Brittleness in the Devonian Horn River shale, British Columbia, Canada

Brittleness in the Devonian Horn River shale, British Columbia, Canada

Journal of Natural Gas Science and Engineering 62 (2019) 247–258 Contents lists available at ScienceDirect Journal of Natural Gas Science and Engine...

3MB Sizes 0 Downloads 10 Views

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering journal homepage: www.elsevier.com/locate/jngse

Brittleness in the Devonian Horn River shale, British Columbia, Canada a,∗

b

b

c

c

Al Moghadam , Nicholas B. Harris , Korhan Ayranci , Juan S. Gomez , Nathalia A. Angulo , Rick Chalaturnykc a b c

T

Petroleum Engineering Technology Department, Northern Alberta Institute of Technology (NAIT), Edmonton, Canada Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, Canada Department of Civil and Environmental Engineering, University of Alberta, Edmonton, Canada

ARTICLE INFO

ABSTRACT

Keywords: Brittleness Horn river Fracability Hardness Young's modulus Poisson's ratio

The Middle and Upper Devonian Horn River Group shale in northeast British Columbia is recognized as a major source of natural gas in Canada. This shale unit is approximately 200–500 m thick and consists of three stratigraphic units in ascending order: Evie Member, Otter Park Member and Muskwa Formation. In this study, Uniaxial Compressive Strength (UCS) tests were conducted on core plugs across stratigraphic units of the Horn River Group. Shear and compressional wave velocities are also measured for the core plugs to establish a relationship between static and dynamic elastic parameters. Brittleness is measured using three independent methods: (1) a hand-held hardness tester is used to measure mechanical hardness along a core; (2) the Elemental Capture Spectroscopy (ECS) log is used to calculate a composition-based brittleness in the same well; (3) Density and sonic logs along with laboratory experiments are used to calculate a brittleness coefficient based on elastic (static and dynamic) parameters. In order to compare brittleness and rock composition, a detailed lithofacies analysis was also conducted. Three major lithofacies were observed in the Horn River Group, including massive mudstones, pyritic mudstones and laminated mudstones. The Muskwa Formation and Evie Member are dominantly represented by massive and pyritic mudstones while the vast majority of the Otter Park Member is represented by laminated mudstones. The three independent measurements of brittleness show similar trends of higher brittleness in Muskwa Formation and Evie Member compared to Otter Park. This indicates that all three methods are suitable in identifying brittle zones in shale formations. The results show a clear relationship between lithofacies and brittleness. Massive mudstones and pyritic mudstones lithofacies show relatively higher brittleness compared to laminated mudstones lithofacies. Additionally, brittleness is influenced by the mineralogy of the rock. Clay content is observed to have the most dominant (inverse) relationship with brittleness coefficients, followed by quartz content. The analysis is further confirmed by calculating the fracability index for the Horn River Group.

1. Introduction The development of shale gas reservoirs requires horizontal drilling and multistage hydraulic fracturing. To improve the economics of these reservoirs, oil and gas companies target sweet spots with respect to hydrocarbon saturation, petrophysical properties, natural fracture systems, and the reservoir's ability to respond to hydraulic fracturing, more specifically brittleness. It is therefore beneficial to identify the zones that will undergo brittle failure. Increasing recognition of the importance of brittleness to the development of shale reservoirs (Wang and Gale, 2009; Rybacki et al., 2016) has focused recent research on a number of aspects of shale geomechanics: compositional controls on brittleness (Aoudia et al., 2010; Dong et al., 2017b); extraction of



elastic moduli from well and seismic data (Abbas et al., 2018a); and derivation of universal definitions of brittleness (Rahimzadeh Kivi et al., 2018). Our research focuses on a comparison of dynamic and static elastic moduli. Dynamic moduli, typically extracted from well logs or seismic data, rely on measurements of wave velocity and rock density, and represent the most abundant and continuous information on geomechanical properties. Static moduli, though more appropriate in describing rock responses to mechanical loads, are almost always derived from sparse and discontinuous lab measurements. Critical research questions focus on the applicability of these measurements for understanding and predicting rock brittleness. Although static moduli measurements are expensive and sparse, they usually have a direct

Corresponding author. E-mail address: [email protected] (A. Moghadam).

https://doi.org/10.1016/j.jngse.2018.12.012 Received 21 August 2018; Received in revised form 20 November 2018; Accepted 11 December 2018 Available online 18 December 2018 1875-5100/ © 2018 Elsevier B.V. All rights reserved.

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 1. Map of Horn River Basin and adjacent areas. The Horn River Basin is highlighted in grey. The location of the well under study is also highlighted (modified after Dong et al., 2017b).

relationship with dynamic moduli (Abbas et al., 2018a, 2018b; Asef and Farrokhrouz, 2017). In this study, we attempt to investigate the applicability of dynamic elastic parameters in determining brittleness coefficients by juxtaposing three independent measures of brittleness. We examine these questions with a data set from the Horn River Group, an important shale gas resource in the Horn River Basin of northeastern British Columbia, Canada (Fig. 1). Previous studies have examined sedimentological and geochemical characteristics of the Horn River Group (e.g., Ayranci et al., 2018a; Ayranci et al., 2018b; Dong et al., 2018a) and related the depositional character to petrophysical properties (Dong et al., 2015, 2017a) and dynamic geomechanical properties (Dong et al., 2017b). In this study, three independent brittleness coefficients are calculated for the Horn River Group:

Fig. 2). No simple parameter indicates whether a rock will fail in a brittle or ductile manner. There are, however, several definitions for rock brittleness available in the literature that can be used to assess the relative brittleness of rocks (Rybacki et al., 2016; Gholami et al., 2016; Rahimzadeh Kivi et al., 2018). Most commonly, rock elastic parameters such as Young's modulus and Poisson's ratio are used to calculate a measure of rock brittleness (Holt et al., 2015; Rickman et al., 2008), where higher Young's modulus and lower Poisson's ratios are correlated to more brittle behavior in rocks. Brittleness coefficients can also be based on rock composition. Rocks composed largely of more brittle minerals such as quartz and carbonates are observed to be more brittle than rocks that contain significant amounts of soft minerals such as clays (Jarvie et al., 2007; Wang and Gale, 2009; Dong et al. 2017b, 2018a; Glorioso and Rattia, 2012). Indentation tests on core surfaces can also be used to obtain measures of rock hardness (Yagiz, 2009; Dong et al., 2017b). These methods have wide applications in mining and drilling sciences, although calibration of the measurements to laboratory or field data is required to obtain reliable results. Finally, another class of brittleness coefficients are based on rock deformation characteristics. These analyses typically use stress-strain relationships during triaxial experiments, focusing on pre-failure region, post-failure region, or a combination of both, to determine the extent of brittle failure (Rybacki et al., 2016). Arguably, these methods should provide the most accurate calculations of brittleness. However, they often require expensive laboratory experiments which can be a deterrent to their use. Extensive laboratory testing shows that brittleness is also dependent on stress and temperature conditions (Rybacki et al., 2015; Rybacki et al., 2016; Rahimzadeh Kivi el al., 2018). Rybacki et al. (2015) demonstrated that elevated temperature and confining stress lead to more ductile behavior in rocks. They concluded that at confining stresses and temperatures corresponding to depths shallower than 4 km, conditions are suitable for brittle or semi-brittle failure. Therefore, under typical reservoir conditions (depths shallower than 4 km), differences in intrinsic rock properties are the determining factor in evaluating relative brittleness. The simpler proxies and correlations that are typically used

• Uniaxial Compressive Strength (UCS) tests on core samples are used

• •

to obtain static Young's modulus and Poisson's ratio. These are compared to dynamic parameters derived for the same samples from compressional and shear wave velocities, enabling us to relate static and dynamic elastic parameters. The static parameters are used to calculate an elasticity-based brittleness coefficient. The electron capture spectroscopy (ECS) log is used to calculate a composition-based brittleness coefficient. An indentation tool is used on a core to obtain rock hardness as a measure of brittleness.

Additionally, the relationship between brittleness, lithofacies, and mineralogy is also investigated. These relationships make it possible to extend the results of this work more regionally in the Horn River Group and possibly other shale formations. 2. Background - nature of rock deformation Rocks dilate during brittle or semi-brittle failure, which indicates creation of fracture volume. On the other hand, rock volume shrinks in ductile failure, and no distinct open fractures form (Hajiabdolmajid and Kaiser, 2003; Tarasov and Potvin, 2013; Hoek and Martin, 2014, 248

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 2. On the left is a typical stress-strain relationship and volumetric strain for a fully ductile failure. On the right is an example of the stress-strain relationship and volumetric strain for an ideal brittle failure.

in literature such as elasticity-based methods, composition-based methods, or indentation testing do not take into account the effect of temperature and confining stress but can be successfully used to obtain relative brittleness in a specific reservoir at a given temperature and stress condition (Rybacki et al., 2016; Jin et al., 2015; Gholami et al., 2016). Brittleness describes the relative ability of a rock to fail in a brittle or semi-brittle manner. When rocks with high brittleness values fail, they tend to form complex fractures (Hajiabdolmajid and Kaiser, 2003; Tarasov and Potvin, 2013; Andreev, 1995). The brittleness coefficient however, does not indicate whether the rock will fracture easily. For example, it has been reported that typically highly brittle rocks such as dolomite can act as fracture barriers, as they are stronger than shales and require higher pressures to fracture (Daniel et al., 2007; Bruner and Smosna, 2011; Jin et al., 2015). Fracture toughness is defined as the ability of rock to resist fracture growth within it. Rocks with low fracture toughness require less energy to allow fracture growth within them. Therefore, ideal candidates for hydraulic fracturing are rocks with high brittleness and low fracture toughness (Bai, 2016; Yuan et al., 2017). Fracture toughness is typically measured in the lab, however several correlations also exist in the literature that relate that to Young's modulus or compressional wave velocity (Jin et al., 2015; Whittaker et al., 1992; Yuan et al., 2017).

Fig. 3. The stratigraphy of the Horn River Basin (Ferri et al., 2011).

and ichnological characterization of the Horn River Group is provided by Ayranci et al. (2018a). This study is conducted on a long core from the well EOG Tattoo DA28-F/094-O-10, collected from the central part of the Horn River Basin (Fig. 1). The cored depth is 3083.2–3316.5 m with minimal length of missing sections (over 200 m core recovery). All three stratigraphic units of Horn River Group are represented in the core. 4. Methodology

3. Geological settings

4.1. Core analysis

The Horn River Group is comprised of three main units, in an ascending order: the Evie and Otter Park Members and the Muskwa Formation (Ferri et al., 2011; Ayranci et al., 2018a, b). The Horn River Group is approximately 200–500 m thick and is estimated to contain 78 Tcf of recoverable gas reserves (Rivard et al., 2014; BC Ministry of Energy and Mines, National Energy Board, 2011). The location and stratigraphy of the Horn River Group are presented in Figs. 1 and 3, respectively. The Evie Member on average has the highest TOC content in the Horn River Group (Dong et al., 2018a) and is also rich in carbonate minerals. The Otter Park Member overlies the Evie Member and typically contains less TOC, and is rich in clay. The Muskawa Formation is typically rich in quartz with moderate amounts of TOC and low clay content (Dong et al., 2018a). The present-day depth to the top of Muskwa Formation ranges from 2175 m to 3000 m (Al-Zahrani, 2011). The Horn River Group is considered to be in the dry gas window (high thermal maturity; Ross and Bustin, 2008). A detailed sedimentological

To understand the relationship between brittleness and lithofacies, we logged the core at high resolution, noting lithology, grain size, physical and biogenic sedimentary structures and recording features as fine as 1 cm. Two hundred fifty samples distributed along the length of the core, each approximately 10 cm in length, were analyzed for TOC content using LECO combustion by GeoMark Research Ltd. and for major, minor and trace elements by Acme Analytical Laboratories, following procedures detailed in Dong et al. (2018a). The mineralogy along the core was obtained from the Elemental Capture Spectroscopy Sonde (ECS) log performed by Schlumberger. A depth shift of 6 m was identified between core and log depths. 4.2. Laboratory experiments Cylindrical specimens for UCS tests were taken from the Horn River 249

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

core using a Kao Ming KMR-700DS hydraulic press. The shale specimens were extracted using a hollow 1” (25.4 mm) steel core barrel rotating at an 88 rev/min (0.264 in/min). Tap water was used as coring fluid to cool the bit during the operation. The upper half of the Evie member in the core was heavily fractured, which made it difficult to take samples large enough to conduct UCS tests. Therefore, fewer samples are tested from the Evie Member. Seventeen samples were successfully obtained for UCS tests. The bedding planes in all samples are parallel to the axis of the cylinders. Attempts to obtain samples with bedding planes perpendicular to the axis of the cylinders, were largely unsuccessful due to core disking. The samples tended to break easily along the bedding planes when cored perpendicular to bedding planes. This issue prevented us from obtaining samples with sufficient length to satisfy the ASTM D-7012 standards for sample's diameter/length ratio. UCS specimens were trimmed to the correct diameter/length ratio, 1” (25.4 mm) in diameter/2” (50.8 mm) in length using a CAT No. 117Mico Instruments Specimen saw. The specimens were preserved in plastic wrap and stored in a sealed container until required for testing purposes. The samples were dried in an oven at 90 °C for at least 48 h prior to testing. Prior to performing the UCS tests, measurements of compressional (P-wave) and shear (S-wave) wave velocities were made on intact specimens. The cylindrical specimens were placed between two metallic endcaps equipped with piezometric crystals. P-wave and S-wave signals were generated by the action of a JSR Ultrasonics DPR300 Pulser/Receiver power source (50 kHz) and processed using an Agilent Technologies DS06014A oscilloscope. The signal was further processed using MATLAB software package to clearly identify both P and S waves’ arrival times. To measure the axial and radial strains during the UCS tests, six OMEGA KFH-Series Linear strain gauges were glued to the body of the shale specimens. Three gauges were installed as axial strain gauges and the other three as lateral strain gauges. The axial/radial strain was calculated from the change in voltage induced by the deformation of the surface of the specimen during the test. The Horn River shale specimens were loaded to failure using an INSTRON 3384-150 kN loading system at a strain rate of 0.12 mm/min. Stress and strain values were recorded during loading for each sample. Young's modulus was calculated as the slope of the best-fit line in the stress-axial strain plot at a load range between 40% and 60% of the peak stress. Poisson's ratio was calculated by dividing the value of Young's modulus by the slope of the best-fit line in the stress-radial strain plot, in the same load range.

BIelastic =

1 E Emin + 2 Emax Emin

max min

max

Eq. (1) shows the brittleness index proposed by Rickman et al. (2008), based on elastic properties. Emin and Emax are the minimum and maximum values of static Young's modulus encountered in the formation under study. Similarly, min and max are the minimum and maximum values of static Poisson's ratio. Brittleness index values in Eq. (1) range between 0 and 1, where 0 indicates ductile and 1 indicates brittle behavior. Eq. (1) has found widespread use due to its simplicity and data availability. Because this equation relies on static moduli, a relationship between dynamic and static elastic parameters must be established in order to calculate brittleness using Eq. (1). It should be noted that Eq. (1) is only suitable to calculate relative brittleness within one reservoir. Eq. (1) cannot be used to compare brittleness between different reservoirs, as temperature and loading conditions typically differ, and both Young's modulus and Poisson's ratio are normalized to the specific formation. Other brittleness equations based on elastic parameters (Rybacki et al., 2016) yield relative values similar to Eq. (1). Due to its widespread use, we applied Rickman's equation to calculate the elasticity-based brittleness index. Several composition-based brittleness coefficients are used in the literature (Jarvie et al., 2007; Wang and Gale, 2009; Glorioso and Rattia, 2012; Rybacki et al., 2016). We propose a modified brittleness criterion based on the equation presented by Rybacki et al. (2016), the most complete form of composition-based brittleness.

BIcomp =

wQFP (Qtz + Fsp + Py ) + wCb Cb wQFP (Qtz + Fsp + Py ) + wCb Cb + wCly (Cly + TOC ) + w 2

We include carbonate content in the numerator of Eq. (2) as a brittle component. In Eq. (2), wQFP , wCb , wCly , and w are weighting factors. We used the value of 1 for all weighting factors except wCb , which was set at 0.5. Quartz, feldspar, pyrite, and carbonate volume fractions were used in the numerator as brittle components. The denominator includes the numerator plus volume fractions of clay, TOC (total organic carbon), and porosity ( ). The fraction of minerals along the core is determined from ECS log, while TOC and porosity were determined from LECO combustion tests and neutron/density logs, respectively. In addition to the brittleness indices introduced in Eq. (1) and Eq. (2), we used a commercially available hardness tester called Equotip Bambino 2, to determine hardness of the core under study. This handheld tool drops a hard metal ball on the face of the core and measures the impact and rebound velocity of the ball as it hits the rock surface. The hardness value is the ratio of the rebound to impact velocity multiplied by 1000 and typically ranges between 200 and 900, where higher hardness values indicate a stronger material. At each depth, we made 3 measurements and used their average as the hardness value for that point. Hardness is measured every 20 cm of the entire length of the core. Table 1 provides a summary of methodologies used for calculating each of the three brittleness measurements in this work.

4.3. Brittleness coefficients Rickman et al. (2008) proposed an averaging method based on elastic parameters that has been consistently used throughout the literature. This method assumes that a higher Young's modulus and lower Poisson's ratio lead to a more brittle rock.

Table 1 A summary of data sources and equations for each brittleness measurement method used in this study. Brittleness Measurement Method Indentation test (hardness) Composition-based method

Elasticity-based method

Sources of Data/Equations used -

1

Measurements taken using Equotip Bambino 2 every 20 cm along the core. Rock composition along the core taken from ECS logs Density and neutron logs are used to estimate porosity values along the core TOC values are measured using LECO Combustion method Eq. (2) is used to calculate brittleness Sonic and density logs are used along with Eq. (3) and Eq. (4) to calculate dynamic Young's modulus and Poisson's ratio Eq. (5) and Eq. (6) are used to convert dynamic elastic properties to static values Static Young's modulus and Poisson's ratio are used in Eq. (1) to calculate brittleness

250

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

5. Results

interval of clay-rich laminated mudstone in the middle of the formation. The Muskwa is largely characterized by high quartz and low clay content and moderate amounts of TOC, except in the laminated mudstone interval. The Otter Park Member is dominated by the laminated mudstone facies, with interbedded carbonate-rich layers. It shows consistently higher clay content, locally as high as 60%, lower quartz content than in the Muskwa Formation, and the lowest TOC content in the Horn River Group. The Evie Member is dominated by pyritic mudstones in the upper half of the unit and massive mudstones in the lower half. The transition from the Evie Member to the overlying Otter Park Member marks a sharp increase in clay content and drop in quartz content. TOC values in Evie are highest in the Horn River Group, particularly in the lower half.

5.1. Core description Ten lithofacies have been documented in the Horn River Group representing various depositional conditions (Ayranci et al., 2018a; Dong et al., 2017a). Seven of these lithofacies were encountered in the EOG Tattoo core and grouped into three based on abundance and similarities in terms of depositional condition. These groups are: massive mudstones, pyritic mudstones, and laminated mudstones. The massive mudstone lithofacies is dark-grey and occurs mainly in Muskwa Formation and Evie Member. It is relatively abundant in TOC and quartz and locally contains thin carbonate-rich mudstone layers, particularly in the Evie Member (Dong et al., 2018a). The pyritic mudstone lithofacies is similar to massive mudstones with common pyrite-rich laminae and lenses. It is also quartz-rich but contains on average somewhat less TOC content. Pyritic mudstones are also commonplace in Muskwa Formation and Evie Member. The laminated mudstone lithofacies is light to dark-grey and is most commonly encountered in the Otter Park Member. Among the Horn River lithofacies, this lithofacies shows typically higher clay content and lower TOC content. Fig. 4 presents the core description, mineralogy, and TOC for the Tattoo core. The Muskwa Formation is largely composed of pyritic mudstones and to a lesser extend massive mudstones, with a thick

5.2. Experimental results Results of the UCS experiments are presented in Table 2. The UCS strength is a measure of the peak stress at failure. UCS values range from 20 to 125 MPa and are systematically high in the Muskwa Formation and low in the Otter Park Member. Only two Evie samples were analyzed; these are comparable to the Otter Park samples. Static Young's modulus values range between 20 and 38 GPa. On average, Muskwa Formation shows an average Young's modulus of 32 GPa, while the average modulus in Otter Park and Evie Member is 28 and 33 GPA, respectively. The Otter Park Member on average exhibits

Fig. 4. Profile of minerals (ECS log - fraction) is provided on the left, core description in the middle, and TOC content (LECO combustion – wt%) is presented on the right. 251

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Table 2 Summary of UCS results. Compressional and shear wave velocities are also included along with dynamic elastic parameters. Sample Number

Depth

Formation

Vp

Vs

Static Young's Modulus

Static Poisson's Ratio

Dynamic Young's Modulus

Dynamic Poisson's Ratio

UCS Strength



m



m/s

m/s

GPa



GPa



MPa

3097.5 3098.5 3118 3120.25 3131 3155.5 3158 3159.25 3183.25 3190 3190.75 3220.75 3223.75 3225.25 3282.25 3284

Muskwa Muskwa Muskwa Muskwa Muskwa Muskwa Muskwa Otter Park Otter Park Otter Park Otter Park Otter Park Otter Park Otter Park Evie Evie

4798 5424 5371 5055 4980 5535 5007 5342 5485 5287 5342 5237 5896 6018 4691 6037

2987 2636 2611 2572 2309 2746 2486 2839 2463 2335 2012 3033 2160 2203 2688 2804

34 34 34 31 27 34 31 37 38 25 20 34 20 22 35 31

0.34 0.13 0.24 0.37 0.28 0.22 0.19 0.38 0.32 0.36 0.17 0.28 0.23 0.25 0.28 0.4

53 48 46 42 37 49 39 53 43 38 30 59 34 35 47 56

0.18 0.35 0.35 0.33 0.36 0.34 0.34 0.30 0.37 0.38 0.42 0.25 0.42 0.42 0.26 0.36

75 125 115 63 54 60 23 31 59 37 48 49 20 20 34 26

1 2 3 4 5 6 7 8 9 11 12 13 14 15 16 17

lower Young's modulus values, however carbonate-rich mudstone layers in Otter Park Member could exhibit high stiffness. Static Poisson's ratio ranges between 0.13 and 0.4 in our samples. The Muskwa Formation exhibits an average Poisson's ratio of 0.25; the Otter Park and Evie Member show an average of 0.28 and 0.34, respectively. Compressional (Vp) and shear (Vs) wave velocity measurements for all samples are also presented in Table 2. Dynamic Young's modulus and Poisson's ratio are calculated using the following equations (Fjaer et al., 2008).

Ed =

d

=

Vs2 (3Vp2 Vp2 Vp2 2(Vp2

however, there is a good correlation between the static and dynamic Young's modulus presented as Eq. (5) (R2 of 0.7).

Es = 1.1149 Ed 0.8686

Eq. (5) shows the relationship between static and dynamic Young's modulus under laboratory conditions, i.e. no confining pressure, dry cores, and room temperature. Assuming Eq. (5) still holds under reservoir conditions, we can use it to estimate in-situ static Young's modulus using dipole sonic and density logs. The applicability of dynamic moduli measurements on dry samples to in-situ conditions may not be strictly valid. This is in part due to presence of water in pores under reservoir conditions. Therefore, high porosity rocks can exhibit higher in-situ dynamic moduli due to water saturation, in the context of poroelasticity. Chang et al. (2006) provides a discussion on the applicability of empirical relationships between dynamic and static parameters. They conclude that for rocks with porosity of less than 25%, the influence of water saturation on dynamic moduli is negligible. Considering the low-porosity nature of Horn River Group, it is an acceptable assumption that the dynamic measurements in the lab and insitu are comparable. Static and dynamic Poisson's ratio show no correlation. However, we found that on average static Poisson's ratio is slightly (22%) lower than the dynamic ratio, albeit with a large scatter. We used the following relationship to estimate static Poisson's ratio from sonic and density logs, based on data presented in Fig. 5.

4Vs2) 2Vs2

3

2Vs2 Vs2 )

5

4

Dynamic Young's moduli in the samples range between 30 and 59 GPa. The average Young's modulus for the Muskwa Formation is 45 GPa, while the Otter Park Member and Evie Member show an average of 42, and 51 GPa, respectively. Similar to static measurements, dynamic data on average indicate lower stiffness values for the Otter Park Member. Dynamic Poisson's ratio ranges between 0.26 and 0.42, averaging 0.32, 0.37, and 0.31 for Muskwa Formation, Otter Park, and Evie Member, respectively. Fig. 5 presents the relationship between static and dynamic Young's modulus (on the left) and Poisson's ratio (on the right). Both dynamic and static elastic moduli are determined from lab results. The static Young's modulus values are consistently lower than the dynamic values;

s

= 0.77 ×

d

6

Fig. 5. Relationship between static and dynamic Young's modulus (left) and static and dynamic Poisson's ratio (right) is illustrated. Static measurements are derived from UCS tests, and the dynamic data are calculated from wave velocity measurements on the same samples. 252

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 6. Profiles of three independent measures of brittleness with depth. The mineralogy and lithofacies are also included. In the lithofacies ribbon, the blue dotted area corresponds to pyritic mudstones lithofacies, the black area corresponds to massive mudstones lithofacies, and the yellow area with dashed lines corresponds to laminated mudstones lithofacies.

5.3. Brittleness coefficients

content decreases in the lower part of the Otter Park, reaching a minimum at the base of the unit (Fig. 6). All three coefficients rebound at the boundary of the Evie member, where there is a sharp decline in clay content. The brittleness coefficients show high relative values throughout Evie, however on average upper Evie shows higher brittleness than lower Evie. The three measures of brittleness used in this work yield consistent results (Fig. 6), indicating that all three methods reliably identify relative brittleness within a formation. In this analysis, Evie Member has the highest relative brittleness, followed closely by Muskwa Formation. The Otter Park Member shows the lowest relative brittleness.

Three independent measures of brittleness are reported in this study, based on elastic properties, rock composition, and indentation characteristic of the rock. Dynamic Young's modulus and Poisson's ratio can be calculated from sonic and density logs (Eq. (3) and Eq. (4)). Static elastic parameters are estimated using the established correlations in this work (Eq. (5) and Eq. (6)). Elasticity-based brittleness can be calculated by incorporating static elastic parameters into Eq. (1). In this work, we used 46 GPa and 15 GPa as Emax and Emin , respectively. Additionally, a value of 0.13 was assigned to min , and 0.4 to max . The limits for elastic parameters were taken from the maximum and minimum values of static elastic parameters that were calculated for the entire length of the core using Eq. (5) and Eq. (6). Eq. (2) is used to calculate composition-based brittleness. Rock composition along the core was taken from ECS logs, while TOC values were obtained from LECO analysis of core plugs. Fig. 6 illustrates variation of the three brittleness coefficients with depth. The mineralogy and lithofacies distributions are also presented. All three brittleness coefficients show similar trends. At the top of Muskwa Formation, all three brittleness coefficients show relatively high values. The brittleness values plunge in the middle of Muskwa, corresponding to the thick interval of the clay-rich laminated mudstones lithofacies. Brittleness coefficients rise again in the lower part of the Muskwa Formation. In the Otter Park Member, brittleness values decrease with depth, as clay content increases. In the middle Otter Park Member, all brittleness values consistently spike in a carbonate-rich laminated mudstone layer. Brittleness decreases again when clay content increases and carbonate

6. Discussions 6.1. Dynamic versus static elastic parameters Sonic and density logs are commonly used to calculate dynamic Young's modulus and Poisson's ratio using Eq. (3) and Eq. (4). Dynamic Young's modulus and Poisson's ratio differ from the static elastic parameters that correlate to a static loading, which is a better representation of the loads encountered during hydraulic fracturing. Geomechanical analyses are typically based on static loads. Since the dynamic elastic parameters are more readily available through logs, it is beneficial to find correlations between dynamic and static mechanical properties. Such correlation facilitates the estimation of static Young's modulus and Poisson's ratio to calculate rock brittleness using logs. This practice has been successfully implemented for sandstones, carbonates, and a few shale formations (Najibi et al., 2015; Abbas et al., 2018a). 253

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 7. Comparison of dynamic and static Young's modulus (GPa) for the core. Brittleness values based on both dynamic and static elastic parameters are also provided.

Elasticity-based brittleness (Eq. (1)) can be also calculated using dynamic elastic parameters. For this purpose, the maximum and minimum bounds (Emax , Emin, etc.) for elastic parameters in Eq. (1) need to be adjusted for the range of dynamic values. Fig. 7 shows elasticitybased brittleness values based on static and dynamic data. The values of brittleness using either dynamic or static data are essentially identical. This is due to the fact that the static and dynamic data follow a similar trend, and the data is normalized based on maximum and minimum values in Eq. (1). Therefore, dynamic elastic parameters can be reliably used to calculate relative brittleness in a formation. Fig. 7 also presents the dynamic and static Young's modulus along the core. The dynamic parameter is calculated using sonic and density logs (Eq. (3) and Eq. (4)). The static Young's modulus is calculated using the correlations developed in this work (Eq. (5) and Eq. (6)). While the trend for both parameters are similar, the static Young's modulus is on average 30% lower than the dynamic modulus. Therefore, using dynamic elastic parameters to model hydraulic fracturing or wellbore stability can lead to large errors in calculations (fracture apertures, borehole movements, stress shadows, etc.), as the loads in these problems are static. While we identified a clear relationship between dynamic and static Young's modulus, no clear trend between dynamic and static Poisson's ratio was observed in our tests. Static and dynamic Poisson's ratio are generally considered to be equal in literature and no clear relationship has been established (Rickman et al., 2008; Archer and Rasouli, 2012;

Fei et al., 2016). Our results confirm the observations reported in the literature. However, we found that static Poisson's ratios measured in this work are on average 22% lower than the dynamic ratios. Therefore, we chose to use Eq. (6) to calculate static Poisson's ratio instead of assuming it to be equal to the dynamic value. Several correlations are proposed in the literature to relate dynamic and static elastic parameters (Najibi et al., 2015; Asef and Farrokhrouz, 2017; Abbas et al., 2018a). However, the applicability of these equations is usually limited to the formation for which they were developed. It is imperative to establish this relationship for the formation of interest to model subsurface rock deformations. In this work, we established a relationship between dynamic and static Young's modulus and Poisson's ratio for the Horn River Group. If the relationships between dynamic and static elastic parameters for a formation of interest are unknown, some of the correlations published in the literature can be applied with caution. Care must be taken that the correlations are developed for a similar type of rock with elastic properties in the same range as the formation of interest. Fig. 8 compares three correlations from the literature to the experimental results reported in this work. While all three correlations overestimate static Young's modulus, the correlation proposed by Asef and Farroukhrouz (2017) produces reasonable results. Table 3 provides a summary of the correlations used in Fig. 8.

254

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

between them arise from consideration of what constitutes brittle or ductile components. Jarvie et al. (2007), consider only quartz as brittle component, while Wang and Gale (2009) identify quartz and dolomite as hard components. Glorioso and Rattia (2012) include all carbonate minerals as brittle components to calculate the composition-based brittleness. After an extensive experimental campaign, Rybacki et al. (2016) concluded that the inclusion of carbonate content as part of brittle component overestimates the brittleness of carbonate-rich rocks. Similar behavior was observed in the Horn River Group when the weighting factor of carbonate minerals in Eq. (2) was increased to 1. The carbonate-rich layers in that case show a significantly higher composition-based brittleness but lower elastic-based brittleness and hardness values, indicative of an over-estimation of composition-based brittleness. However, complete exclusion of carbonates as brittle components in this dataset significantly underestimates the brittleness index. Using a weighting of wCb = 0.5 in Eq. (2) alleviates the problem of overestimation of brittleness. It is also theoretically sound, as carbonates have less strength than quartz and therefore their influence on brittleness should be weighed less (Rybacki et al., 2015). Differences in mineralogy and lithofacies appear to control brittleness trends (Figs. 6 and 9). Fig. 10 relates clay and quartz content to brittleness values, demonstrating a clear inverse relationship between clay content and brittleness. Region A in Fig. 10 represents high quartz and low clay content samples, primarily associated with pyritic mudstones lithofacies (average pyrite content is less than 4%). These are the most brittle samples observed in this core. Region B signifies a lower quartz content than region A but also fairly low clay content (higher in TOC and carbonate contents). These samples are typically massive mudstones and are moderately high in relative brittleness. Region C shows samples with high clay content and lower quartz content compared to regions A and B, associated with the laminated mudstones, which has the lowest brittleness values encountered in the core. Fig. 10 shows that the clay content is the dominant factor influencing brittleness values in the Horn River Group, consistent with other studies of the Devonian Woodford Shale (Aoudia et al., 2010), the Horn River shale (Dong et al., 2017a), and Duvernay Formation (Dong et al., 2018b), while quartz content plays a secondary role. The impact of quartz content on brittleness depends on the type of quartz present in the rock (biogenic versus detrital). A comprehensive discussion on types of quartz in Horn River Group is provided by Dong et al. (2017b). Brittleness coefficients measured in this work did not correlate to any other minerals, including pyrite, carbonate and TOC (Moghadam et al., 2018; see “Data for: Brittleness in the Devonian Horn River Shale, British Columbia, Canada”).

Fig. 8. Relationship between elasticity-based and composition-based brittleness indices. The data is further grouped according to the dominant lithofacies observed in the core. Table 3 Summary of the correlations used in Fig. 8. Es and Ed are static and dynamic Young's modulus in GPa, and is porosity in fraction. No.

Equation

Author

1

Es = 0.018Ed2 + 0.422Ed Es = 1.153Ed 15.2 Es = 0.88Ed (1 ) 3.7

Lacy (1997)

2 3

Nur and Wang (1999) Asef and Farrokhrouz (2017)

6.2. Mineralogical and sedimentological controls on brittleness We evaluated three independent brittleness coefficients for the Horn River Group shale and documented similar stratigraphic trends (Fig. 6). Fig. 9 shows the relationship between the elasticity-based and composition-based brittleness. A clear positive trend is identifiable. In order to observe the influence of lithofacies on brittleness, the data is grouped into pyritic, massive, and laminated mudstones facies. The pyritic facies on average show the highest brittleness values, followed by massive mudstones and laminated mudstones lithofacies. The several composition-based brittleness indices proposed in the literature (Jarvie et al., 2007; Wang and Gale, 2009; Glorioso and Rattia, 2012; Rybacki et al., 2016) all account for the ratio of hard/ brittle minerals to brittle plus ductile minerals. The slight differences

6.3. Fracability index Yuan et al. (2017) developed a fracability index (FI) that incorporates brittleness, fracture toughness, and gradient of minimum insitu horizontal stress. Gamma ray, density, and sonic logs are used to estimate Mode-I (tensile) and Mode-II (shear) fracture toughness. Eq. (7) and Eq. (8), shown below, outline the methodology to calculate Mode-I and Mode-II fracture toughness using log data,

KIC = 0.45 KIIC = 2.121

0.151 exp(Vcl ) + 0.201 ln(DT ) 0.245 exp(Vcl ) + 1.152 ln(DT )

0.877 8.378

7 8

where KIC and KIIC are Mode-I and Mode-II fracture toughness in MPa.m0.5, is the formation density in gr/cm3 from density log, DT is the acoustic travel time in μs/ft from sonic log, Vcl is the clay content of the rock in fraction. Yuan et al. (2017) used a gamma ray log to estimate Vcl . For the Horn River Group however, gamma ray is more correlated to TOC rather than clay content (Ayranci et al., 2018b), typical of Devonian shales (for example, Hemmesch et al., 2014). Therefore, we used clay content from ECS log to calculate Mode-I and Mode-II fracture toughness values.

Fig. 9. Relationship between elasticity-based and composition-based brittleness indices. The data is further grouped according to the dominant lithofacies observed in the core. 255

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 10. Scatter plot of quartz versus clay content for the core under study. The points are color coded based on brittleness values. Area A contains points from pyritic mudstones lithofacies, while area B and C refer to massive mudstones and laminated mudstones lithofacies, respectively.

Fig. 11. Profile of the fracability index (FI) and fracture toughness (MPa.m0.5) versus depth calculated using Eq. (7), Eq. (8), and Eq. (9). Hardness measurements and mineralogy are also provided. Region A is an example of a low FI zone and B is an example of a high FI zone.

Once Mode-I and Mode-II fracture toughness values are estimated, Eq. (9) can be used to calculate the fracability index.

FI = BI ×

2 1 × KIC + KIIC h

similar results. We used Schlumberger's in-situ stress analysis for h values across the formation. These data are publicly available through British Columbia Oil and Gas Commission. Fig. 11 presents the fracability index and fracture toughness profiles. The hardness values and mineralogy are also plotted. The fracability index in the core follows the same trend as the brittleness values. The results show a high fracability index in Muskwa formation with the exception of the clay-rich layer in the middle of Muskwa. Otter Park member shows a low fracability index throughout, while the Evie member presents the best fracturing target.

9

FI is the fracability index in MPa2.m0.5. KIC and KIIC are Mode-I and Mode-II fracture toughness in MPa.m0.5. h is the gradient of minimum horizontal stress in MPa/100m. BI is the brittleness index. We used the composition-based brittleness coefficient as BI in Eq. (9), but elasticitybased brittleness or normalized hardness values in Eq. (9) would yield 256

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Fig. 12. Pictures of the Horn River core corresponding to regions A and B in Fig. 11. On the left, picture of zone A corresponds to a low FI zone. On the right, picture of zone B corresponds to a high FI zone.

The region marked as A in Fig. 11 is chosen as an example of a low FI zone, while region B is chosen an example of a high FI zone. Fig. 12 presents two images of the Horn River core, corresponding to regions A and B. On the left, an image of zone A provides an example of a low FI region in the core. Twenty fractures were identified in zone A and marked by blue lines. On the right, an image of zone B with higher FI values is provided. 52 fractures were identified in zone B. The high FI zones in the core consistently show higher number of fractures compared to the lower FI zones such as in Otter Park. This is further evidence that regions with high FI or hardness values, are more susceptible to fracturing. Yang et al. (2018) analyzed a long core from Horn River shale to investigate the number of natural fractures in the formation with respect to the hardness measurements. Hardness values were measured with a tool similar to the one used in this study. They concluded that approximately 80% of natural fractures encountered in the Horn River core were in regions with hardness values above 600. Their results provide additional confirmation that fracability index, hardness, and brittleness measurements used in this work can be successfully implemented to identify targets for hydraulic fracturing operations.

laboratory experiments or appropriate correlations.

• For Horn River Group, the pyritic mudstones lithofacies have the •

Acknowledgements We thank British Columbia Oil and Gas Commission for access to cores. We are grateful for the funding support from Natural Sciences and Engineering Research Council of Canada (grant CRDPJ 445064–12) and co-funders ConocoPhillips Canada, Devon Canada, Husky Energy, Imperial, Nexen-CNOOC, and Shell Canada. Appendix A. Supplementary data Supplementary data related to this article can be found at https:// doi.org/10.1016/j.jngse.2018.12.012.

7. Conclusions The following conclusions are drawn from this work:

References

• Using UCS experiments •

• •

highest brittleness, followed by the massive mudstones lithofacies. The laminated mudstones lithofacies common in the Otter Park member, shows the lowest brittleness. Clay content plays the most dominant role in determining high brittleness regions. Quartz content plays a secondary role. Regions with clay content below 20% and quartz content above 60% are the most suitable fracturing targets in Horn River Group.

and wave velocity measurements, a relationship between dynamic and static elastic parameters is proposed for Horn River Group (Eq. (5) and Eq. (6)). Based on three independent measures of brittleness, and a fracability model, the Evie Member is identified as the most suitable target for hydraulic fracturing, followed by the Muskwa Formation. The Otter Park Member has considerably lower brittleness values. In the core studied in this work, there is a clay-rich layer in the middle of Muskwa Formation that has low brittleness. Well logs and core hardness measurements can be used reliably to determine relative brittleness in shale formations. Dynamic elastic parameters from well logs produce nearly identical values for elasticity-based brittleness when compared to static elastic parameters. Static Young's modulus was found to be 30% lower than the dynamic Young's modulus from well logs. Therefore, for modelling rock deformation, static values need to be obtained through

Abbas, A.K., Flori, R.E., Alsaba, M., 2018a. Estimating rock mechanical properties of the Zubair shale formation using a sonic wireline log and core analysis. J. Nat. Gas Sci. Eng. 53, 359–369. Abbas, A.K., Flori, R.E., Alsaba, M., Dahm, H., Alkamil, E.H.K., 2018b. Integrated approach using core analysis and wireline measurement to estimate rock mechanical properties of the Zubair Reservoir, Southern Iraq. J. Petrol. Sci. Eng. 166, 406–419. Al-Zahrani, A.A., 2011. Interpretation of 3D Multicomponent Seismic Data for Investigating Natural Fractures in the Horn River Basin, Northeast British Columbia. Master’s thesis. University of Calgary, Calgary, Canada. Andreev, G.E., 1995. Brittle Failure of Rock Materials: Test Results and Constitutive Models. A. A. Balkema, Rotterdam, pp. 446. Archer, S., Rasouli, V., 2012. A log based analysis to estimate mechanical properties and in-situ stresses in a shale gas well in North Perth Basin. In: 1st International Conference on Petroleum and Mineral Resources. Curtin Research Publications. Asef, M.R., Farrokhrouz, M., 2017. A semi-empirical relation between static and dynamic elastic modulus. J. Petrol. Sci. Eng. 157, 359–363. Aoudia, K., Miskimins, J.L., Harris, N.B., Mnich, C.A., 2010. Statistical analysis of the effects of mineralogy on rock mechanical properties of the Woodford shale and the associated impacts for hydraulic fracture treatment design. In: Paper Presented at 44th U.S. Rock Mechanics Symposium and 5th U.S.-Canada Rock Mechanics Symposium. American Rock Mechanics Association.

257

Journal of Natural Gas Science and Engineering 62 (2019) 247–258

A. Moghadam et al.

Hemmesch, N.T., Harris, N.B., Mnich, C.A., Selby, D., 2014. A sequence-stratigraphic framework for the upper devonian Woodford shale, Permian basin, west Texas. AAPG (Am. Assoc. Pet. Geol.) Bull. 98 (1), 23–47. Hoek, E., Martin, C.D., 2014. Fracture initiation and propagation in intact rock-a review. J. Rock Mech. Geotech. Eng. 6 (4), 287–300. Holt, R.M., Fjaer, E., Stenebraten, J.F., Nes, O., 2015. Brittleness of shales: relevance to borehole collapse and hydraulic fracturing. J. Petrol. Sci. Eng. 131, 200–209. Jarvie, D.M., Hill, R.J., Ruble, T.E., Pollastro, R.M., 2007. Unconventional shale-gas systems: the Mississippian Barnett shale of north-central Texas as one model for thermogenic shale-gas assessment. AAPG (Am. Assoc. Pet. Geol.) Bull. 91, 475–499. Jin, X., Shah, S.N., Roegiers, J., Zhang, B., 2015. An integrated petrophysics and geomechanics approach for fracability evaluation in shale reservoirs. SPE J. 20 (03), 518–526. Lacy, L.L., 1997. Dynamic rock mechanics testing for optimized fracture designs. In: Paper Presented at SPE Annual Technical Conference and Exhibition. Society of Petroleum Engineers. Moghadam, A., Harris, N.B., Ayranci, K., 2018. Data for: Brittleness in the Devonian Horn River Shale, British Columbia, Canada. Mendeley Data. https://data.mendeley.com/ datasets/r538bbwx95/draft?a=49ba3b14-c2eb-4e24-8634-5fd63972ec29. Najibi, A.R., Ghafoori, M., Lashkaripour, G.R., Asef, M.R., 2015. Empirical relations between strength and static and dynamic elastic properties of Asmari and Sarvak limestones, two main oil reservoirs in Iran. J. Petrol. Sci. Eng. 126, 78–82. Nur, A., Wang, Z., 1999. Seismic and Acoustic Velocities in Reservoir Rocks: Recent Developments, vol 10 Society of Exploration Geophysicists. Rahimzadeh Kivi, I., Ameri, M., Molladavoodi, H., 2018. Shale brittleness evaluation based on energy balance analysis of stress-strain curves. J. Petrol. Sci. Eng. 167, 1–19. Rickman, R., Mullen, M.J., Petre, J.E., Grieser, B., Kundert, D., 2008. A practical use of shale petrophysics for stimulation design optimization: all shale plays are not clones to Barnett shale. In: Paper Presented at SPE Annual Technical Conference and Exhibition, Society of Petroleum Engineers. Rivard, C., Lavoie, D., Lefebvre, R., Sejourne, S., Lamontange, C., Duchesne, M., 2014. An overview of Canadian shale gas production and environmental concerns. Int. J. Coal Geol. 126, 64–76. Ross, D.J.K., Bustin, R.M., 2008. Characterizing the shale gas resource potential of Devonian–Mississippian strata in the Western Canada sedimentary basin: application of an integrated formation evaluation. AAPG Bull. 92 (1), 87–125. Rybacki, E., Meier, T., Dresen, G., 2016. What controls the mechanical properties of shale rocks? – Part II: brittleness. J. Petrol. Sci. Eng. 144, 39–58. Rybacki, E., Reinicke, A., Meier, T., Makasi, M., Dresen, G., 2015. What controls the mechanical properties of shale rocks? – Part I: strength and Young's modulus. J. Petrol. Sci. Eng. 135, 702–722. Tarasov, B., Potvin, Y., 2013. Universal criteria for rock brittleness estimation under triaxial compression. Int. J. Rock Mech. Min. Sci. 59, 57–69. Wang, F.P., Gale, J.F.W., 2009. Screening criteria for shale-gas systems. Gulf Coast Assoc. Geol. Soc. Trans. 59, 779–793. Whittaker, B.N., Singh, R.N., Sun, G., 1992. Rock Fracture Mechanics Principles, Design and Applications (Developments in Geotechnical Engineering). Elsevier Science Ltd. Yagiz, S., 2009. Assessment of brittleness using rock strength and density with punch penetration test. Tunn. Undergr. Space Technol. 24, 66–74. Yang, S., Harris, N.B., Dong, T., Wu, W., Chen, Z., 2018. Natural Fractures and Mechanical Properties in a Horn River Shale Core from Well Logs and Hardness Measurements. SPE Reservoir Evaluation & Engineering Pre-print, May, 2018 (SPE174287-PA). Yuan, J., Zhou, J., Liu, S., Feng, Y., Deng, J., Xie, Q., Lu, Z., 2017. An improved fracability-evaluation method for shale reservoirs based on new fracture toughness-prediction models. SPE J. 22 (05), 1704–1713.

Ayranci, K., Harris, N.B., Dong, T., 2018a. Sedimentological and ichnological characterization of the middle to upper Devonian Horn River Group, British Columbia, Canada: insights into mudstone depositional conditions and processes below storm wave base. J. Sediment. Res. 88, 1–23. Ayranci, K., Harris, N.B., Dong, T., 2018b. High resolution sequence stratigraphic reconstruction of mud-dominated systems below storm wave base: a case study from the Middle to Upper Devonian Horn River Group, British Columbia, Canada. Sediment. Geol. 373, 239–253. Bai, M., 2016. Why are brittleness and fracability not equivalent in designing hydraulic fracturing in tight shale gas reservoirs. Petroleum 2 (1), 1–19. BC Ministry of Energy and Mines, National Energy Board, 2011. Ultimate Potential for Unconventional Natural Gas in Northeastern British Columbia's Horn River Basin. Oil and Gas Reports 2011-1. Burner, K.R., Smosna, R., 2011. A Comparative Study of the Mississippian Barnett Shale, Fort Worth Basin, and Devonian Marcellus Shale, Appalachian Basin. NETL (National Energy Technology Laboratory) Report, Department of Energy, US Federal Government. Chang, C., Zoback, M.D., Khaksar, A., 2006. Empirical relations between rock strength and physical properties in sedimentary rocks. J. Petrol. Sci. Eng. 51, 223–237. Daniel, M.J., Ronald, J.H., Tim, E.R., 2007. A comparative study of the mississippian barnett shale, fort worth basin, and devonian marcellus shale, appalachian basin. AAPG (Am. Assoc. Pet. Geol.) Bull. 91 (4), 475–499. Dong, T., Harris, N.B., Ayranci, K., 2018a. Sea level fluctuation and organic matter accumulation in Horn River Shale, Northeastern British Columbia, Canada: insights into sediment flux, redox conditions, and productivity from major and trace element geochemistry. Geol. Soc. Am. Bull. 130, 859–880. Dong, T., Harris, N.B., Ayranci, K., Twemlow, C.E., Nassichuk, B.R., 2015. Porosity characteristics of the Devonian Horn River shale, Canada: insights from lithofacies classification and shale composition. Int. J. Coal Geol. 141, 74–90. Dong, T., Harris, N.B., Ayranci, K., Twemlow, C.E., Nassichuk, B.R., 2017a. The impact of composition on pore throat size and permeability in high maturity shales: middle and Upper Devonian Horn River Group, northeastern British Columbia, Canada. Mar. Petrol. Geol. 81, 220–236. Dong, T., Harris, N.B., Ayranci, K., Yang, S., 2017b. The impact of rock composition on geomechanical properties of a shale formation: middle and Upper Devonian Horn River Group shale. Northeast British Columbia: AAPG (Am. Assoc. Pet. Geol.) Bull. 101 (2), 177–204. Dong, T., Harris, N.B., Knapp, L.J., McMillan, J.M., David, L., Bish, D.L., 2018b. The effect of thermal maturity on geomechanical properties in shale reservoirs: an example from the Upper Devonian Duvernay Formation, Western Canada Sedimentary Basin. Mar. Petrol. Geol. 97, 137–153. Fei, W., Huiyuan, B., Jun, Y., 2016. Correlation of dynamic and static elastic parameters of rock. Electron. J. Geotech. Eng. 21 (4), 1551–1560. Ferri, F., Hickin, A.S., Huntley, D.H., 2011. Besa River Formation, Western Liard Basin, British Columbia (NTS094N): Geochemistry and Regional; Correlations. British Columbia Ministry of Energy and Mines Geoscience Reports, pp. 1–18. Fjaer, E., Holt, R.M., Horsrud, P., Raaen, A.M., Risnes, R., 2008. Petroleum Related Rock Mechanics, second ed. Elsevier, Amsterdam, pp. 491. Gholami, R., Rasouli, V., Sarmadivaleh, M., Minaeian, V., Fakhari, N., 2016. Brittleness of shale gas reservoirs: a study from the north Perth basin, Australia. J. Nat. Gas Sci. Eng. 33, 1244–1259. Glorioso, J.C., Rattia, A., 2012. Unconventional reservoirs: basic petrophysical concepts for shale gas. In: Paper Presented at SPE/EAGE European Unconventional Resources Conference and Exhibition. Society of Petroleum Engineers. Hajiabdolmajid, V., Kaiser, P., 2003. Brittleness of rock and stability assessment in hard rock tunneling. Tunn. Undergr. Space Technol. 18 (1), 35–48.

258