile silica according to this equation by the authors cited above:

This reaction begins in the temperature range of 50–60°C (122–140°F), prior to the onset of hydrocarbon generation (Hoffman and Hower, 1979; Pollastro, 1993). A portion of the generated silica is exported to interbedded sandstones where it is precipitated as quartz overgrowths (Boles and Franks, 1979), but euhedral quartz overgrowths can likewise be found during SEM analysis of mature mudrocks (Thyberg et al., 2010). The process also generates large volumes of water that can contribute to the total pore water volume.

In sediments containing common or abundant biogenic silica, typically in the form of opaline radiolaria or diatoms, the generation of authigenic quartz occurs in stepwise transitions beginning with the conversion of detrital, amorphous, hydrous, biogenic opal-A to more ordered opal-CT, and a second transition from opal-CT to quartz (Weller and Behl, 2017). The temperatures at which these transitions occur are affected by the presence and proportional volume of other detrital components, particularly aluminum. The aluminum increases the temperature at which opal-A converts to opal-CT but decreases the temperature at which opal-CT converts to quartz (see Figure 6). According to the diagram by Weller and Behl (2017) in Figure 6, the temperature range for the first transition is from 45 to 50°C (113–122°F) (pre-oil-window), and the temperature range for the transition to quartz is between 60 and 90°C (140 and 194°F), which corresponds to the early oil window.

The analysis of Mesozoic mudrocks is based on 18 data points from 12 wells from Alberta and Saskatchewan, Canada, and from Montana, Wyoming, and Texas, USA. The formations are all Late Cretaceous, and the stratigraphic units include the Colorado Group in Alberta and Saskatchewan, Canada, and Montana, USA; the Mowry Shale in Wyoming, USA; and the Woodbine Shale in Texas, USA.

There are two depth groupings in the data set—the first between 240 and 587 m (787 and 1926 ft) and the second between 1879 and 3128 m (6165 and 10,262 ft). Compositionally, the Mesozoic and Cenozoic mudrocks display a wide range in quartz volume from approximately 35 to 85% with the majority around 40 to 50% (Figure 7, Table 1). Vitrinite reflectance data indicate immature to early oil window maturation levels for the shallow group, and oil to wet-gas window maturation in the deeper group. Figure 8 is a plot of total porosity by depth for the Cretaceous rocks in this study together with the compaction curve from Figure 5. Average porosity is 29% in the shallow group and 7% in the deeper group. Of note here is that the compaction curve derived from Baldwin (1971) predicts slightly higher porosities for many of the deeper samples. However, Baldwin (1971) used porosity data from both unconsolidated deep sea muds, terrigenous clay and other mud types, as well as assumptions based on water content, and bulk densities from various subsurface formations. The resulting composite may not fully account for the effects of cementation.

Figures 9A, B are thin-section and SEM photomicrographs of an example from the shallower group at a depth of 326 m (1070 ft). It is from the Milk River Shale (Colorado Group) in Montana, USA, with a measured porosity of 27%, and it contains abundant detrital quartz in the very fine to coarse silt range. XRD analysis indicates that the sample is composed of 53 vol. % quartz and 30 vol. % total clay. Visual estimates from the thin section are in the range of 20–30% and suggest that most of the quartz is contained in the detrital grains. Many of the Western Interior Cretaceous mudrocks were deposited in open to restricted shelf environments, and it is common to find burrows and heterogeneity in thin sections. A conventional (broken surface) SEM image of this sample reveals a weakly to moderately compacted fabric expressed by crenulated clay particles and little to no authigenic cement (Figure 9B). In this case, the “terminal” XRD quartz volume is predominantly detrital.

A more deeply buried (3128 m [10,262 ft]) Cretaceous mudrock example from the Woodbine Shale in Texas is shown in Figures 9C, D. In this case, clay is the most abundant constituent (at 60 vol. %) with quartz silt as a minor component (at 30 vol. %). The thin section, however, contains very little silt-size material. The Argon-Ion-Milled SEM image shows a moderately to well-compacted fabric with scattered detrital grains that have induced local, differential compaction. Measured porosity is 10%, which is less than half of the porosity in the Milk River Shale sample. The terminal XRD quartz volume in the XRD data for this sample includes authigenic quartz.

Figure 10A illustrates the change in static Young’s modulus with depth for the Mesozoic mudrocks and ranges from 0.14 to 1.72 GPa (0.02 ×3 10⁶ to 0.25 × 10⁶ psi) in the shallow group and from 8.36 to 30.52 GPa (1.21 × 10⁶ to 4.43 × 10⁶ psi) in the deeper group. PR though exhibits no correlation with depth and ranges broadly within each depth group from 0.17 to 0.38 in the shallow group and from 0.20 to 0.29 in the deeper group (Figure 10B). The relationship between total porosity and Young’s modulus is illustrated in Figure 11 and shows an abrupt increase in Young’s modulus below 20% porosity. This corresponds to a depth of approximately 800–1000 m (2625–3281 ft) on the Cretaceous trendline in Figure 8.

The Marcellus data set includes 27 samples from wells in Pennsylvania and West Virginia, USA, ranging in depth from 1745 to 2996 m (5725 to 9829 ft) (Table 1). Quartz ranges from 25 to 52% by volume and averages 37%, and total clay ranges from 24 to 56% by volume and averages 41% (Figure 12). Total, normalized organic volume ranges from 2 to 28% and averages 9%. Total porosity ranges from 4 to 19% and averages 8% and, unlike the Cretaceous mudrocks, shows no correlation with depth (Figure 13A). Vitrinite or vitrinite-equivalent reflectance values range from 1.28 to 4.27% indicating condensate-wet gas window to over-mature thermal conditions.

Figure 13B is a depth plot of the static Young’s modulus data from the Marcellus and shows no correlation with depth. Values range from 6.34 to 27.27 GPa (0.92 × 10⁶ to 3.95 × 10⁶ psi) with an average of 14.35 GPa (2.08 × 10⁶ psi). A depth plot of PR is shown in Figure 13C and also shows no correlation with depth. PR ranges from a low of 0.17–0.31 and averages 0.24.

Thin-section and argon ion milled—SEM analyses typically document abundant quartz silt as in a sample from the Marcellus Shale at a depth of 2182 m (7159 ft) (Figure 14). Poorly developed quartz overgrowths are common on quartz silt grains (Figure 14B), and the clay particles display differential compaction around the rigid grains. XRD data indicate that the sample is composed of 45 vol. % clay and 32 vol. % quartz. This is in rough agreement to the amount of quartz silt in the thin-section photomicrograph in Figure 14A.

Two, broadly defined lithologies comprise the Woodford Shale of Oklahoma. The most common rock type is organic-rich, siliceous shale containing varying amounts of quartz silt, and the secondary lithology is organic, black chert. The black chert beds range from a few centimeters to upward of 7 or 8 cm (2.8 to 3.1 in.) in thickness and are locally interbedded with the siliceous shale. The cherts are typically more quartz rich than the siliceous shales, but the two lithologies exhibit an overlap in mineralogy (Figure 15). Vitrinite or vitrinite-equivalent reflectance values range from 0.76 to 1.32% in the Woodford indicating oil to dry-gas window maturation (Table 1).

Young’s modulus in the Woodford siliceous shale ranges from 15.17 to 32.47 GPa (2.20 × 10⁶ to 4.71 × 10⁶ psi) and averages 21.76 GPa (3.16 × 10⁶ psi) (Figure 16A), and PR ranges from 0.17 to 0.24 and averages 0.20 (Figure 16B). BIs range from 0.46 to 0.84 and average 0.62 (Table 1).

Young’s modulus in the Woodford cherts is higher compared to the siliceous shale ranging from 24.34 to 63.71 GPa (3.53 × 10⁶ to 9.24 × 10⁶ psi) and averages 39.40 GPa (5.71 × 10⁶ psi) (Figure 16A). Poisson’s ratio ranges from 0.13 to 0.25 and averages 0.20 (Figure 16B). The BI of the chert beds ranges from 0.66 to 0.91 and averages 0.75 (Table 1).

The plot of mechanical properties for the siliceous shale and chert lithologies by depth in Figure 16 reveals no correlation with current burial depth. However, for depths between 2000 and 4100 m (6562 and 13,451 ft), the cherts have substantially higher Young’s moduli than the siliceous shales. Qualitatively, we would expect the cherts to be harder than the shales, but in some cases, the cherts yield Young’s moduli that are two to three times the siliceous shale moduli for a given BI.

Figure 17A is a thin-section image of a siliceous shale in the Woodford from a depth of 3935 m (12,910 ft). Quartz comprises 31 vol. % of the sample with clay and organic matter together making up 49 vol. %, and the BI is 0.39 because of the low quartz volume. Most of the white grains are quartz, and a visual estimate of about 20% is not substantially less than the XRD volume of 30%, indicating that detrital quartz is contributing to the mineralogy-based calculation of BI. The AIM–SEM image in Figure 17B reveals common clay and mica particles, with scattered quartz grains with euhedral outlines suggesting the presence of quartz overgrowths. The Young’s modulus is 20.0 GPa (2.90 × 10⁶ psi), and PR is 0.17.

In contrast is the chert lithology represented by the sample from a depth of 2445 m (8022 ft) highlighted in Figure 17C. It contains much more quartz (76 vol. %) and less clay than the silty shale, but thin-section analysis indicates little silt, and partially compacted, silica-replaced algal cysts within a lamina, and compacted cysts in the matrix around it. The AIM–SEM image (Figure 17D) reveals minimal clay particles, and abundant authigenic quartz, which is expressed by interlocking euhedral crystals of various sizes. The BI is 0.84, Young’s modulus is 29.9 GPa (4.34 × 10⁶ psi), and the PR is 0.20. Based on thin-section observations, the terminal XRD quartz volume in the Woodford cherts is mostly to entirely composed of authigenic quartz, with only traces of detrital quartz.

To evaluate the impact of diagenesis on mudrocks of differing compositions and the resulting changes in brittleness, we will outline two pathways to brittleness that start with two different initial compositions (Figure 18) and relate the key processes broadly to depth and temperature using Figure 19. The purpose of this discussion is to focus on compaction and the dominant, quartz cement-generating processes in siliceous mudrocks, but it is understood that the panorama of mudrock diagenesis is broader and more complex than is being treated here. Concomitant generation of hydrocarbons can effect changes in pH and CO³ isotope composition just to name two of the complexities (see for example Elmore et al., 2016; Wickard et al., 2016).

Pathway 1: Clay/Detrital quartz-dominated mudrock—The compaction curve shown in Figure 5 indicates that the porosity in a water-laden, smectite-rich mud is reduced from 75% at the time of deposition to less than 20% by 1500 m (4921 ft) where temperatures could range between 40 and 50°C (104 and 122°F). The illitization of smectite could be underway at this point generating silica and quartz cement, and the volume of illite in the resulting mixed-layer illite-smectite would be in the range of 25%. By 2500 m (8202 ft), mechanical compaction has probably reached its maximum with interparticle porosity near 10%, and the temperature could be 75°C (167°F). From this point, the key inorganic, diagenetic process is the generation of authigenic quartz cement by the illitization of smectite.

By 3500 m (11,483 ft), a mudrock formation can be exposed to temperatures possibly in excess of 100°C (212°F), and most of the authigenic quartz from smectite alteration has been generated. The ratio of illite to smectite in mixed-layer clay is nearly 50%. This is also within the range of homogenization temperatures for hydrocarbon-filled fluid inclusions (McLimans, 1987). For the purposes of this discussion, the hypothetical formation used here has gained most or all of its brittleness because of the generation of authigenic quartz cement at this point. The deep Cretaceous, Marcellus, and Woodford siliceous shale formations have reached these depths. Their compaction is essentially complete, and their mechanical properties have been augmented by the development of quartz cement—mostly by the illitization of smectite. As a result, these formations show no correlation between current depth and rock strength. The BIs for these formations include this authigenic quartz cement and the original quartz fraction still present in the rock.

Pathway 2: Biogenic quartz/clay-dominated mudrock—Isaacs (1981) reports that in diatomaceous rocks of the Monterey Formation in California, USA, the sediments underwent a loss of porosity at the depth corresponding to the transformation of opal-A to opal-CT, and another loss in porosity during the transition from opal-CT to quartz at greater depths. These porosity losses are attributed to the dissolution and reprecipitation of silica during both transitions.

During the first step, the porosity drops from a depositional porosity of 55–70% to 25–40% in the opal-CT facies, depending on bulk composition. As mentioned previously, this occurs between 40 and 50°C (104 and 122°F), depending on aluminum content. Behl and Garrison’s (1994) opal diagenesis chart is included in Figure 19, and this temperature corresponds to the early illitization of smectite. The next step occurs at the transition from opal-CT to quartz where porosity drops to 10–20% at 65–80°C (149–176°F), which corresponds to the later phases of illitization. Isaacs (1981) suggests that the compactional porosity loss is due to the loss of “framework strength” during the transition from one phase to the next. When Isaacs’ (1981) stepwise “curve” for biogenic quartz porosity loss is compared to the compaction curve in Figure 5, we see that they begin with similar porosities at the surface and end with similar porosities below 3000 m (9843 ft).

Each of these steps also produces increases in bulk density as well. Isaacs (1981) reports an average of 0.7–1.1 g/cc in diatomaceous rocks (opal-A) and 1.4–1.7 g/cc densities in opal-CT phase rocks. Bulk densities in the quartz phase range from 1.8 to 2.1 g/cc. Parallel to this, Weller and Behl (2017) report abrupt increases in formation hardness using a rebound hammer system between the various opal phases with the transition from opal-A to opal-CT having the most significant increase, and the opal-CT to quartz transition being less pronounced.

Figure 20 illustrates the porosity types and loss in a biogenic silica-rich sediment using SEM images from conventional cores from the Monterey Formation in Ventura and Kerns counties, California, USA. The uppermost sample from a depth of 1483 m (4865 ft) contains abundant diatom frustules with large volumes of intraparticle porosity in addition to the porosity associated with the enclosing matrix. Porosity in this sample was measured at 35%, and the BI is 0.20 because of the absence of XRD-detectable crystalline quartz. The sample at a depth of 1855 m (6086 ft) is characterized by silicified opal-CT lepispheres and has a porosity of 26%, and a BI of 0.13, again, low because of the absence of crystalline quartz. The sample at a depth of 2930 m (9613 ft) contains abundant authigenic quartz with no indications of opal varieties either visually or in the XRD data. The porosity is 5%, and the BI is 0.70.

The full data set for this analysis of BI and rock mechanical properties is plotted in Figure 21 and illustrates graphically a nebulous trend of increasing Young’s modulus with increasing BI. For example, note the broad range of Young’s modulus from less than 1 GPa (0.15 × 10⁶ psi) to over 25 GPa (3.63 × 10⁶ psi) for rocks with BI around 0.40 and also that Young’s moduli around 15 GPa are associated with BIs ranging from less than 0.40 to more than 0.60. It appears then that bulk XRD mineralogy may only weakly control the mechanical properties in siliceous mudrocks. This scatter may be due to the presence of ductile clays and variable proportions of detrital quartz, which contributes to the total XRD quartz volume but would not comprise a cement.

Data from the Cretaceous mudrocks particularly highlight the role that increasing burial depth with increasing compaction and cementation plays in increasing rock strength. The more tightly compacted deep formations (Figure 8) have a significantly higher average Young’s modulus at 18.85 GPa (2.74 × 10⁶ psi) compared to 0.78 GPa (0.11 × 10⁶ psi) for the shallow systems and higher BI (averaging 0.54) compared to the shallow formations (averaging 0.46). The increase in BI with depth in the Cretaceous mudrocks may reflect the addition of quartz cement from clay diagenesis as discussed previously.

Figure 21 also reveals a trend in the average values for the deep Cretaceous, Marcellus, and Woodford siliceous shale facies despite the scatter of the data set as a whole, and the presence of detrital quartz. In contrast, the average Young’s modulus for the Woodford chert is 39.40 GPA (5.72 × 10⁶ psi) (large, black cross symbol) and lays high off-trend of the deep Cretaceous, Marcellus, and Woodford Shale facies, given its BI. The substantially higher proportion of biogenic quartz compared to detrital quartz in the XRD data results in a much higher Young’s modulus compared to the silty mudrocks. If we were to use the trendline algorithm to predict a Young’s modulus value for these cherts, we would get a value of only 27 GPa (3.92 × 10⁶ psi).

Brittleness in siliceous mudrocks is determined at the time of deposition and is created during the course of burial diagenesis. Environmental controls on the accumulation and proportion of clay, detrital quartz, and biogenic silica set the stage for two pathways. Both are characterized by compaction and the creation of authigenic quartz; however, the rate of compaction and the source of the authigenic quartz are characteristics of the individual pathways. The magnitude of compactional porosity loss in both trajectories is significant and similar.

In muds containing abundant clay (mostly smectite), compaction is more or less a continuous process during the early phases of burial with interparticle spacing decreasing along a power-law curve. The detrital smectite begins to convert to illite at temperatures in the range of 50–60°C (122–140°F) prior to hydrocarbon generation. This process yields abundant authigenic quartz, some of which remains in the sediment as a cement and serves to increase formation brittleness.

Fine-grained sediment containing abundant biogenic silica likewise follows a path in which compaction characterizes the early burial history. However, most of the compactional porosity loss occurs in a stepwise manner at two different depths/temperature points. The first point is the depth/temperature at which opal-A is converted to opal-CT, and the second is the depth/temperature at which opal-CT is converted to quartz. At each point, the porosity decreases whereas the volume of quartz and hardness of the formation increases.

Both detrital and authigenic quartz contribute to XRD quartz volumes and to an XRD-based, arithmetic BI. In sandy or silty mudrock formations, the XRD quartz volume includes the detrital sand and silt, which increases the calculated BI of the formation. However, the granular, detrital quartz may not contribute to the brittleness of the rock much like isolated ball bearings mixed into a block of modeling clay do not contribute to the mechanical properties of the modeling clay. In this case, the brittleness of the formation is primarily a function of the volume of authigenic quartz generated by the alteration of smectite to illite and, to a lesser degree, compaction. Formations rich in biogenous quartz cement and no or little detrital quartz can exhibit very high Young’s moduli that tie well with the BI because of the absence or paucity of ductile clay, and of detrital quartz grains that cannot contribute to brittleness.