A combined coal rock hydraulic fracture identification method and medium

Abstract

The application discloses a combined coal rock hydraulic fracture identification method and medium: a combined coal rock sample is placed in a true triaxial loading space, and three principal stresses are applied to the combined coal rock sample to obtain target stress control data; according to the target stress control data and a preset injection rate, a fracturing fluid is injected into a fracturing borehole in the combined coal rock sample, and fracturing process response data is obtained; pump injection pressure data is divided into stages to obtain fracturing stage identification results, and pressure characteristic parameters are extracted based on the fracturing stage identification results; after fracturing is completed, crack distribution data of the combined coal rock sample is obtained, and a crack interface extension path is identified based on the crack distribution data; the pressure characteristic parameters, the preset injection rate, the crack interface extension path and the target stress control data are associated and analyzed to identify a hydraulic fracture extension mode of the combined coal rock sample. The application realizes identification of the hydraulic fracture extension mode under different vertical stress constraint states.

Description

Technical Field

[0001] This application relates to the field of coal and rock mass analysis, and in particular to a combined method and medium for identifying hydraulic fractures in coal and rock. Background Technology

[0002] In engineering fields such as coal mining, roof depressurization, and unconventional reservoir permeability enhancement, hydraulic fracturing of coal-rock masses with hard roofs or distinct layered structures is a crucial technique for controlling surrounding rock structure, expanding the weakened area, and improving fracture seepage conditions. Especially under hard roof conditions, the roof rock mass is usually not a single homogeneous stratum, but rather a composite coal-rock structure composed of two or more layers of high-strength, thick, and mechanically different rock layers combined with a coal seam. These composite coal-rock masses exhibit well-developed interlayer interfaces, with differences in mechanical parameters such as elastic modulus, tensile strength, and fracture toughness between different layers, along with strong heterogeneity and anisotropy. Therefore, when fracturing fluid drives fracture propagation in such media, the fracture propagation process is not only controlled by fluid pressure but also influenced by interlayer interface constraints, lithological differences, and in-situ stress states. The fracture propagation behavior is more complex than in single homogeneous rock samples, directly affecting the roof weakening range, depressurization effect, and fracturing results.

[0003] Existing analytical methods for hydraulic fracture propagation behavior mostly focus on homogeneous rock samples or single rock layers. They typically rely on peak pressure, fracturing pressure, or single-stage pressure variation characteristics in the pumping pressure curve to determine fracture initiation and propagation, or on empirically summarizing fracture propagation paths based on observed fracture morphology after fracturing. These methods are applicable to conditions where fracture propagation paths are relatively simple and interface effects are not significant. However, for composite coal-rock masses with significant interlayer interfaces and lithological differences, under the coupling effect of true triaxial stress and injection rate, significant stress redistribution and propagation impedance changes occur near the interlayer interfaces at the fracture tip. This leads to dynamic transformations during fracture propagation, including cross-layer propagation, interface-based propagation, interface-arrested propagation, and multi-branch competitive propagation. Simultaneously, different fracture branches may exhibit mutual induction, mutual inhibition, or competitive dominance, resulting in complex evolutionary characteristics in the pumping pressure response, such as post-peak fluctuations, local rebounds, and continuous attenuation. Based solely on a single pump pressure parameter, it is difficult to distinguish the differences in mechanical response corresponding to the above-mentioned different expansion behaviors; based solely on the post-fracturing fracture morphology, it is also difficult to reconstruct the dynamic competitive evolution mechanism of fractures during the fracturing process.

[0004] Furthermore, during the fracturing of composite coal-rock masses, different combinations of geostress, particularly the difference between vertical stress and horizontal principal stress, alter the stress boundary conditions near the interlayer interfaces. This affects whether the fracture tip can overcome interlayer resistance to achieve cross-layer propagation or instead deflects along the interface. Different injection rates also change the process of fluid pressure build-up and energy release, altering the relative strength between the fracture propagation driving force and the interlayer constraint. In other words, the hydraulic fracture propagation mode in composite coal-rock masses is not determined by a single stress parameter or pressure characteristic, but rather by the combined effects of stress constraint state, injection rate conditions, pressure response during fracturing, and post-fracturing fracture spatial distribution. Current technology lacks an identification method capable of uniformly characterizing and correlating this multi-source information, making it difficult to accurately reveal the formation mechanism of competitive fracture propagation modes under different vertical stress constraint states and injection rate conditions, and also hindering the effective identification and evaluation of their mechanical responses.

[0005] Therefore, accurately identifying the competitive propagation modes of hydraulic fractures and their corresponding mechanical responses in combined coal-rock masses with obvious interlayer interfaces, lithological differences, and heterogeneity has become a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0006] To address the problem in existing technologies that make it difficult to accurately identify the competitive propagation mode of hydraulic fractures and their mechanical response in combined coal and rock masses under the coupled action of true triaxial stress and injection rate, this application provides a method and medium for identifying hydraulic fractures in combined coal and rock masses. By performing correlation analysis on the combined pressure characteristic parameters, fracture interface propagation path, injection rate and stress control data, the method can identify the propagation mode of hydraulic fractures under different vertical stress constraints.

[0007] One aspect of this application provides a combined method for identifying hydraulic fractures in coal and rock, comprising:

[0008] Obtain composite coal and rock sample data;

[0009] The composite coal and rock sample was placed in a true triaxial loading space, and triaxial principal stress was applied to the composite coal and rock sample to obtain target stress control data; among which, the target stress control data included the maximum horizontal principal stress, the minimum horizontal principal stress, the vertical principal stress, and the vertical stress difference coefficient determined by the vertical principal stress.

[0010] Based on the target stress control data and the preset injection rate, fracturing fluid is injected into the fracturing borehole in the combined coal and rock sample, and the fracturing process response data is obtained; wherein, the fracturing process response data includes at least the pumping pressure data;

[0011] The pump injection pressure data is divided into stages to obtain fracturing stage identification results, and pressure characteristic parameters are extracted based on the fracturing stage identification results;

[0012] After hydraulic fracturing, fracture distribution data of the combined coal and rock samples were acquired, and fracture interface propagation paths were identified based on the fracture distribution data.

[0013] Correlation analysis was performed on pressure characteristic parameters, preset injection rate, fracture interface propagation path, and target stress control data to identify hydraulic fracture propagation modes of composite coal and rock samples under different vertical stress constraints and different injection rates.

[0014] Another aspect of this application provides a computer-readable storage medium storing computer instructions that, when executed by a processor, implement a combined coal and rock hydraulic fracture identification method of this application.

[0015] Compared to existing technologies, the advantages of this application are:

[0016] This application does not infer fracture morphology solely based on a single pump pressure parameter or the fracture propagation law of homogeneous rock samples. Instead, it first obtains the maximum horizontal principal stress, minimum horizontal principal stress, vertical principal stress, and vertical stress difference coefficient under true triaxial loading conditions, thereby characterizing the vertical stress constraint state of the composite coal-rock mass during the fracturing process. Since different vertical stress constraint states will change the stress state and propagation resistance of the fracture tip near the interlayer interface, it can provide stress boundary basis for identifying the transformation between cross-layer propagation, interface propagation, interface arrest, and multi-branch competitive propagation.

[0017] Furthermore, this application divides the pump injection pressure data into stages and extracts pressure characteristic parameters such as peak pressure, duration of rapid pressurization stage, pressure fluctuation amplitude, pressure fluctuation frequency, pressure decay amplitude, and duration of depressurization stage. This allows for the characterization of the pressure accumulation process before and after fracture initiation, the disturbance process during the interaction between the fracture and the interlayer interface, and the depressurization and release process after fracturing. This enables the capture of the corresponding dynamic mechanical response during the competitive propagation of fractures.

[0018] Furthermore, this application acquires fracture distribution data after fracturing and identifies fracture interface propagation paths based on the fracture distribution locations in each layer and their positional relationship with interlayer interfaces. This ensures that the pressure response information during fracturing corresponds to the actual fracture paths formed post-fracturing, avoiding the one-sidedness caused by inferring fracture propagation modes solely from pump pressure curves. Finally, this application establishes a correlation between fracture propagation modes and mechanical responses under different vertical stress constraint states and different injection rates through correlation analysis of combined pressure characteristic parameters, preset injection rates, fracture interface propagation paths, and target stress control data. This enables accurate identification and evaluation of the competitive propagation modes and evolution mechanisms of hydraulic fractures in combined coal and rock masses. Attached Figure Description

[0019] This application will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0020] Figure 1 This is an exemplary flowchart of a combined coal and rock hydraulic fracture identification method of this application;

[0021] Figure 2 This is a schematic diagram of the large-size true triaxial coal and rock mass fracturing test system of this application;

[0022] Figure 3 This is a schematic diagram of the sample preparation for this application;

[0023] Figure 4 This is a schematic diagram showing the arrangement of the acoustic emission sensors in this application;

[0024] Figure 5 This is a graph showing the pumping pressure and confining pressure of this application;

[0025] Figure 6 This is a graph showing the changes in pump injection pressure at different injection rates according to this application;

[0026] Figure 7 This is a graph showing the changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at a pumping rate of 30 mL / min according to this application.

[0027] Figure 8 This is a graph showing the changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at a flow rate of 60 mL / min according to this application.

[0028] Figure 9 This is a schematic diagram of the unfolded six sides of the sample of this application;

[0029] Figure 10 This is a characteristic diagram of the surface crack distribution of the W-30 group of samples when k=1.4 in this application;

[0030] Figure 11 This is a characteristic diagram of the surface crack distribution of the W-30 group of samples when k=1.0 in this application;

[0031] Figure 12 This is a characteristic diagram of surface crack distribution of W-30 group samples when k=0.6 in this application;

[0032] Figure 13 This is a characteristic diagram of the surface crack distribution of the W-30 group of samples when k=0.2 in this application;

[0033] Figure 14 This is a characteristic diagram of the surface crack distribution of the W-60 group of samples when k=1.4 in this application;

[0034] Figure 15 This is a characteristic diagram of the surface crack distribution of the W-60 group of samples when k=1.0 in this application;

[0035] Figure 16 This is a characteristic diagram of the surface crack distribution of the W-60 group of samples when k=0.6 in this application;

[0036] Figure 17 This is a characteristic diagram of the surface crack distribution of the W-60 group of samples when k=0.2 in this application. Detailed Implementation

[0037] The methods and systems provided in the embodiments of this application will be described in detail below with reference to the accompanying drawings. It should be noted that the following embodiments are used to illustrate the technical concept, implementation path, and application method of this application, and are not intended to limit the scope of protection of this application.

[0038] This application targets composite coal-rock strata or their corresponding composite coal-rock mass structures that exhibit significant interlayer interfaces, lithological differences, and heterogeneity. During hydraulic fracturing, these strata are influenced by factors such as geostress state, injection parameters, interlayer interface mechanical properties, and lithological differences. Hydraulic fractures in these formations are prone to dynamic transformations between various modes, including cross-layer propagation, interface propagation, interface arrest, and multi-branch competitive propagation. Existing technologies typically rely on single pump pressure parameters, single monitoring methods, or the assumption of homogeneous rock formations to assess fracture propagation behavior. This approach struggles to accurately identify the true propagation modes of fractures under composite coal-rock conditions, thus hindering direct application to parameter optimization, fracture orientation determination, and propagation risk control in field fracturing operations.

[0039] Based on this, such as Figure 1 As shown, this application proposes a combined coal and rock hydraulic fracture identification method for engineering applications. It focuses on target stress control data, pumping pressure stage characteristics, and fracture interface propagation path information. Through joint analysis of pressure response data, formation structure data, construction parameter data, and fracture distribution information during fracturing, it identifies hydraulic fracture propagation modes under different vertical stress constraints and injection rates. The identification results can be further used for optimizing fracturing construction parameters, assessing interlayer penetration capability, providing early warning of interface fracture arrest risks, and evaluating the formation effect of target fracture networks, thus providing technical support for the design and on-site construction control of combined coal and rock reservoir fracturing.

[0040] I. Construction of the Experimental System and Identification Platform:

[0041] This embodiment utilizes a large-scale true triaxial coal and rock fracturing test machine to simulate the entire fracturing process of different types of rocks under underground stress structures. It can also observe and analyze the crack propagation patterns, pump pressure response characteristics, and multi-field coupling effects after fracturing. A schematic diagram and physical image of the system are shown below. Figure 2 As shown.

[0042] like Figure 2 As shown, the test platform uses a KDLS-V true triaxial apparatus as the main loading unit, applying triaxial principal stress to the sample through a five-sided loading method to simulate the actual stress state of the underground reservoir. The loading in each of the three directions can be controlled independently or synchronously, allowing for the construction of different triaxial stress combinations according to the identification requirements of this application. The pressure chamber and loading cylinder of the triaxial apparatus are mounted on a bottom fixed frame, and the five loading plates are detachable to allow for changing the position of the acoustic emission sensors as needed for the experiment. A moving guide rail is provided on the front cover of the pressure chamber for sample loading and unloading. The device also features load and displacement parameter setting and over-limit protection functions, enabling multi-directional synchronous and stable loading, providing the basic hardware conditions for subsequent acquisition of target stress control data.

[0043] The testing machine control system mainly includes: a fracturing / confining pressure system, a servo hydraulic system, a back pressure metering system, a data acquisition and control system, and auxiliary systems. The fracturing / confining pressure system is used for sample confining pressure loading and fracturing process control; the servo hydraulic system consists of a high-pressure servo loading pump and a servo oil source, providing power to five axial loading cylinders; the back pressure metering system applies back pressure at the model outlet that is higher than atmospheric pressure and close to the simulated formation pressure to ensure stable effluent flow and improve outlet fluid metering accuracy; the data acquisition and control system includes permeability testing, fracturing pressure monitoring, flow monitoring, displacement acquisition, acquisition cards, PLC, sensors, an automated control system, and data processing software. Through the coordinated operation of these systems, a complete monitoring chain covering pre-fracturing, fracturing, and post-fracturing stages can be formed, providing a data foundation for subsequent pumping pressure stage identification, pressure characteristic parameter extraction, and fracture path identification.

[0044] In this embodiment, the injection device uses a KDHB-70 single-cylinder benchtop high-precision constant-speed and constant-pressure tracking pump with a flow rate range of 0.1–200 mL / min and a maximum output pressure of approximately 50 MPa. It can achieve constant injection rate or staged injection rate control and display pressure, flow rate, and other parameter curves in real time. A ZR-Ⅲ type piston container is used to store the pre-pressurized fracturing fluid and stably delivers the fracturing fluid into the fracturing holes within the sample under the drive of the injection pump. Since this application requires comparing the pump pressure response and fracture propagation mode of combined coal and rock samples under different injection rate conditions, this type of high-precision injection equipment can provide stable and repeatable flow boundary conditions for the experiment.

[0045] In addition, to assist in the identification of fracture propagation paths, this embodiment also includes an acoustic emission signal detection system and a tracer development method. The acoustic emission system is used to record transient elastic wave signals caused by fracture initiation, propagation, and branching evolution during fracturing; the tracer is used to assist in identifying the orientation of surface fractures after fracturing and to provide a visual basis for classifying fracture interface propagation paths.

[0046] Thus, this embodiment forms an identification platform that integrates true triaxial stress simulation, stable flow injection, process signal acquisition, and post-compression crack observation.

[0047] II. Data Acquisition and Sample Preparation of Combined Coal and Rock Samples:

[0048] The composite coal and rock sample data includes the sample's layered structure, lithological data, interlayer interface data, and fracturing pore data. In this embodiment, a three-layer composite coal and rock formation consisting of coarse-grained sandstone, coal, and medium-grained sandstone is selected as the sample. Figure 3 As shown in the figure, the upper part is coarse-grained sandstone, the middle part is a coal seam, and the lower part is medium-grained sandstone. The overall dimensions are 200mm×200mm×200mm, with the upper coarse-grained sandstone being 60mm thick, the middle coal seam being 80mm thick, and the lower medium-grained sandstone being 60mm thick.

[0049] The three-layer structure is adopted for two reasons. First, in actual coal-bearing strata, the hard roof is not a single homogeneous layer, but often interlayered or sandwiched with weaker layers such as coal, mudstone, or carbonaceous mudstone. Second, the significant differences in lithology and mechanical properties between the upper and lower strata and the coal seam can form stable interlayer interfaces and heterogeneous structures, thus providing a real medium basis for phenomena such as cross-layer propagation, deflection along interfaces, interface arrest, and branching competition during fracture propagation. In particular, this embodiment uses medium-grained sandstone and coarse-grained sandstone as the upper and lower surrounding rocks of the coal seam, respectively, instead of using the same lithological materials. This further differentiates the propagation resistance, interface crossing ability, and deflection trend of fractures at the upper and lower interfaces, making it easier to identify the competitive propagation behavior of fractures at different interlayer interfaces.

[0050] The coarse-grained sandstone, medium-grained sandstone, and coal seam used in the sample were all obtained by cutting actual rock blocks on site. During preparation, each rock block was first machined to the designed dimensions using wire cutting; then, the contact surfaces were polished to ensure flat end faces and good interlayer adhesion; subsequently, surface dust and debris were cleaned, and epoxy resin was evenly applied to the upper and lower interlayer interfaces. The three parts were then aligned, pressed, and fixed according to a predetermined layer sequence to prevent misalignment during curing; after the adhesive had cured, the outer surface of the sample was trimmed and numbered, ultimately obtaining a combined coal-rock mass sample that met the experimental requirements.

[0051] The resulting sample data includes at least: information on the layered structure of the upper coarse-grained sandstone, the middle coal seam, and the lower medium-grained sandstone; information on the lithological and mechanical differences between each layer; information on the interlayer interfaces between the coarse-grained sandstone and the coal seam, and between the coal seam and the medium-grained sandstone; and data on the fracturing holes formed by subsequent drilling and sealing.

[0052] Based on this, pretreatment of the sample is required to meet the requirements of fluid injection, borehole sealing, and pressure stability in true triaxial hydraulic fracturing tests. First, the outer surface of the sample is measured and calibrated to complete the borehole center positioning, marking, and drilling direction confirmation. The drilling point is then marked at the center of the coal seam in the middle of the sample. Next, the sample is fixed on the drilling machine table, and a 10mm diameter, 100mm long drill bit is used for drilling. During drilling, debris is removed by controlling the feed rate and intermittent purging. After drilling, a leveling drill bit of the same size is used to level the bottom of the hole to ensure a flat bottom, preventing uneven stress during fracturing. Finally, a fracturing borehole with a diameter of 10mm and a length of 100mm is formed.

[0053] Subsequently, a 7mm diameter, 180mm long hollow aluminum tube was selected as the fracturing tube, and a connector was installed at the end of the tube for connection with the injection system and pressure sealing. After inserting the fracturing tube into the borehole, the borehole opening and surrounding area were cleaned and dried. Then, epoxy resin was injected in stages into the borehole opening and the gap between the tubes using a glue gun, allowing the resin to fully fill and form a continuous sealing layer. At the same time, the outside of the borehole opening was wrapped and reinforced. After the resin cured, a combined coal and rock sample capable of withstanding subsequent fracturing injection was obtained.

[0054] Thus, this embodiment completes the acquisition and construction of composite coal and rock sample data, enabling the sample to simultaneously possess: obvious interlayer interfaces; differences in lithology and mechanical properties; and fracturing pores and sealing structures within the central coal seam; thereby meeting the prerequisites for subsequent path identification based on the positional relationship between fractures and interlayer interfaces in this application.

[0055] III. Construction of Target Stress Control Data:

[0056] This embodiment constructs target stress control data under different vertical stress constraint states by setting the maximum horizontal principal stress, the minimum horizontal principal stress, and the vertical principal stress.

[0057] Specifically, as shown in Table 1, this embodiment maintains the maximum horizontal principal stress. and minimum horizontal principal stress Constant, take respectively , By adjusting the vertical principal stress Construct different triaxial stress combinations, The pressure was set to 12 MPa, 10 MPa, 8 MPa, and 6 MPa, respectively, forming four different vertical stress constraint states. To characterize the deviation of the vertical principal stress from the two horizontal principal stresses, this embodiment further introduces a vertical stress difference coefficient k, with corresponding values ​​of 1.4, 1.0, 0.6, and 0.2.

[0058] Table 1 Test Scheme

[0059]

[0060] Therefore, the triaxial stress combination can be matched with the corresponding vertical stress difference coefficient k to obtain target stress control data under different vertical stress constraint states. For this application, the target stress control data is not only used to describe the experimental boundary conditions, but also serves as a core working condition variable in subsequent crack propagation mode identification. When k is large, the vertical constraint is enhanced, and the stress state and propagation impedance of the crack tip near the interface will change significantly, which is more likely to induce higher crack initiation pressure, stronger post-peak fluctuations, and more complex interface interactions; while when k is small, the vertical constraint is weakened, and the crack is more likely to form a relatively simple propagation path in the coal seam according to the principal stress control direction.

[0061] To ensure stability and repeatability during triaxial stress loading, this embodiment employs a graded loading method to apply the target stress control data. Specifically, this includes: First, simultaneously loading the triaxial principal stresses to the minimum horizontal principal stress value and stabilizing the pressure for 10 minutes; Second, simultaneously increasing the maximum horizontal and vertical principal stresses to the smaller of the two values ​​and stabilizing the pressure for another 10 minutes; Third, loading the larger of the two values ​​at a lower rate to the set target value and stabilizing the pressure for 30 minutes. This graded loading method prevents premature failure of the specimen under uneven stress and ensures that the triaxial stress reaches a stable equilibrium before injection, thus guaranteeing that the subsequent pumping pressure response primarily reflects crack initiation and propagation behavior, rather than transient loading interference.

[0062] IV. Acquisition of fracturing process response data and stage division of pump injection pressure data, and identification of fracturing stages:

[0063] After obtaining the target stress control data, this embodiment injects fracturing fluid into the fracturing borehole according to the current fracturing conditions and collects data on the change of pumping pressure over time, thereby generating fracturing process response data.

[0064] To facilitate post-fracturing fracture development and path identification, this embodiment pre-adds the aqueous fluorescent dye LUYOR-6200 to the piston container. The water-to-crystal ratio used in this embodiment is 1:500. Because this tracer is readily soluble in water and emits bright fluorescence under ultraviolet light, it can assist in identifying the fracture surface propagation path after fracturing and correlate it with the pumping pressure information.

[0065] Before fracturing, such as Figure 4 As shown, acoustic emission sensors are arranged at designated positions on the loading plate and two lateral loading plates of the true triaxial apparatus, with two sensors on each loading plate, for a total of six channels. Each sensor is pre-coated with coupling agent and securely fixed to the loading plate using clamps. The acquisition parameters are shown in Table 2.

[0066] Table 2 Acquisition Parameters for Acoustic Emission

[0067]

[0068] By ensuring stable signal response in each channel through steps such as grounding and tapping calibration, the acoustic emission system provides auxiliary information on the dynamic evolution of cracks, which helps to verify whether the post-peak fluctuation stage corresponds to crack adjustment, branch competition, or enhanced interface interaction, thereby improving the reliability of the identification results.

[0069] After completing sensor installation and stabilizing the operating conditions, the sample was placed in the true triaxial apparatus chamber, and the fracturing tube on the sample was connected to the pipeline inside the chamber. The sealing sleeve and loading plate were then installed. Subsequently, according to the operating conditions set in Table 1, the high-pressure precision injection pump was started, and fracturing fluid was continuously injected into the sample at a preset injection rate. In this embodiment, two injection rates were set: 30 mL / min, corresponding to group W-30; and 60 mL / min, corresponding to group W-60. Each injection rate corresponds to four vertical stress constraint states, constituting a total of 8 sets of tests. The sample numbers and operating conditions are shown in Table 1.

[0070] During the injection process, data on the change of pump injection pressure over time are continuously collected, and the pump injection pressure data at each moment is matched with the corresponding injection time, injection rate, and current target stress control data to generate fracturing process response data. This fracturing process response data includes at least: the current fracturing condition; the injection rate; the pump injection pressure time-series curve; triaxial stress change information; and the acoustic emission response signal.

[0071] Figure 5 Eight sets of pumping pressure and triaxial stress control curves under different operating conditions are shown. Figure 5 (a) W-30-1 shows the pumping pressure and triaxial stress curves under operating condition W-30-1. Figure 5 (b) W-30-2 shows the pumping pressure and triaxial stress curves under operating condition W-30-2. Figure 5 (c) W-30-3 shows the pumping pressure and triaxial stress curves under operating condition W-30-3. Figure 5 (d) W-30-4 shows the pumping pressure and triaxial stress curves under operating condition W-30-4. Figure 5 (e) W-60-1 shows the pumping pressure and triaxial stress curves under operating condition W-60-1. Figure 5 (f) W-60-2 shows the pumping pressure and triaxial stress curves under operating condition W-60-2. Figure 5 (g) W-60-3 shows the pumping pressure and triaxial stress curves under operating condition W-60-3. Figure 5 (h) W-60-4 shows the pumping pressure and triaxial stress curves under operating condition W-60-4.

[0072] Depend on Figure 5 As shown in (a) W-30-1, under this condition, the pumping pressure rapidly increases from near zero after the injection begins, reaching a peak of approximately 20.32 MPa in the early stage, followed by a relatively long post-peak evolution process. This post-peak evolution process is characterized by a plateau segment superimposed with multiple sawtooth fluctuations, indicating that after crack initiation under this condition, a stable single channel does not immediately form, but rather there is a significant process of propagation path adjustment, branching disturbance, and interface interaction. At the same time, the triaxial stress curve remains nearly horizontal overall, only decreasing synchronously at the end of unloading, indicating that the target stress control boundary is stable, and the pumping pressure fluctuation mainly originates from crack propagation behavior rather than disturbances in the loading system.

[0073] Depend on Figure 5 (b) As shown in W-30-2, under this operating condition, the pumping pressure also rises rapidly in the early stage, reaching a peak of approximately 17.03 MPa. After the peak, the pressure quickly drops to a lower level, then gradually rises slowly and tends to a relatively stable plateau state. Figure 5 (a) Compared to W-30-1, the fluctuation amplitude of the post-peak evolution stage under this condition is significantly reduced, and no continuous and significant sawtooth fluctuations appear. This indicates that under the stress and injection conditions corresponding to this condition, the dominant fracture channel forms relatively quickly after the peak, and the subsequent propagation process is relatively concentrated and stable. Its triaxial stress curve remains basically stable throughout the injection stage, which also indicates that this curve can be used as an effective input for subsequent fracture mode identification.

[0074] Depend on Figure 5(c) As shown in W-30-3, under this condition, the pumping pressure reaches a peak of approximately 12.96 MPa after a rapid increase, followed by a prolonged post-peak evolution phase. This phase does not directly enter a stable plateau, but rather exhibits a complex pattern of multiple local fluctuations, periodic declines, and subsequent re-increases. This indicates that during crack propagation, there are multiple channel reconstructions, reopening after local obstruction, and dynamic adjustment behaviors under the influence of interlayer interfaces. The triaxial stress curve remains generally stable, only decreasing rapidly during unloading, demonstrating that the complex post-peak fluctuations under this condition also have a clear indicative significance for crack response.

[0075] Depend on Figure 5 As shown in (d)W-30-4, under this operating condition, the pumping pressure reaches a peak of approximately 11.50 MPa in the early stage and then rapidly declines, subsequently showing a slow upward trend. The post-peak fluctuations are relatively small, with only slight fluctuations occurring in local periods, indicating a more stable and continuous propagation process overall. This characteristic suggests that under this operating condition, a relatively stable main channel is formed quickly after crack initiation, and the degree of branch competition and interface disturbance is weak during subsequent crack propagation. Its triaxial stress curve also remains basically horizontal and stable, providing reliable boundary conditions for the stage division of the pumping pressure curve.

[0076] Depend on Figure 5 As shown in (e)W-60-1, under this condition, the pumping pressure rapidly increases to a peak of approximately 22.11 MPa, then falls back to near a plateau. During the post-peak evolution phase, it exhibits a pronounced plateau-superimposed sawtooth-shaped fluctuation characteristic, and the fluctuation process lasts for a relatively long time. This type of post-peak pressure response indicates that, under the corresponding conditions, crack propagation does not proceed steadily along a single path, but is more susceptible to the influence of interlayer interfaces and heterogeneous structures, resulting in strong path deflection, branching competition, and channel adjustment. Under this condition, the triaxial stress curve is essentially a horizontal straight line during injection, indicating that the post-peak fluctuation has high identification effectiveness.

[0077] Depend on Figure 5 (f) As shown in W-60-2, under this operating condition, the pump injection pressure reaches a peak of approximately 18.54 MPa during the rapid pressurization phase, then drops back and remains near a relatively high plateau in the subsequent phase, accompanied by some local fluctuations. Figure 5 (e) Compared to W-60-1, its post-peak fluctuation amplitude is slightly smaller, but multiple disturbances can still be observed, indicating that channel adjustment and interface interference still exist in the crack propagation process under this condition, only the overall competition is weaker than that of W-60-1. Figure 5 (e) W-60-1. The triaxial stress curve remains basically stable, further proving that this type of fluctuation can be attributed to the crack propagation dynamics.

[0078] Depend on Figure 5As shown in (g)W-60-3, under this operating condition, the pumping pressure rapidly reaches a peak of approximately 13.35 MPa in the early stage, then quickly drops back to a lower level, and in the post-peak stage, it generally exhibits a slow and continuous climb with relatively weak local fluctuations. This characteristic indicates that under this operating condition, the dominant flow channel is established relatively quickly after crack initiation, the crack path is relatively clear during the post-peak propagation process, and the degree of interface disturbance and branch competition is low. The triaxial stress curve is generally stable, only decreasing synchronously when the pump is stopped and unloaded.

[0079] Depend on Figure 5 As shown in (h)W-60-4, under this operating condition, the pumping pressure reaches a peak of approximately 11.70 MPa in the early stage, then falls back and enters a relatively stable post-peak evolution phase, generally characterized by a slow climb with only minor fluctuations. This operating condition is similar to... Figure 5 (g) Similar to W-60-3, both exhibit rapid formation of the main fracture channel and relatively stable subsequent propagation. However, differences exist in their peak levels and later evolution slopes, indicating that the fracture propagation dynamics still exhibit distinguishable response characteristics under different injection parameters or stress combinations. Their triaxial stress curves also remain stable, meeting the boundary condition requirements for subsequent identification and analysis.

[0080] Combination Figure 5 (a) W-30-1 to Figure 5 As shown in (h)W-60-4, the triaxial stress curves under all eight working conditions remained basically horizontal and stable before and after injection, with a synchronous decrease only occurring during the pump shutdown and unloading phase. This indicates that the target stress control data obtained in this embodiment has high stability and control accuracy. Based on this, the pump injection pressure curves of each group can be divided into a rapid pressurization stage, a post-peak evolution stage, and a depressurization stage. The duration, fluctuation characteristics, and turning points of each stage differ under different working conditions. These differences can be used to characterize the processes of fracture initiation, interface effects, and post-pressure relief under combined coal and rock conditions.

[0081] In this embodiment, the segmentation of the pump injection pressure data is based on the pressure change relationship corresponding to continuous sampling times. Let the pump injection pressure data be denoted according to the sampling time sequence as follows: The pressure change at adjacent sampling times is denoted as The entire pumping pressure data is traversed to determine the sampling point corresponding to the maximum pressure as the peak point. Then, using this peak point as the boundary and combining it with the time when fracturing fluid injection stops, the pressure changes of adjacent sampling times in different sections are judged, thereby completing the identification of the three stages.

[0082] The first preset condition corresponds to the identification condition for the rapid pressurization stage. Specifically, in multiple consecutive sampling times (N ≥ 3), if the pressure changes of adjacent intervals are all greater than zero, and the corresponding pressure values ​​within the continuous segment do not exceed the pressure value corresponding to the peak point, then the continuous segment is identified as the rapid pressurization stage. The rapid pressurization stage refers to the segment in which the pumping pressure continuously rises from its initial value and maintains positive growth before reaching the peak point. It is used to characterize the pressure response process corresponding to the continuous accumulation of internal pressure, fracture initiation, and initial propagation of the sample after fracturing fluid injection.

[0083] The second preset condition corresponds to the identification condition of the post-peak evolution stage. Specifically, in multiple consecutive sampling times (N ≥ 8) after the peak point and before the cessation of fracturing fluid injection, if adjacent pressure changes simultaneously contain both positive and negative values, the corresponding segment is identified as the post-peak evolution stage. This criterion indicates that the post-peak evolution stage is not a monotonous pressure reduction process, but rather, after the overall peak, the pressure curve exhibits characteristics such as fluctuation, rebound, plateau, local rise, or readjustment due to factors such as fracture propagation path adjustment, interface deflection, local fracturing arrest, branch competition propagation, and re-initiation. Therefore, in this embodiment, the post-peak evolution stage can manifest as one or more combinations of post-peak decline, plateau fluctuation, slow rise, local re-rise, and stabilization. Essentially, within the continuous sampling segment after the peak point, adjacent pressure changes no longer maintain the same sign, but instead simultaneously contain both positive and negative values. This identification method can separate the complex dynamic propagation process of cracks under the constraints of interlayer interfaces and the heterogeneous structure of coal and rock from simple pressure curve changes, and form an independent stage that can be used for subsequent pressure feature extraction.

[0084] The third preset condition corresponds to the identification condition of the pressure relief stage. Specifically, in multiple consecutive sampling times (N ≥ 3) after the fracturing fluid injection stops, if the pressure change of adjacent intervals is less than zero, then the continuous segment is identified as the pressure relief stage. This stage corresponds to the process after the fracturing fluid stops being continuously supplied, the fluid pressure inside the sample is released, the fluid in the fracture leaks out, and the system pressure decays. Its typical characteristic is that the pressure decreases continuously over time.

[0085] Based on the above-described stage division method, this embodiment will Figure 5 (a) W-30-1 to Figure 5(h) The original pumping pressure curves for each working condition shown in W-60-4 are transformed into stage information that can be analyzed in a structured manner. This solves the problem in existing technologies where judgment is based solely on a single fracturing pressure, peak pressure, or stable pressure, which is insufficient to reflect the dynamic propagation process of fractures under combined coal and rock conditions. Furthermore, the stage division results and characteristic parameters can be used not only for fracture propagation behavior identification but also for evaluating fracturing construction parameters, providing early warning of propagation risks, and optimizing subsequent working conditions. This provides data support for monitoring, analysis, and decision-making during the fracturing stimulation of combined coal and rock reservoirs.

[0086] V. Extracting pressure characteristic parameters based on fracturing stage identification results:

[0087] After obtaining the fracturing stage identification results, this embodiment further extracts pressure characteristic parameters to characterize the dynamic response characteristics of the fractures in the combined coal and rock samples under different working conditions. The extracted pressure characteristic parameters include at least: peak pressure; duration of the rapid pressurization stage; pressure fluctuation amplitude; pressure fluctuation frequency; pressure decay amplitude; and duration of the depressurization stage.

[0088] Among them, the peak pressure and the duration of the rapid pressurization stage mainly characterize the pressure accumulation response before and after the fracture initiation; the pressure fluctuation amplitude and the pressure fluctuation frequency mainly characterize the propagation disturbance response that occurs during the interaction between the fracture and the interlayer interface; and the pressure decay amplitude and the duration of the depressurization stage mainly characterize the fracture depressurization and release response after the completion of fracturing.

[0089] In this embodiment, Figures 5 to 8 The pumping pressure curves and related characteristic parameter variations under different injection rates and different vertical stress difference coefficients are presented. Figure 5 Eight sets of pumping pressure and triaxial stress control curves under different operating conditions are shown. Figure 6 The pump pressure variation at different injection rates is shown. Figure 7 The changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at an injection rate of 30 mL / min are shown. Figure 8 The changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at an injection rate of 60 mL / min are shown.

[0090] like Figure 6 As shown, while keeping the vertical stress difference coefficient k constant, increasing the injection rate from 30 mL / min to 60 mL / min will systematically affect the stage evolution characteristics and pressure characteristic parameters of the pumping pressure curve. Based on this pattern, this embodiment can use the injection rate as one of the operating condition input parameters, and incorporate it together with the initiation pressure, stabilization pressure, duration, and post-peak fluctuation characteristics into a unified identification framework to determine the crack propagation efficiency, the speed of main channel establishment, and the degree of interface interference.

[0091] Specifically, Figure 6 (a) shows the changes in pump pressure at different injection rates under the condition of k=1.4. Figure 6 (a) It can be seen that when the injection rate is 30 mL / min, the pump pressure curve first enters the rapid pressurization stage, and after reaching the peak, it drops significantly. Then, it remains at a relatively stable level for a long time, accompanied by some fluctuations. When the injection rate is increased to 60 mL / min, the duration of the rapid pressurization stage is similar, but the duration of the fluctuation evolution zone after the peak drop is significantly shortened.

[0092] Further integration Figure 6 (a) According to the corresponding characteristic parameters, the initiation pressure increased from 20.32 MPa to 22.11 MPa, an increase of 8.81%; the stable pressure increased from 17.71 MPa to 18.54 MPa, an increase of 4.69%; and the duration decreased from 1587 s to 551 s, a reduction of 65.28%. This indicates that under strong vertical constraints, increasing the injection rate accelerates the establishment of net pressure within the borehole, allowing the crack to initiate and enter the main propagation process in a shorter time. However, a slightly higher pressure level is required to maintain crack propagation. Therefore, in this embodiment, the combination of "short duration + higher pressure plateau" corresponding to a higher injection rate can serve as one of the important criteria for identifying the response of rapidly propagating cracks.

[0093] Figure 6 (b) shows the changes in pump pressure at different injection rates under the condition of k=1.0. Figure 6 (b) It can be seen that when the injection rate is 30 mL / min, the pump pressure curve basically skips the obvious strong fluctuation stage and enters a relatively stable slow climbing zone after the rapid pressure increase reaches the peak and falls back; while when the injection rate is increased to 60 mL / min, a more obvious pressure fluctuation appears after the peak, and then transitions to the stable stage, and the duration of the stable stage is significantly shortened.

[0094] Combination Figure 6 The changes in characteristic parameters in (b) show that the initiation pressure increased from 17.03 MPa to 18.54 MPa, an increase of 8.87%; the stabilization pressure increased from 13.01 MPa to 15.38 MPa, an increase of 18.22%; and the duration decreased from 1533 s to 549 s, a reduction of 64.19%. This indicates that under moderately high vertical constraints, the increased injection rate not only compressed the overall process duration but also significantly raised the pressure level in the post-peak evolution stage, suggesting that fracture channel formation and maintenance are more sensitive to higher flow rates. Therefore, in this embodiment, when the combined characteristic of "significant increase in stabilization pressure + significant reduction in duration" is identified, it can be further determined that the injection rate has an enhancing effect on the post-peak channel maintenance behavior under this condition.

[0095] Figure 6 (c) shows the changes in pump pressure at different injection rates under the condition of k=0.6. Figure 6 (c) It can be seen that at an injection rate of 30 mL / min, the pump pressure curve rapidly increases to the peak and then enters the post-peak evolution stage, which includes a certain degree of fluctuation and plateau change. At an injection rate of 60 mL / min, the rapid pressure increase stage is more concentrated, the pump pressure reaches the peak faster, and then drops rapidly after the peak and shows a continuous slow upward trend. The plateau evolution is more monotonous and the duration is significantly shortened.

[0096] according to Figure 6 In (c), the characteristic parameters show that the initiation pressure increased from 12.96 MPa to 13.35 MPa, an increase of 3.01%; the stable pressure increased from 9.84 MPa to 9.87 MPa, remaining essentially the same; and the duration decreased from 1479 s to 558 s, a reduction of 62.27%. This indicates that under moderately low vertical constraints, the injection rate has a weak impact on the pressure level itself, but its impact on the entire timescale is still significant. Therefore, in this embodiment, if the identification result shows "small changes in initiation pressure and stable pressure, but a significant compression of the post-peak duration," it can be determined that the injection rate mainly changes the fracture propagation efficiency, rather than significantly altering the fracture initiation threshold.

[0097] Figure 6 (d) shows the changes in pump pressure at different injection rates under the condition of k=0.2. Figure 6 (d) It can be seen that at an injection rate of 30 mL / min, the pump pressure curve quickly enters the pressurization stage and reaches the peak value. After the pressure drops after the peak, it experiences a short period of stabilization and then rises again before entering a relatively stable plateau. When the injection rate is increased to 60 mL / min, the pressurization process is more concentrated. After the pump pressure reaches the peak value, it quickly enters the stabilization stage and ends in a short period of time before entering the depressurization stage.

[0098] Combination Figure 6 The characteristic parameters in (d) show that the initiation pressure increased from 11.50 MPa to 11.70 MPa, an increase of 1.74%; the steady-state pressure increased from 8.76 MPa to 9.74 MPa, an increase of 11.19%; and the duration decreased from 1512 s to 543 s, a reduction of 64.09%. This indicates that under low vertical constraint conditions, the crack initiation threshold is still mainly determined by the triaxial stress boundary, while a higher injection rate is more reflected in the faster establishment of the post-peak plateau, a slight increase in plateau pressure, and a significant compression of the overall duration.

[0099] comprehensive Figure 6 (a) to Figure 6(d) It can be seen that, when the vertical stress difference coefficient remains consistent, the most stable and significant effect of increasing the injection rate is reflected in the overall duration of the rapid compression pressurization stage and the post-peak evolution stage. The durations of the four sets of experiments were shortened by 65.28%, 64.19%, 62.27%, and 64.09%, respectively, with the entire process at high injection rates only about 34% to 38% of the original duration. In contrast, the overall variation in initiation pressure was 1.74% to 8.87%, while the steady-state pressure generally showed an upward trend.

[0100] like Figure 7 and Figure 8 As shown, when the injection rate is fixed, as the vertical stress difference coefficient k increases, the pumping pressure curve generally shows a rise in the peak value and a shift in the post-peak plateau, indicating that enhanced vertical constraint will significantly increase the pressure level required for crack initiation and continued propagation.

[0101] Specifically, Figure 7 The changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at an injection rate of 30 mL / min are shown. Figure 7 (a) is the pump injection pressure curve. Figure 7 (b) shows the curve of characteristic parameter variation. (From...) Figure 7 (a) and Figure 7 (b) It can be seen that at an injection rate of 30 mL / min, with k=0.2 as the baseline, when k increases sequentially to 0.6, 1.0, and 1.4, the initiation pressure increases from 11.50 MPa to 12.96 MPa, 17.03 MPa, and 20.32 MPa, respectively, representing increases of 12.70%, 48.09%, and 76.70%; the stabilization pressure increases from 8.76 MPa to 9.84 MPa, 13.01 MPa, and 17.71 MPa, respectively, representing increases of 12.33%, 48.52%, and 102.17%; and the duration changes from 1512 s to 1479 s, 1533 s, and 1587 s, with changes of only -2.18%, +1.39%, and +4.96%. This indicates that under an injection rate of 30 mL / min, the vertical stress difference coefficient has a significant controlling effect on the initiation pressure and stabilization pressure, while its effect on the duration is relatively weak.

[0102] Further integration Figure 7 As can be seen from the curve shape in (a), as k increases, the pressure level of the post-peak plateau gradually shifts upward, and the post-peak fluctuation is more obvious under high k conditions, reflecting that the crack propagation resistance increases, the interface interaction is enhanced, and the channel adjustment is more frequent under stronger vertical constraints.

[0103] Figure 8 The changes in pumping pressure and characteristic parameters under different vertical stress difference coefficients at an injection rate of 60 mL / min are shown. Figure 8(a) is the pump injection pressure curve. Figure 8 (b) shows the curve of characteristic parameter variation. (From...) Figure 8 (a) and Figure 8 (b) It can be seen that at an injection rate of 60 mL / min, with k=0.2 as the baseline, when k increases to 0.6, 1.0, and 1.4, the initiation pressure increases from 11.70 MPa to 13.35 MPa, 18.54 MPa, and 22.11 MPa, respectively, representing increases of 14.10%, 58.46%, and 88.97%; the stable pressure changes from 9.74 MPa to 9.87 MPa, 15.38 MPa, and 18.54 MPa, respectively, representing increases of 1.34%, 57.91%, and 90.35%; and the duration changes from 543 s to 558 s, 549 s, and 551 s, with fluctuations of only +2.76%, +1.10%, and +1.47%. This further illustrates that at higher injection rates, the vertical stress difference coefficient also mainly controls the pressure level, while having a relatively small impact on the overall time scale. Furthermore, combined with... Figure 8 (a) It can be seen that the peak value is higher and the post-peak fluctuation is more prominent under the high k condition, indicating that under the coupling effect of strong vertical constraint and high injection rate, the crack is more likely to exhibit complex expansion behaviors such as interface deflection, reopening after local obstruction and branch adjustment.

[0104] comprehensive Figure 7 and Figure 8 It can be seen that, under both injection rates, as k increases from 0.2 to 1.4, the pumping pressure curves generally show a pattern of peak increase and post-peak pressure plateau shift, with both initiation and stabilization pressures continuously increasing, while the duration only shows slight fluctuations. That is, the injection rate more strongly regulates the time scale and post-peak evolution rhythm, while k more strongly regulates the pressure level and the intensity of interfacial interactions.

[0105] VI. Acquisition of fracture distribution data and identification of fracture interface propagation path after fracturing:

[0106] After fracturing, it is necessary to obtain the fracture distribution data of the combined coal and rock samples and identify the fracture interface propagation path based on the fracture distribution data.

[0107] In this embodiment, after the sample fracturing is completed, the injection pump and acoustic emission system are first shut off, and then the sample is unloaded in stages in the reverse order of loading to avoid premature breakage during unloading. After unloading, the distribution of post-fracturing cracks is recorded and statistically analyzed using a six-sided unfolded diagram of the sample, such as... Figure 9As shown. A fluorescent tracer pre-added to the fracturing fluid was used to assist in the observation and calibration of surface fracture orientation under ultraviolet light irradiation. Fracture distribution data included: surface fracture distribution data of the sample; internal fracture orientation inferred from surface fractures; fracture distribution locations in the upper coarse-grained sandstone, middle coal seam, and lower medium-grained sandstone; and the positional relationships of the fractures with the upper interface (coal-coarse-grained sandstone interface) and the lower interface (coal-medium-grained sandstone interface).

[0108] The following path determination information can be extracted: the intersection of the crack trajectory and the interlayer interface; whether the crack trajectory forms a continuous distribution section on both sides of the interface; whether the crack trajectory forms an extension section along the interlayer interface; whether the crack tip terminates at the interlayer interface; and whether there are two or more crack branches near the interlayer interface, with each branch extending in a different direction.

[0109] Based on the above information, this embodiment divides the crack interface propagation path into: interlayer propagation path: the crack trajectory passes through the interlayer interface and forms continuous distribution segments on both sides of the interface; interface propagation path: after the crack trajectory reaches the interlayer interface, it forms a clear extension segment along the interface; interface arrest path: the crack tip terminates at the interlayer interface, and no continuous crack segment connected to it is identified on the other side of the interface; multi-branch competing propagation path: two or more crack branches are identified near the interlayer interface, and each branch extends along different extension directions.

[0110] like Figures 10 to 13 As shown, under an injection rate of 30 mL / min, there are significant differences in the crack distribution characteristics corresponding to different k values.

[0111] for Figure 10 W-30-1 (k=1.4) shows "I"-shaped cracks in the coal seam and "Y"-shaped cracks on the surface of the coarse-grained sandstone. Combined with the continuous distribution characteristics on both sides of the interface, one of the cracks can be identified as a trans-layer extension path that crosses the upper interface; at the same time, the other path does not cross the layer and is accompanied by certain branching characteristics.

[0112] for Figure 11 W-30-2 (k=1.0) exhibits an "L" shape in its crack pattern. After reaching the lower interface, the crack extends along the interface, but no continuous cross-crack is formed on the other side of the interface. Therefore, it is identified as an extension path along the lower interface.

[0113] for Figure 12 W-30-3 (k=0.6) exhibits an "I"-shaped crack, with one horizontal crack extending along the upper interface, which can be identified as an extension path along the upper interface. It also has certain branches, indicating a competitive extension tendency near the interface.

[0114] for Figure 13W-30-4 (k=0.2) consists mainly of multiple "I"-shaped fractures extending along the interior of the coal seam, without obvious trans-layer or interface extension characteristics. It can be classified as a path type dominated by the coal seam extension and with weak interface effects.

[0115] like Figures 14 to 17 As shown, the overall complexity of the fracture increases under an injection rate of 60 mL / min.

[0116] for Figure 14 W-60-1 (k=1.4) has fractures distributed in both the coal seam and medium-grained sandstone on the left surface. Combined with the continuous fracture distribution on the surface of the medium-grained sandstone, it can be identified as a trans-layer extension path that passes through the lower interface. At the same time, there are complex fractures and secondary branches near the interface, which can further identify the multi-branch competitive extension characteristics, accompanied by local interface fracture arrest.

[0117] for Figure 15 W-60-2 (k=1.0) exhibits a combination of "X" and "Y" shaped cracks. The cracks do not penetrate the layers but terminate near the upper interface and are accompanied by branching expansion. They can be identified as crack arrest paths at the upper interface and have obvious multi-branch competitive expansion characteristics.

[0118] for Figure 16 W-60-3 (k=0.6) exhibits an "X" shape + "I" shape in its crack pattern. After the main crack reaches the upper interface, it extends along the interface, thus identifying it as an extension path along the upper interface.

[0119] for Figure 17 W-60-4 (k=0.2) consists mainly of two "I"-shaped cracks within the coal seam, which do not penetrate the layers or extend along the interface, indicating that the interface effect is not significant.

[0120] Therefore, this embodiment, through post-compression crack distribution data, can not only identify whether cracks penetrate layers, but also further identify whether the interaction between cracks and interlayer interfaces is "interface propagation" or "interface arrest," and can identify whether there is multi-branch competitive propagation near the interface. Unlike traditional methods that rely solely on pump-compression curves to infer crack patterns, this significantly improves the realism and accuracy of pattern recognition.

[0121] VII. Correlation analysis of pressure characteristic parameters, injection rate, fracture path, and target stress control data:

[0122] After acquiring pressure characteristic parameters, preset injection rate, fracture interface propagation path, and target stress control data, the above data are correlated and analyzed to form a basis for identifying hydraulic fracture propagation behavior under different working conditions, and the corresponding target hydraulic fracture propagation mode is output.

[0123] Specifically, while keeping the vertical stress difference coefficient k constant, the pressure characteristic parameter set under different injection rate conditions is mapped to the crack interface propagation path to identify the impact of injection rate changes on crack complexity, number of branches, and propagation rhythm.

[0124] Combined with Table 3 and Figure 6 It can be seen that, under the same k value, when the injection rate is increased from 30 mL / min to 60 mL / min, the crack propagation results are more likely to exhibit multi-branched, compound-oriented, or complex crack combinations. This indicates that when the injection rate increases, the volume and energy of fluid input into the sample per unit time increase, and the crack completes initiation, adjustment, and propagation in a shorter time, making it easier to form multiple competing propagation channels near the interlayer interface.

[0125] Table 3. Statistical analysis of hydraulic fracture propagation characteristics under different injection rates.

[0126]

[0127] Figure 6 (b) Under the corresponding k=1.0 condition, the cracks at 30 mL / min mainly have a simple propagation mode, while at 60 mL / min, it is easier to form a propagation result with multiple principal cracks and multiple secondary cracks coexisting. Figure 6 (a) Under the condition of k=1.4, increasing the injection rate further increases the complexity of the cracks and the number of secondary cracks; Figure 6 (d) Under the condition of k=0.2, although both injection rate conditions can result in an "I"-shaped main fracture, the number of secondary fractures increases with the higher injection rate, indicating that the increase in injection rate still promotes the improvement of fracture complexity.

[0128] Further integration Figure 6 (a) to Figure 6 As shown in the pressure curve (d), under high injection rate conditions, the rapid pressurization phase is more concentrated, the post-peak evolution phase is completed in a shorter time, and the overall duration is significantly shortened. For example, in Figure 6 (a) to Figure 6 (d) In the four operating conditions, after the injection rate was increased from 30 mL / min to 60 mL / min, the duration was shortened by 65.28%, 64.19%, 62.27%, and 64.09%, respectively. Therefore, in this embodiment, "significantly shortened duration, concentrated post-peak fluctuations, and increased number of crack branches" can be regarded as a type of identification feature of operating conditions dominated by high injection rates.

[0129] It should be noted that while an increased injection rate significantly alters the complexity and competition of fracture propagation, it does not necessarily determine whether a fracture will penetrate a layer. Whether a fracture penetrates an interlayer interface must be determined in conjunction with the vertical stress difference coefficient k in the target stress control data. Therefore, in this embodiment, the injection rate is primarily used as a working condition parameter characterizing fracture propagation efficiency, branch competition intensity, and post-peak evolution rhythm in mode identification.

[0130] While keeping the injection rate constant, the pressure characteristic parameter set under different vertical stress difference coefficients k is mapped to the crack interface propagation path to identify the control effect of vertical stress constraints on crack interface behavior and propagation mode.

[0131] Combined with Table 4, Figure 7 and Figure 8 It can be seen that, under the same injection rate, as the vertical stress difference coefficient k increases, the fracture path gradually changes from relatively simple propagation within the coal seam to propagation along the interface, interface arrest, or cross-layer propagation. This indicates that as the vertical stress difference coefficient k increases, the vertical constraint on fracture propagation is enhanced, and the stress redistribution and propagation resistance near the interface become more pronounced, thereby strengthening the interaction between the fracture and the interlayer interface.

[0132] Table 4. Statistics on hydraulic crack propagation characteristics under different vertical stress difference coefficients k.

[0133]

[0134] For example, at an injection rate of 30 mL / min, Figure 7 (a) and Figure 7 (b) shows that as k increases from 0.2 to 1.4, the initiation pressure increases sequentially from 11.50 MPa to 12.96 MPa, 17.03 MPa, and 20.32 MPa, while the stabilization pressure increases sequentially from 8.76 MPa to 9.84 MPa, 13.01 MPa, and 17.71 MPa, although the duration of cracking changes relatively little. Correspondingly, the crack path gradually changes from simple propagation to interface deflection, enhanced interface effects, and even local per-layer cracking.

[0135] At an injection rate of 60 mL / min Figure 8 (a) and Figure 8 (b) Similarly, it is shown that as k increases, the initiation pressure increases sequentially from 11.70 MPa to 13.35 MPa, 18.54 MPa, and 22.11 MPa, while the stabilization pressure changes sequentially from 9.74 MPa to 9.87 MPa, 15.38 MPa, and 18.54 MPa, although the duration remains relatively constant. The corresponding crack path results show that under higher k values, complex interface responses such as interface arrest, interface crossing, and multi-branch propagation are more likely to occur.

[0136] Therefore, in this embodiment, "high initiation pressure, high stabilization pressure, enhanced post-peak fluctuations, and significant crack interface effects" can be used as a type of identification characteristic for high vertical constraint conditions. In other words, the vertical stress difference coefficient k is mainly used to characterize the pressure level required for crack propagation and the coupling strength between the crack and the interlayer interface, rather than primarily to characterize the duration of the entire process.

[0137] In this embodiment, for ease of engineering use, the above-mentioned correlation analysis results can be organized into a rule base, a tag base, or a training sample set. One implementation method is to establish a data record item containing the following fields for each working condition: pressure characteristic parameter group; injection rate parameter; target stress control data; fracture interface propagation path label; target hydraulic fracture propagation mode label.

[0138] The aforementioned data records can be stored in a database or pattern recognition module for matching and identification of subsequent new operating conditions. For the operating condition to be identified, the pumping pressure data acquired in real time or in history can be extracted into a set of pressure characteristic parameters. These parameters are then combined with the corresponding injection rate and target stress control data to perform similarity matching, rule judgment, or classification identification with the established data records, thereby outputting the target hydraulic fracture propagation pattern.

[0139] Based on the above correlation analysis results, this embodiment can further classify the fracture propagation behavior under different working conditions into several identifiable target hydraulic fracture propagation modes, and use the corresponding modes as the output results of fracturing effect evaluation and parameter optimization.

[0140] In this embodiment, the target hydraulic fracture propagation mode does not refer solely to the geometry itself, but rather to a category of propagation behavior characterized by pressure response features, injection rate, target stress control data, and fracture interface propagation path. Specifically, it may include, but is not limited to, the following modes:

[0141] The dominant trans-layer propagation mode corresponds to a crack trajectory that crosses the interlayer interface, forming continuous crack distribution sections on both sides of the interface. For this type of mode, its trans-layer characteristics can be directly determined through crack path identification results, and further confirmation that it is trans-layer propagation behavior formed under high resistance conditions can be achieved by combining high initiation pressure and stabilization pressure. For example, in Figure 6 (a) Under the corresponding high-k operating conditions, and Figure 7 and Figure 8As shown in the high vertical stress difference condition, cracks can exhibit propagation results that cross the upper or lower interface. This type of mode typically corresponds to higher initiation pressure and higher stabilization pressure, indicating that under strong vertical constraints, the cracks require a higher driving force to overcome interfacial resistance and achieve interlayer extension. In practical applications, identifying the dominant interlayer propagation mode indicates that the current parameter combination has a strong interlayer modification capability, making it suitable for fracturing scenarios that require opening up adjacent layers or improving vertical connectivity.

[0142] The dominant mode of propagation along the interface occurs when a crack, after reaching the interlayer interface, does not directly cross it but continues to extend along the interface direction. For this type of mode, the direction of propagation along the upper or lower interface can be determined by identifying the crack's interface propagation path, and confirmed by combining this with a moderate-level pressure plateau and certain post-peak fluctuation characteristics. For example, Figure 7 The corresponding medium k working conditions, and Figure 8 In operating conditions near k=0.6, fractures are more likely to deflect and extend along the interface after contact. This pattern indicates that the interface is the preferred path for fracture propagation under current conditions; although fractures have the ability to extend, they are not yet capable of stably crossing the interface. In practical applications, identifying the dominant mode of propagation along the interface can be used to assess whether fracturing stimulation tends to propagate within the layer or through interface slip, thus providing a basis for fracture network morphology control and prediction of the extent of layer stimulation.

[0143] Interfacial arrest-dominant mode is characterized by fracture tips terminating at the interlayer interface, with no continuous fracture segment forming on the other side of the interface. For this type of mode, the absence of effective extension on the other side of the interface can be determined by fracture path results, combined with enhanced post-peak fluctuations and significant local adjustments without continuous penetration in the pressure response characteristics. For example, under certain high-k or high-flow-rate conditions, although the fracture reaches the vicinity of the interface and undergoes some degree of interfacial interaction, it fails to achieve complete penetration due to interfacial obstruction or local stress redistribution, thus exhibiting arrest. In this case, the pump pressure curve typically shows some fluctuations in the post-peak stage, reflecting repeated adjustments at the fracture tip near the interface. In practical applications, after identifying the interfacial arrest-dominant mode, it can be determined that the existing parameter combination is insufficient to achieve stable penetration, requiring adjustments to the injection rate, volume, or stress control strategy to improve cross-interfacial extension capabilities.

[0144] A multi-branch competitive propagation dominant mode corresponds to the formation of two or more branches during fracture propagation, with these branches competing to propagate in different directions. This type of mode can be identified by combining the number of primary fractures, secondary fractures, and their distribution characteristics in fracture images, and further confirmed by pressure response characteristics such as concentrated fluctuations and compressed duration during the post-peak evolution stage. For example, under higher injection rates, some conditions exhibit complex fractures, coexistence of multiple primary fractures, or simultaneous development of multiple secondary fractures. This type of mode indicates that after a high fluid energy input in a short period, fractures are more likely to form multiple propagation channels and compete for propagation. In practical applications, identifying the multi-branch competitive propagation dominant mode can be used to assess the promoting effect of the current parameter combination on fracture network complexity, thereby providing a basis for unconventional reservoir volume stimulation, fracture network enhancement, and parameter optimization.

[0145] In this embodiment, the identification results of the target hydraulic fracture propagation mode can be output in any one or more of the following ways: outputting the corresponding mode category label; outputting the fracture interface propagation path corresponding to the mode category; outputting the pressure characteristic parameter range corresponding to the mode; outputting the similarity or matching degree between the current working condition and the known mode sample; and outputting parameter adjustment suggestions to guide the optimization of subsequent fracturing operations.

[0146] For example, when the working condition to be identified is characterized by "high initiation pressure, high stable pressure, obvious post-peak fluctuations and crack crossing the interface", the output can be "cross-layer propagation dominant mode"; when the working condition to be identified is characterized by "significantly shortened duration, increased number of branches and concentrated post-peak fluctuations", the output can be "multi-branch competitive propagation dominant mode"; when the working condition to be identified is characterized by "medium pressure level, obvious interface deflection and extension along the interface", the output can be "interface propagation dominant mode".

[0147] In summary, this application integrates pump pressure time-series data, injection rate, target stress control data, and fracture path data into a unified identification framework, avoiding one-sided judgments based solely on a single pump pressure index or a single fracture morphology. Through pressure feature parameter extraction and correlation analysis, it achieves a structured expression of fracture propagation behavior, facilitating the formation of reusable, comparable, and scalable data models. It can distinguish the impact of injection rate on fracture complexity and propagation rhythm, as well as the impact of the vertical stress difference coefficient k on pressure level and interfacial interaction intensity, thereby improving the accuracy of pattern recognition.

[0148] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of this application.

Claims

1. A combined method for identifying hydraulic fractures in coal and rock, characterized in that, include: Obtain composite coal and rock sample data; The composite coal and rock sample was placed in a true triaxial loading space, and triaxial principal stress was applied to the composite coal and rock sample to obtain target stress control data; among which, the target stress control data included the maximum horizontal principal stress, the minimum horizontal principal stress, the vertical principal stress, and the vertical stress difference coefficient determined by the vertical principal stress. Based on the target stress control data and the preset injection rate, fracturing fluid is injected into the fracturing borehole in the combined coal and rock sample, and the fracturing process response data is obtained; wherein, the fracturing process response data includes at least the pumping pressure data; The pump injection pressure data is divided into stages to obtain fracturing stage identification results, and pressure characteristic parameters are extracted based on the fracturing stage identification results; After hydraulic fracturing, fracture distribution data of the combined coal and rock samples were acquired, and fracture interface propagation paths were identified based on the fracture distribution data. Correlation analysis was performed on pressure characteristic parameters, preset injection rate, fracture interface propagation path, and target stress control data to identify hydraulic fracture propagation modes of composite coal and rock samples under different vertical stress constraints and different injection rates.

2. The combined coal and rock hydraulic fracture identification method according to claim 1, characterized in that: The combined coal and rock sample data includes: the layered structure of the sample, lithological data, interlayer interface data, and fracturing pore data; The composite coal and rock sample includes at least: an upper rock layer, a middle coal seam, and a lower rock layer arranged vertically in sequence; wherein at least one of the upper rock layer and the lower rock layer has different lithological or mechanical properties from the middle coal seam, so as to form a composite coal and rock structure with obvious interlayer interfaces and heterogeneity. Interlayer interface data includes interlayer interface data between the upper rock strata and the middle coal seam, as well as interlayer interface data between the middle coal seam and the lower rock strata; The fracturing borehole data includes fracturing borehole data located in the middle coal seam, as well as fracturing pipe data installed in the fracturing borehole and sealed.

3. The combined coal and rock hydraulic fracture identification method according to claim 2, characterized in that: The target stress control data are obtained, including: Set target values ​​for the maximum and minimum horizontal principal stresses, and keep the maximum and minimum horizontal principal stresses constant; Adjust the loading value of the vertical principal stress to construct multiple triaxial stress combinations under different vertical stress constraint states; Based on each triaxial stress combination, the corresponding vertical stress difference coefficient is calculated; By matching the triaxial stress combination with the corresponding vertical stress difference coefficient, target stress control data corresponding to different vertical stress constraint states are obtained. The vertical stress difference coefficient characterizes the degree of deviation of the vertical principal stress from the maximum and minimum horizontal principal stresses.

4. The combined coal and rock hydraulic fracture identification method according to claim 3, characterized in that: Acquire response data during the fracturing process, including: Determine the current fracturing conditions based on the target stress control data; Fracturing fluid is continuously injected into the fracturing borehole at a constant injection rate or a staged injection rate to form a fracturing loading process coupled with a true triaxial stress environment. During the injection process, data on the change of pumping pressure over time is collected, and the data on the change of pumping pressure over time is matched with the corresponding injection time to obtain the pumping pressure data; By correlating the injection rate, current fracturing conditions, and pump pressure data, fracturing process response data is generated.

5. The combined coal and rock hydraulic fracture identification method according to any one of claims 2 to 4, characterized in that: The fracturing stage identification results are obtained, including: The pump pressure data were arranged in time sequence according to the injection time to obtain the pressure change sequence corresponding to each sampling time. Based on the pressure change sequence, calculate the pressure change between adjacent sampling times; Identify the continuous time segment before the peak point where the pressure change meets the first preset condition as the rapid pressurization stage; Identify the continuous time segment after the peak point where the pressure change meets the second preset condition, and use it as the post-peak evolution stage; Identify the continuous time segment in which the pressure change after stopping the injection of fracturing fluid meets the third preset condition, and use it as the pressure relief stage; The rapid pressurization stage, the post-peak evolution stage, and the depressurization stage are output as the fracturing stage identification results.

6. The method for identifying combined coal and rock hydraulic fractures according to claim 5, characterized in that: The first preset condition is: the pressure change at consecutive sampling times is greater than zero, and the corresponding pressure value does not exceed the peak point in the pump pressure data; The second preset condition is: in multiple consecutive sampling times after the peak point, the adjacent pressure changes include both positive and negative values, so as to characterize the post-peak fluctuation evolution of the crack under the constraint of the interlayer interface. The third preset condition is: in multiple consecutive sampling moments after the fracturing fluid injection is stopped, the pressure change is less than zero in each adjacent moment.

7. The combined coal and rock hydraulic fracture identification method according to claim 5, characterized in that: Pressure characteristic parameters are extracted based on the fracturing stage identification results, including: Peak pressure and duration of rapid pressurization phase are extracted based on the rapid pressurization phase to characterize the pressure accumulation response before and after crack initiation; The amplitude and frequency of pressure fluctuations are extracted based on the post-peak evolution stage to characterize the extended disturbance response during the interaction between the crack and the interlayer interface. The pressure attenuation amplitude and duration of the decompression phase are extracted to characterize the fracture decompression release response after fracturing. Peak pressure, duration of rapid pressurization phase, pressure fluctuation amplitude, pressure fluctuation frequency, pressure decay amplitude, and duration of depressurization phase are used as pressure characteristic parameters.

8. The method for identifying combined coal and rock hydraulic fractures according to claim 5, characterized in that: Identifying crack interface propagation paths based on crack distribution data includes: Acquire fracture distribution data of the combined coal and rock samples after fracturing. The fracture distribution data includes fracture distribution data on the sample surface and / or fracture distribution data inside the sample. Based on fracture distribution data, the distribution locations of fractures in the upper rock strata, middle coal seam, and lower rock strata were identified, as well as the positional relationships between the fractures and the interlayer interfaces between the upper rock strata and the middle coal seam, and between the middle coal seam and the lower rock strata. Based on the distribution of cracks in each layer and the positional relationship between cracks and interlayer interfaces, crack interface propagation paths are classified; among them, crack interface propagation paths include cross-layer propagation paths, along-interface propagation paths, interface-arrested crack paths, and multi-branch competing propagation paths.

9. The method for identifying combined coal and rock hydraulic fractures according to claim 8, characterized in that: The crack interface propagation paths are classified, including: Extract the intersection of the crack trajectory with the interlayer interface, the continuous distribution of the crack trajectory on both sides of the interlayer interface, the extension section of the crack trajectory along the interlayer interface, the termination position of the crack tip at the interlayer interface, and the number of crack branches near the interlayer interface. When a crack trajectory crosses the interlayer interface and forms a continuous distribution segment on both sides of the interlayer interface, the corresponding crack interface propagation path is determined as the cross-layer propagation path. When the crack trajectory reaches the interlayer interface and forms an extension segment along the interlayer interface, the corresponding crack interface extension path is determined as the interface extension path. When the tip of the crack trajectory terminates at the interlayer interface, and no continuous distribution segment connected to the corresponding crack trajectory is identified on the other side of the interlayer interface, the corresponding crack interface propagation path is determined as the interface crack arrest path. When two or more crack branches are identified near the interlayer interface, and each crack branch extends in a different direction, the corresponding crack interface extension path is determined as a multi-branch competitive extension path.

10. A computer-readable storage medium storing computer instructions that, when executed by a processor, implement the method as described in any one of claims 1 to 9.

Images