The samples for the experiments were sourced from the Jurassic lacustrine shale in northeastern Sichuan. Because cores obtained from the field typically have irregular shapes, they require processing before being used in experiments. The sampling directions for the triaxial rock mechanics experiments were set to be longitudinal and transverse (Xuefeng et al., 2024; Biswas et al., 2024; Banerjee et al., 2024) (Figure 1B). In the course of conducting high-temperature and high-pressure triaxial experiments, samples were consistently obtained in the transverse direction to investigate the influence of confining pressure and temperature on rock mechanical properties. Furthermore, control experiments for rock mechanics were established in both longitudinal and horizontal orientations, considering the actual temperature and pressure conditions of the reservoirs in the study area and averaging conditions based on prior experiments. The experimental conditions for the rock mechanics tests were established as follows: Ying Mountain block (YS5 and YS8 wells): 20 MPa and 50°C; Pingchang block (PY1-2H, PY1-5H, and PA101 wells): 30 MPa and 70°C; Longgang block (DY1 well): 30 MPa and 70°C. The sampling directions for tensile experiments were set to be vertical and horizontal. Taking into account the coverage of the samples, differences in reservoir conditions, vertical segment distribution, and core collection status, along with the distribution of lithofacies, sedimentary structures, and fractures in the study area, 260 samples from nine different categories across six wells—PY1-2H, PY1-5H, PA101, DY1, YS5, and YS8—were collected and processed (Figure 1A). Figure1C is a histogram of the main lithologies in the area.

The mechanical parameters of rock cores are determined using a high-temperature and high-pressure triaxial testing apparatus, specifically the RTR-1000 Static (Dynamic) Triaxial Rock Mechanics Testing System from GCTS Corporation in the United States. This comprehensive setup consists of four main components: a high-temperature and high-pressure triaxial chamber, a confining pressure system, an axial pressure system, and an automatic data acquisition and control system. The apparatus complies with ASTM D2664-04 standards and the recommended methods for rock mechanics experiments by the International Society for Rock Mechanics and Rock Engineering (Qinghui et al., 2012a; Jarvie et al., 2007; Reedy et al., 2024). The testing system boasts a maximum axial load of 1,000 kN, a maximum confining pressure of 140 MPa, a pore pressure of 140 MPa, a dynamic frequency of 10 Hz, and a temperature of 150°C. The experimental control precision is as follows: pressure: 0.01 MPa; liquid density: 0.01 g/cm³; deformation: 0.001 mm. The testing is conducted at an ambient temperature ranging from 200°C to 230°C and a humidity of 64% RH, utilizing a lateral isobaric triaxial testing method (Li et al., 2024; Chakladar et al., 2024).

During the tensile strength testing, a 50 KN electronic servo system is employed to conduct a splitting test on standard specimens until the rock reaches its tensile limit and fractures. Corresponding parameters are then calculated. The testing procedures are strictly implemented according to the DZ/T 0276.21-2015 “Rock Tensile Strength Testing” standard. The experimental control precision for this test is: pressure of 0.05 MPa and deformation of 0.001 mm (Li et al., 2024).

For whole-rock XRD mineral analysis, a DX-2700 X-ray diffractometer is used to determine the mineral composition and clay content of reservoir rocks. Powder samples are thoroughly crushed using a sample crusher to a size of approximately 5 mm. Based on the diffraction data obtained from the testing, including diffraction curves, d-values, relative intensities, and diffraction peak widths, analysis software is utilized on a computer to conduct whole-rock and clay mineral composition analysis (Cao Feng et al., 2024).

The porosity of rocks (helium method) is calculated using a PHI-220 automatic porosity tester. This method exploits the principle that the diffusion rate of helium in pores is directly proportional to porosity. By measuring the diffusion rate of helium in shale, porosity can be calculated. The porosity and permeability of the samples are analyzed using the GB-T 29172-2012 core analysis method (Xuefeng et al., 2024).

The mineral composition of shale is primarily comprised of clastic components (such as quartz and felspar), matrix materials (like clay minerals), and cementing materials (such as carbonate minerals). The traditional classification divides it into four major categories based on the content of quartz-feldspar, clay minerals, and carbonate minerals: I, Clayey Shale (CAS); II, Silty Shale (SYS); III, Calcareous Shale (CS); IV, Mixed Shale (BS).IV. Mixed shale (BS). This classification is based on the mineral content characteristics of the study area and the influence of organic carbon in shale oil reservoirs, with 1% and 2% designated as the thresholds for total organic carbon (TOC) content. Three categories are identified: low, medium, and high organic carbon (Ehsan et al., 2024a; Amjad et al., 2023) (Figure 2). Furthermore, four lithofacies are classified: I. Long-feldspar shale facies (feldspar + quartz > 50%); II. Calcareous shale facies (carbonate minerals > 50%); III. Clay-rich shale facies (clay minerals > 50%); IV. Mixed shale facies (no mineral component exceeds 50%). The lithology analyzed displays considerable diversity, particularly with higher amounts of clay minerals (Ehsan et al., 2024b). The mineral composition of Jurassic shale is predominantly quartz and clay minerals, along with lesser amounts of plagioclase and calcite, and includes small quantities of pyrite. In the Liangshang section, plagioclase content is markedly higher compared to the Danzhai and Dongyuemiao sections, suggesting a more substantial impact from proximate sediment sources. In general, the mineral compositions vary greatly among different lithologies. Sandstones are mainly characterized by high contents of quartz and feldspar, whereas shales show substantial variations, predominantly composed of quartz and clay minerals. Notably, the calcareous shale in the Danzhai section has a higher concentration of calcite. Within the study area, the rocks predominantly display three categories of lamination structures: organic-rich lamination, sand-rich lamination, and calcareous lamination (Table 1).

Results from the tests of rock mechanics parameters under triaxial loading conditions are shown in Tables 2, 3 presents the stress-strain curves of selected rock samples, indicating that shale samples transition from elastic deformation to fracturing under low-temperature pressure, whereas at high-temperature pressure, the peak stress is reached more slowly, and significant residual stress is maintained after rupture. As the confining pressure increases, the pores and microfractures within the rock are generally subjected to intense compression from the surrounding rock. The stress-strain curves are categorized into five types based on the morphology and characteristic points of the three stages of the stress-strain relationship. The first three types indicate strong brittleness in the rocks, mainly represented by sandstone, leucogranite shale, and laminated clay shale, with notable differences in mechanical properties observed in both horizontal and vertical orientations (Table 3). The complexity of the fracture patterns in rock samples is a fundamental indication of brittleness. This study primarily identified single shear failure as the fracture mode, with samples largely remaining intact after fracturing. As the uniformity of the rock increases, the elastic properties become stronger, resulting in more complex fractures. Furthermore, if the rock contains layered structures, the fracture mode varies with the orientation of these layers. Supplementary Explanation S1 illustrates that at a pressure of 20 MPa, the fracture characteristics of the rock samples are predominantly of the splitting type, with shear fractures being secondary. Cracks mainly develop axially, showing complex morphologies, with notable fragmentation of the samples. Under the condition of 30MPa, the crack extends to another group of cracks and terminates, accompanied by microcracks at the end of longer cracks. Under this confining pressure, double shear type is dominant, and the fracture surface no longer follows the axis and presents a higher angle with the horizontal.The types of fractures in the rock can be classified into three categories: tensile failure occurs under conditions of high brittleness when axial stress surpasses compressive strength, resulting in tensile fractures along the axial direction, characterized by one or more sets of splitting cracks. Shear failure arises in rocks displaying strong plasticity, with large plastic deformations and evident lateral deformation leading to plastic shear failures. There is also a composite failure type, where the rock samples show both pronounced tensile splitting and shear fractures simultaneously. As the conditions transition from low temperature and low confinement pressure to high temperature and high confinement pressure, the fracture modes of the rock specimens change from a complex splitting-tensile type to a composite tensile-shear type, ultimately resulting in a pure shear type.

Table 4 presents the partial results of rock mechanics parameter tests conducted under tensile strength loading conditions. The results of tensile strength tests for different lithologies in the Jurassic Formation this time vary significantly. Overall, the tensile strength of sandstones from the Lianggaoshan Formation is relatively high, generally exceeding 10 MPa. In contrast, the tensile strength of limestones and mudstones containing shells in the Da′anzai and Dongyuemiao sections is relatively low. Additionally, felsic shales and mixed shales have higher tensile strengths than clayey shales. The experiment revealed three primary fracture modes: simple uniform longitudinal splitting, parallel longitudinal splitting, and intersecting shear and tensile fractures (Supplementary Explanation S2).

R. Rickman et al. explored the brittleness index derived from elastic parameters, employing statistical regression methods. Their findings indicate that higher Young′s modulus and lower Poisson′s ratio correlate with enhanced brittleness. Additionally, they introduced a normalized calculation formula for brittleness based on Young′s modulus and Poisson′s ratio (Rickman et al., 2008) (Equations 1–3). B b r i t = 0.5 E b r i t + 0.5 μ b r i t (1) E b r i t = E − E min E max − E min (2) μ b r i t = μ max − μ μ max − μ min (3)

In this formula, Emax, Emin, μmax, and μmin denote the maximum and minimum values of the Young′s modulus and Poisson′s ratio for the rock samples, respectively. Researchers in China analyzed the mineral composition of the Barrt shale in the Fort Worth Basin and the Lianggaoshan Formation in northeastern Sichuan. They found that the Lianggaoshan Formation has a relatively lower content of brittle minerals and a higher content of clay compared to the Barnett shale in North America, although the general distribution remains similar (Jarvie et al., 2007; Mohammad Mahdi and Reza, 2015). The brittleness index derived from this formula demonstrates a good correlation with the fracture breakdown pressure and extension pressure of shale samples in field fracturing.

The assessment of rock brittleness features through the rock stress-strain curve has been extensively utilized in engineering applications. It reflects the complete process of rock deformation, fracture, and ultimate loss of load-bearing capacity under external loads. It enables the quantitative acquisition of rock characteristics across various stress states. This represents the most intuitive and effective approach for assessing the brittleness of rock materials. The assessment of brittleness characteristics using the stress-strain curve has gained significant application in engineering. This method mainly originates from characteristic points related to the fracture transformations of rock, such as peak stress, peak strain, residual stress, and residual strain. Formulas for assessing brittleness are developed based on the energy area or variations in characteristic points. This study focuses on the following approaches for the evaluation of brittleness (Chen et al., 2018) (Equations 4–6): B i = B i 1 + B i 2 (4) B i 1 = σ P − σ i / δ p ε P − ε i / ε P (5) B i 2 = σ P − σ r / δ p ε r − ε p / ε P (6)

In the equation, δ _i and δ _p are defined as the initiation stress and the peak stress, respectively. ɛ _i and ɛ _p represent the initiation strain and the peak strain. Additionally, δ _r denotes the residual stress, while ɛ _r signifies the residual strain. This approach considers the changes in both the pre-peak and post-peak curves, and it allows for relatively simple calculations.

The assessment of sandstone brittleness, grounded in the complete stress-strain curve and energy variations during the rock fracture process, is categorized into three main stages: work hardening and linear elastic damage, plastic deformation and damage, and failure and softening. For each sample, the stress-strain curve is analyzed, and adjustments to the theoretical programming code are made. Remapping, energy block division, and calculation of brittleness-sensitive parameters across various stages are conducted. The fuzzy analytic hierarchy process, based on energy evolution values from different stages, is used to ascertain weights and contribution values. This entire procedure is repeated three times to derive the final average (Table 5).

The brittleness data derived from rock mechanics experiments or mineral composition tests are generally reliable, but they pose challenges such as high costs, time consumption, and the difficulty of performing comprehensive single-well profile analysis due to data dispersion. Various researchers have developed evaluation methods for shale reservoir fracturability based on logging data, which varies continuously with depth and is highly accurate. Acquiring dynamic parameters of rock mechanics from logging data is a key approach for assessing shale brittleness. Typically, dynamic rock mechanics parameters such as Young′s modulus, Poisson′s ratio, internal friction angle, and cohesion are calculated using P-wave travel time, S-wave travel time, and density logging data. The following formulas are used to calculate the dynamic rock mechanics parameters (Equations 7–12): Dynamic Young ′ s  Modulus : E = ρ b 3 ∆ t s 2 − 4 ∆ t p 2 ∆ t s 2 ∆ t s 2 − ∆ t p 2 × 9.299 × 10 7 (7) Dynamic Poisson ′ s  Ratio : μ = ∆ t s 2 − 2 ∆ t p 2 2 ∆ t s 2 − ∆ t p 2 (8) Compressive Strength : σ c = 0.0045 E 1 − V s h + 0.008 E V s h (9) Tensile Strength : σ t = 0.0045 E 1 − V s h + 0.008 E V s h 12 (10) Cohesion : C = 4.69433 × 10 7 × ρ b 2 1 + μ 1 − μ 1 − 2 μ 1 + 0.78 V s h ∆ t p 4 (11) Internal Friction Angle : φ = π 12 2 1 − μ d 1 − μ d + 1 (12)

In this equation, ρ b is defined as the DEN value; E denotes the dynamic Young′s modulus, expressed in GPa; μ represents the dynamic Poisson′s ratio, a dimensionless quantity; σ c denotes the compressive strength, expressed in MPa; σ t represents the tensile strength, quantified in MPa; C denotes the cohesion, expressed in MPa; φ represents the internal friction angle, quantified in degrees; Δt _s and Δt _p represent the travel times of shear and compressional waves, respectively, expressed in μs/m; ρ denotes the volume density of the geological formation, quantified in g/cm³.

The brittleness index, derived from mechanical experiments, is applied to the well-logging evaluation model, five evaluation factors that significantly influence brittleness are taken into account: E/μ, compressive strength, tensile strength, cohesion, and internal friction angle (Fang et al., 2007; Lei et al., 2007a; Lei et al., 2007b). The Analytic Hierarchy Process (AHP) is employed to quantify the five parameters and conduct pairwise comparisons, thereby constructing a judgment matrix for the calculation of weights (Table 6).

A judgment matrix (A) is constructed for the factors contributing to the comprehensive brittleness indicator (BI), simultaneously, (A) is normalized by columns, followed by the calculation of the eigenvector (ω) and the product (Aω) (Tables 7, 8).

The matrix (A) is subjected to consistency evaluation, and the consistency ratio is calculated according to the following formula (Equations 13–15): C R = C I R I (13) C I = λ max − n n − 1 (14) λ max = ∑ i = 1 n A ω i n ω i (15)

The average random consistency index (RI) for (n = 5) is identified as 1.1075, and the calculated consistency ratio (CR = 0), which is below 0.1, signifies that the consistency has been achieved. It should be noted that the assessment of ambiguity in this study is conducted based on correlation, thereby incorporating a certain level of subjectivity. The weights assigned to (BI), compressive strength, tensile strength, cohesion, and internal friction angle in the comprehensive brittleness indicator (BI) derived from rock mechanics are determined to be 0.393, 0.302, 0.139, 0.095, and 0.072, respectively. The comprehensive brittleness evaluation metric for the Jurassic shale in northeastern Sichuan is derived as follows (Equations 16–18): B I = 0.393 E μ + 0.302 σ c + 0.139 σ t + 0.095 C + 0.072 φ (16)

Ultimately, the standardization was performed using the range variation method, represented by the following formula: S p = X − X min X max − X min (17) S n = X max − X X max − X min (18)

In this equation, S _p and Sn indicate the standardized values of positive and negative indicators, respectively; X signifies the parameter value; X _max refers to the maximum parameter value; and X _min indicates the minimum parameter value.

The comprehensive brittleness indicator (BI), which was derived from the normalization of E/μ, compressive strength, tensile strength, cohesion, and internal friction angle, displays consistency with the fracture breakdown pressure and extension pressure observed during field fracturing. It was found that a higher brittleness index corresponds to lower fracture breakdown and extension pressures, suggesting relatively good compressibility (Figure 10). The strength of the comprehensive brittleness indicator (BI) is attributed to its consideration of a wide range of brittleness influencing factors and its solid correlation with the brittleness index obtained from the full stress-strain curve. This indicator effectively reflects the intrinsic failure characteristics of the rock and facilitates continuous evaluation in well logging, indicating that the brittleness evaluation model possesses generalizability. By applying this indicator to assess the brittleness of the PA1 upper and lower sections in the deep layers of northeastern Sichuan, it was observed that the upper section exhibits a higher brittleness index (Figure 11), indicating a greater tendency for the shale to develop a network of fractures, which is crucial for future shale gas exploitation.