A method and device for determining the effect of a fracturing reconstruction with double intelligent correction

Abstract

The application provides a method and device for determining fracturing reconstruction effect through double intelligent correction, which comprises the following steps: collecting multi-source data of a target block to form a three-dimensional geomechanical model of the target block; searching for an optimal parameter combination in a high-dimensional uncertainty space according to a simulated fracture geometry and actual pump injection pressure data, automatically performing history matching on a pump pressure curve to obtain a once-corrected fracture geometry; simulating production dynamic data and reservoir reconstruction volume through a numerical simulation model of the target block and a production system, establishing a reservoir simulation proxy model, and inverting reservoir characteristic parameters and the fracture geometry as a twice-corrected fracture geometry; and predicting long-term productivity and effective reconstruction volume of the target block according to the reservoir characteristic parameters and the twice-corrected fracture geometry. In the above scheme, the accuracy of the fracture geometry and productivity prediction is improved through double intelligent correction and full-process coupling.

Description

Technical Field

[0001] This application belongs to the field of oil and gas exploration technology, and in particular relates to a method and device for determining the fracturing effect with dual intelligent correction. Background Technology

[0002] With the increasing energy demand and rapidly changing energy landscape, ensuring stable energy supply necessitates a focus on research into unconventional oil and gas resources, and increasing the development and utilization of unconventional natural gas is an inevitable choice for energy security. Unconventional reservoirs typically require hydraulic fracturing to reduce seepage resistance, increase drainage area, and improve oil and gas production. However, in actual fracturing operations, unconventional reservoirs present complex conditions such as strong heterogeneity, disordered natural fractures, and the unavailability of formation oil and gas, which inevitably lead to technical problems such as inaccurate parameter prediction, lag in dynamic control, and large errors in manual parameter adjustment.

[0003] Intelligent fracturing technology deeply integrates oil and gas reservoir fracturing and well completion with intelligent algorithms. It uses different types of algorithms to extract fracturing characteristic parameters, characterize high-dimensional complex nonlinear relationships, find the optimal combination of fracturing parameters, dynamically control the pumping program, and evaluate the fracturing operation effect.

[0004] However, existing post-fracturing fracture characterization and volumetric assessment involve complex and cumbersome geological and engineering data, and post-fracturing fracture morphology is disconnected from subsequent production prediction and drainage analysis. Existing intelligent fracturing methods suffer from high data processing complexity, weak model generalization ability, strong reliance on expert experience, and delayed decision response, which severely restricts the refinement and intelligentization of fracturing design.

[0005] There is currently no effective solution to the above problems. Summary of the Invention

[0006] The purpose of this application is to provide a method and apparatus for determining the fracturing effect with dual intelligent correction, which can achieve efficient and accurate prediction of the production capacity of the target block.

[0007] This application provides a method and apparatus for determining the fracturing effect with dual intelligent correction, which is implemented as follows: A method for determining the fracturing effect with dual intelligent correction, the method comprising: Collect multi-source data of the target block, calculate mechanical parameters using the multi-source data, and obtain discrete well point data of the target block; The discrete well point data is interpolated into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; Based on the fracturing construction parameters of the target block and the dual medium model of pores and fractures, natural fractures are implanted into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing construction parameters, a numerical model of fracture propagation at the mine scale is established to output the simulated fracture geometry. Based on the simulated crack geometry and the actual pumping pressure data, the optimal parameter combination is searched in the high-dimensional uncertainty space, and the pumping pressure curve is automatically historically fitted to obtain the crack geometry after one correction. The corrected fracture geometry is coupled to the reservoir numerical model through an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. By using a numerical simulation model of the target block reservoir and the production regime, dynamic production data and reservoir stimulation volume were obtained through simulation. Based on the simulated production dynamics data and actual production history data, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry. The inverted fracture geometry is used as the fracture geometry after secondary correction. Based on reservoir characteristic parameters and fracture geometry after secondary correction, the long-term productivity and effective stimulation volume of the target block are predicted.

[0008] In one implementation, a reservoir simulation proxy model is established based on simulated production dynamic data and actual production history data to invert reservoir characteristic parameters and fracture geometry, including: Based on the distribution range of prior parameters, an initial sample set is generated using the Latin hypercube sampling method. The prior parameters include at least one of the following: reservoir porosity, permeability, fracture conductivity, and fracture half-length. The initial sample set is input into a mesh model containing fractures, the corresponding production dynamic data is calculated, and batch numerical simulations are performed on the production dynamic data to construct simulation training data of parameter combinations and production response. The production dynamic data includes at least one of the following: bottom hole flowing pressure, daily gas production, and daily water production. Using the simulated training data, a multi-response parameter surrogate model is constructed using a neural network algorithm, wherein the multi-response parameter surrogate model is used to predict production dynamic data under a given combination of parameters; The multi-response parameter surrogate model is combined with Markov chain Monte Carlo to establish an automatic history fitting model. In the automatic history fitting model, the actual production history data is used as a constraint, and the parameter combination with the highest matching degree with the actual production history data is searched in the parameter space through a preset sampling method. Based on the preset convergence criteria, it is determined whether sampling needs to continue. If the convergence criteria are not met, sampling is performed again for the next round of simulation calculation and proxy model update. If the convergence criteria are met, the parameter combination that best matches the actual production historical data is taken as the optimal historical fitting solution set. The optimal historical fitting solution set includes at least one of the following: posterior parameter distribution characteristics, optimal parameter combination and its uncertainty range, and historical fitting quality evaluation index. By fitting the optimal solution set in history, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry.

[0009] In one implementation, multi-source data of the target block is collected, and mechanical parameters are calculated using the multi-source data to obtain discrete well point data of the target block, including: Collect geological data, logging data, drilling data, and completion data for the target block. The geological data includes at least one of the following: regional structure, reservoir geological description, stratigraphic information, and fault / natural fracture distribution. The logging data includes at least one of the following: geostress and geological parameters. The drilling data includes at least one of the following: wellhead coordinates, wellbore trajectory, and casing program. The completion data includes at least one of the following: perforation information, fracturing design, construction curves, flowback data, and gas testing data. Geological data, well logging data, drilling data, and well completion data are used as multi-source data to calculate mechanical parameters, wherein the mechanical parameters include at least one of the following: vertical depth, sounding depth, Young's modulus, Poisson's ratio, and in-situ stress.

[0010] In one implementation, based on the fracturing operation parameters of the target block and a dual-medium model of pores and fractures, natural fractures are embedded into the three-dimensional geomechanical model through a discrete fracture network model. A field-scale fracture propagation numerical model is then established using the fracturing operation parameters to output the simulated fracture geometry, including: A discrete crack network model conforming to statistical laws was generated using the Monte Carlo method. The generated discrete fracture network model is embedded into a three-dimensional geomechanical model to obtain a numerical model of fracture propagation at the mine scale. In the numerical model of crack propagation at the mine scale, natural cracks are regarded as frictional interfaces without flow conductivity. The propagation mode of hydraulic cracks under the interference of natural cracks is determined based on the crack intersection criterion. The extension direction of hydraulic cracks after intersecting with natural cracks is determined by analyzing the stress state near the tip of the hydraulic crack. If the maximum tensile stress on one side of the natural crack is greater than or equal to the tensile strength of the rock, the hydraulic crack will pass through the natural crack and continue to propagate. If the shear stress and normal stress acting on the natural crack satisfy the linear friction law, the hydraulic crack is captured by the natural crack and deflected into the natural crack.

[0011] In one implementation, based on the fracturing operation parameters of the target block and a dual-medium model of pores and fractures, natural fractures are embedded into the three-dimensional geomechanical model through a discrete fracture network model. A field-scale fracture propagation numerical model is then established using the fracturing operation parameters to output the simulated fracture geometry, including: The rock deformation and fluid flow at each time step are iteratively coupled and solved in the following manner to obtain the fluid pressure distribution: Repeat the following steps until the model converges: make a preliminary estimate of the fluid pressure inside the fracture; use the rock elastic deformation theory to describe the displacement discontinuity surface, and use the preliminary estimated fluid pressure inside the fracture to calculate the displacement discontinuity surface; after obtaining the displacement discontinuity surface, use the Newton-Raphson iteration method to calculate the fluid pressure distribution, and compare the calculated pressure with the preliminary estimated pressure to verify the convergence; The mass balance equation is used to ensure that the total injected fluid volume is equal to the sum of the fluid volume in the crack and the fluid loss volume; If the time step does not meet the calculation requirements, the fluid pressure inside the crack is estimated and the calculation is performed for the next time step. If the time step meets the calculation requirements, the iterative coupling solution ends.

[0012] In one embodiment, a preliminary estimate of the fluid pressure within the crevice is made, including: Calculate the pressure drop within a single crack using the following formula:

[0013] in, This represents the pressure drop within a single crack. Indicates internal pressure. Indicates the distance along the crack. Indicates the consistency index. Represents the power-law exponent of fluid. Indicates the crack height. Indicates the crack width. Indicates fluid flow rate; Based on the injection displacement of each fracture, the total injection displacement is calculated using the following formula:

[0014] in, Indicates the total injection displacement. Let N represent the injection displacement of the i-th fracture, and N represent the number of fractures. Calculate the fracture inlet pressure using the following formula:

[0015] in, The inlet pressure of the crack. Indicates wellhead pressure, For the perforation friction voltage drop, For wellbore friction pressure drop, Let be the perforation friction pressure drop at the i-th crack. For the density of the proppant slurry, The density of water, Let be the number of perforations at the i-th crack. Let be the diameter of the perforation at the i-th crack. Let be the wellbore friction pressure drop at the i-th fracture. Indicates the wellbore friction coefficient. Indicates the location of the crack. Represents the power-law exponent of fluid. This indicates the flow rate in the current section of the wellbore. Based on the pressure drop within the single fracture, the total injection volume, and the fracture inlet pressure, the pressure drop along the fracture is calculated using lubrication theory, and then fully coupled with rock deformation for iterative solution to obtain the fluid pressure within the fracture.

[0016] In one embodiment, the displacement discontinuity is calculated using a preliminary estimate of the fluid pressure within the gap, including: Divide the boundary into multiple planar units; Calculate the stress generated by the displacement discontinuity of each planar element; The stresses generated by the displacement discontinuities of each plane element in multiple plane elements are linearly combined to obtain the stress at any point.

[0017] In one embodiment, the stresses generated by the displacement discontinuities of each planar element in a plurality of planar elements are linearly combined to obtain the stress at any point, including: Calculate the stress generated at any point within the region due to the displacement discontinuity at point η on the boundary using the following formula:

[0018] in, Let η be the displacement discontinuity at point η on the boundary, which generates stress at point ξ within the region. For displacement, This is the tensor field representing the effect of concentrated force on displacement. Here, j and k represent the component coordinate indices of the stress tensor, which are the integration boundaries.

[0019] In one implementation, calculating the stress generated by the displacement discontinuity of each planar element includes: Under given boundary conditions, the stresses generated by the normal displacement discontinuity and the shear displacement discontinuity of each planar element are obtained by solving the following system of equations:

[0020] in,

[0021] in, Let be the total normal stress generated by the surface displacement discontinuity of planar element j. Let J be the total shear stress generated by the surface displacement discontinuity of planar element j. This represents the discontinuity in the normal displacement of planar element j, corresponding to the crack opening. The shear displacement discontinuity of planar element j represents the relative slippage between crack surfaces, and N is the number of planar elements. Let be the elastic influence coefficient matrix under plane strain. Let be the normal stress response induced at plane element i by the discontinuity of shear displacement in plane element j. Let be the distance between the center of crack element i and the center of crack element j. The height of the crack. and These are empirical constants determined to match the actual 3D crack solution.

[0022] In one embodiment, after calculating the stress generated by the displacement discontinuity of each planar element, the method further includes: The relative net pressure is calculated as the stress shadowing effect according to the following formula, wherein the stress shadowing effect is used to characterize the induced stress compression of the internal crack by the external cracks on both sides.

[0023] in, Indicates relative net pressure. This indicates the net pressure inside the crack. This represents the minimum horizontal principal stress. This represents the maximum horizontal principal stress. The stress shadowing effect optimizes the stress generated by the displacement discontinuity of each planar element.

[0024] A dual-intelligent correction device for determining the fracturing effect includes: The acquisition module is used to acquire multi-source data of the target block, calculate mechanical parameters through the multi-source data, and obtain discrete well point data of the target block. An interpolation module is used to interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block. The output module is used to embed natural fractures into the three-dimensional geomechanical model through a discrete fracture network model based on the dual medium model of pores and fractures using the fracturing construction parameters of the target block, and to establish a field-scale fracture propagation numerical model in combination with the fracturing construction parameters, so as to output the simulated fracture geometry. The primary correction module is used to find the optimal parameter combination in a high-dimensional uncertainty space based on the simulated crack geometry and the actual pumping pressure data, and to automatically fit the pumping pressure curve to obtain the crack geometry after primary correction. A module is established to couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model, thereby establishing a numerical simulation model of the target block reservoir. The simulation module is used to simulate production dynamics data and reservoir stimulation volume using a numerical simulation model of the target block reservoir and production regime. The secondary correction module is used to establish a reservoir simulation proxy model based on the simulated production dynamic data and the actual production history data, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. The prediction module is used to predict the long-term productivity and effective stimulation volume of the target block based on reservoir characteristic parameters and fracture geometry after secondary correction.

[0025] A computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0026] The method and apparatus for determining the fracturing effect with dual intelligent correction provided in this application collect multi-source data of the target block, calculate mechanical parameters through the multi-source data to obtain discrete well point data of the target block; interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; based on the dual medium model of pores and fractures, using the fracturing construction parameters of the target block, embed natural fractures into the three-dimensional geomechanical model through a discrete fracture network model, and establish a field-scale fracture propagation numerical model in combination with the fracturing construction parameters to output the simulated fracture geometry; based on the simulated fracture geometry and actual pumping pressure data, search for... The optimal parameter combination is found, and the pump pressure curve is automatically historically fitted to obtain the fracture geometry after primary correction. This primary corrected fracture geometry is then coupled to the reservoir numerical model via an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. Using the target block reservoir numerical simulation model and production regime, dynamic production data and reservoir stimulation volume are simulated. Based on the simulated dynamic production data and actual historical production data, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry. The inverted fracture geometry is used as the secondary corrected fracture geometry. Based on the reservoir characteristic parameters and the secondary corrected fracture geometry, the long-term production capacity and effective stimulation volume of the target block are predicted. In other words, geological modeling, fracture propagation, fracture correction, reservoir coupling, parameter inversion, and production capacity assessment form a complete closed loop. The first correction of fracture morphology is performed through pump pressure history fitting, and the second correction of fracture parameters is performed through production history fitting. Through dual intelligent correction and full-process coupling, the accuracy of fracture morphology and production capacity prediction is significantly improved, while reducing manual intervention and computational costs. This enables rapid, accurate, and quantitative evaluation of fracturing stimulation effects and guides on-site fracturing optimization. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart of one embodiment of the rapid post-pressure effect evaluation method based on the integration of geology, engineering, and intelligence provided in this application; Figure 2 This is a schematic diagram of the iterative scenario provided in this application; Figure 3 This is a flowchart of the method for automatically inverting reservoir parameters and fracture geometry parameters provided in this application; Figure 4 This is a schematic diagram of the post-pressure evaluation system based on the integration of geology, engineering, and intelligence provided in this application; Figure 5 This is a flowchart of one embodiment of the method for determining the fracturing effect of dual intelligent correction provided in this application; Figure 6 This is a hardware structure block diagram of an electronic device for a dual intelligent correction method for determining the fracturing effect provided in this application; Figure 7 This is a schematic diagram of the module structure of one embodiment of the fracturing modification effect determination device with dual intelligent correction provided in this application. Detailed Implementation

[0029] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0030] It should also be noted that in the embodiments of this specification, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0031] This example provides a rapid post-pressure effect evaluation method based on geological-engineering-intelligent integration, which may include: Step 1: Collect data from the target block to establish an integrated geological and engineering shared database; establish a refined three-dimensional geomechanical model based on well logging empirical formulas and Kriging interpolation algorithms; Specifically, geological data, logging data, drilling data, completion data, and production data of the target block can be collected. A detailed three-dimensional geomechanical model can be established based on logging data such as vertical depth, depth sounding, Young's modulus, Poisson's ratio, and geostress.

[0032] Step 2: Establish a mine-scale crack propagation model based on the displacement discontinuity method; the crack propagation model is used to simulate and calculate crack morphology; Specifically, the physical mechanisms of stress, fluid, and fractures can be coupled based on a pore-fracture dual-medium model; natural fractures can be incorporated into a discrete fracture network model, including the number of groups, number of fractures, density, azimuth, length, width, and permeability of natural fractures.

[0033] Step 3: Establish a crack propagation proxy model based on machine learning algorithms; the crack propagation proxy model is used for automatic correction of crack geometry. Specifically, the actual pumping pressure can be automatically fitted and the fracture geometry corrected based on a fracture propagation proxy model; an automatic historical fitting workflow for pump pressure can be formed based on the proxy model. Alternatively, actual production data can be automatically fitted and reservoir and fracture characteristic parameters can be inverted based on a reservoir simulation proxy model; an automatic historical fitting workflow for production capacity can be formed based on the proxy model.

[0034] Step 4: Couple the fracture propagation model with the reservoir simulation model based on the embedded discrete fracture model; That is, the corrected fracture morphology and natural fracture distribution can be implanted into the reservoir simulation model based on the embedded discrete fracture model.

[0035] Step 5: Predict production capacity and reservoir stimulation volume based on reservoir simulation model; Step 6: Establish a reservoir simulation proxy model based on machine learning algorithms; the reservoir simulation proxy model is used to invert reservoir and fracture characteristic parameters.

[0036] Step 7: Predict long-term productivity and effective stimulation volume based on the second revised fracture and reservoir characteristic parameters.

[0037] Specifically, the optimal solution set can be selected based on minimizing the global error as the objective function, and the optimal solution set can be used in the reservoir simulation model to predict long-term production capacity and effective stimulation volume according to the method described above.

[0038] The following section provides a detailed explanation of the aforementioned rapid post-pressure effect evaluation method based on the integration of geology, engineering, and intelligence. Figure 1 As shown, a rapid evaluation method for post-pressure effects based on integrated geology, engineering, and intelligence can include the following steps: Step 101: Collect data from the target block to establish an integrated geological and engineering shared database, and establish a three-dimensional geomechanical model based on well logging empirical formulas.

[0039] Specifically, the collected multi-source data can be standardized to establish an integrated shared database for geological engineering. Mechanical parameters are calculated based on well logging empirical formulas (e.g., using sonic logging to calculate dynamic Young's modulus and Poisson's ratio). Then, discrete well point data are interpolated into continuous three-dimensional spatial data to establish a three-dimensional geomechanical model, providing a geomechanical background for subsequent fracture propagation simulation. This three-dimensional geomechanical model includes the three-dimensional geological property distribution of the target block (e.g., porosity, permeability, Young's modulus, Poisson's ratio, in-situ stress, etc.).

[0040] The aforementioned collection of target block data to establish an integrated geological and engineering shared database can include the collection of geological data, logging data, drilling data, completion data, and production data for the target block. Specifically, the aforementioned geological data may include, but is not limited to: regional structure, reservoir geological description, stratigraphic information, and fault / natural fracture distribution; the aforementioned logging data may include: geostress, geomechanical parameters, etc.; the aforementioned drilling data may include: wellhead coordinates, wellbore trajectory, casing program, etc.; the aforementioned completion data may include: perforation information, fracturing design, construction curves, flowback and testing data, etc.; and the aforementioned production data may include: production curves, cumulative production data, etc.

[0041] In one implementation, a three-dimensional geomechanical model can be established based on well logging data such as vertical depth, depth sounding, Young's modulus, Poisson's ratio, and geostress. Specifically, a refined three-dimensional geomechanical model can be established by combining geological data of the study area (e.g., regional structure, microstructure, reservoir description, stratigraphic information, fault / natural fracture distribution, etc.), well logging data (e.g., geostress, geomechanical parameters, etc.), drilling data (e.g., wellhead coordinates, wellbore trajectory, casing program, etc.), and completion data (e.g., perforation information, fracturing design, construction curves, flowback and gas testing data, etc.).

[0042] Step 102: Based on the above three-dimensional geomechanical model and fracturing construction parameters, establish a mine-scale fracture propagation model, wherein the fracture propagation model is used to simulate and calculate the fracture morphology.

[0043] That is, the input to this step is the aforementioned three-dimensional geomechanical model and fracturing construction parameters (e.g., displacement, sand volume, fluid volume, perforation parameters, etc.), and the output is the simulated fracture morphology, including fracture length, width, and height. Specifically, based on the pore-fracture dual-medium model, the fluid-structure interaction problem involving multi-scale natural fractures can be studied, considering the key physical mechanisms of fracture propagation, including stress shadowing, dynamic distribution of fluids among multiple fractures, and the interaction between hydraulic fractures and natural fractures. Natural fractures are incorporated through a Discrete Fracture Network (DFN) model to establish a field-scale numerical model of fracture propagation. That is, based on elasticity theory, fractures are considered as displacement discontinuities, thus allowing the calculation of the displacement discontinuities and induced stresses of a single three-dimensional fracture, and the accurate and efficient solution of the interactions between multiple three-dimensional fractures.

[0044] In one implementation, based on a pore-fracture dual-medium model that couples stress, fluid, and the physical mechanisms between fractures, the dual-medium model divides the reservoir into a matrix system and a fracture system. The matrix system primarily provides storage space, while the fracture system provides seepage channels. In implementation, the Picard iteration method can be used to iteratively couple rock deformation and fluid flow at each time step. Specifically, the iterative coupling solution can include: S1: Make a preliminary estimate of the fluid pressure inside the gap; S2: The displacement discontinuity surface is described using the theory of rock elastic deformation. Based on the simplified three-dimensional DDM method, the displacement discontinuity surface is calculated by estimating the pressure. S3: After obtaining the displacement discontinuity surface, the fluid pressure distribution is calculated using the Newton-Raphson iteration method. The calculated pressure is compared with the estimated pressure to verify its convergence. If the convergence criterion is not met, the pressure is corrected, and S2 and S3 are repeated until the model converges. S4: The mass balance equation is used to ensure that the total injected fluid volume is equal to the sum of the fluid volume in the crack and the fluid loss volume; S5: If the time step does not meet the requirements, continue to estimate the fluid pressure inside the crack, and then perform the calculation for the next time step. If the time step meets the calculation requirements, the simulation ends.

[0045] Natural fractures are incorporated into a discrete fracture network model, including their number, density, azimuth, length, width, and permeability. The presence of natural fractures significantly influences the propagation path of hydraulic fractures, which may extend along natural fractures to form complex fracture networks. Based on this, a Monte Carlo method can be used to generate a statistically consistent discrete fracture network and embed it into the fracture propagation model. In this model, natural fractures are treated as non-conductive frictional interfaces. The propagation mode of hydraulic fractures under the interference of natural fractures is determined based on the fracture intersection criterion, and the extension direction of the hydraulic fracture after intersecting with the natural fracture is determined by analyzing the stress state near the tip of the hydraulic fracture. If the maximum tensile stress on one side of the natural fracture (the side farther from the hydraulic fracture) is greater than or equal to the tensile strength of the rock, the hydraulic fracture will pass through the natural fracture and continue to propagate; if the shear stress and normal stress acting on the natural fracture satisfy the linear friction law, the hydraulic fracture will be captured by the natural fracture and deflected into it.

[0046] Step 103: Using the crack morphology obtained from the above simulation and the actual pumping pressure data, a crack propagation model is established based on a machine learning algorithm, wherein the crack propagation proxy model is used for automatic correction of crack geometry.

[0047] That is, the input to this step is the fracture morphology obtained from the above simulation and the actual pumping pressure data, and the output is the corrected fracture morphology. This output improves the accuracy of the fracture morphology through the first correction. Specifically, a crack propagation surrogate model is established using a preset model, and sampling is performed by combining Markov chains and Monte Carlo simulation. The optimal parameter combination is found in the high-dimensional uncertainty space to achieve automatic historical fitting of the pumping pressure curve and determine the parameters to be inverted (discharge rate, sand concentration, internal friction coefficient, etc.), thereby achieving automated calibration of the fracture morphology. The preset model consists of multiple decision trees, each of which focuses on the residual of the previous tree, branches outward from the root node, and uses a gradient algorithm to find a new method for building decision trees to reduce the residual of model training. Finally, the tree ensemble model is obtained by summing to predict the final result. In the sampling process, a Markov process is introduced into the Monte Carlo simulation to achieve dynamic simulation where the sampling distribution changes as the simulation progresses, thereby overcoming the limitation of existing Monte Carlo integrals which can only perform static simulations.

[0048] Building a crack propagation surrogate model based on machine learning algorithms can automatically fit the actual pumping pressure and correct the crack geometry. Specifically, automatically fitting the actual pumping pressure can be done as follows: Figure 2 As shown, after iteration, the simulated pumping pressure can match the actual pumping pressure well. The automatic pumping pressure fitting process mainly corrects four parameters to be inverted, including: effective perforation number, effective perforation diameter, drag reduction coefficient, and casing pressure gradient. After correction, the fracture network morphology, main fracture morphology, and fracture parameters after fracturing are obtained. The fracture network volume refers to the sum of the main fracture volume and the activated natural fracture volume.

[0049] In one implementation, forming an automatic pump pressure history fitting workflow based on a proxy model may include: S1: Collect on-site fracturing construction data, generate an input data table for all fracturing clusters / layers / segments, and prepare the simulation pumping program using an automated construction curve import tool; S2: Identify uncertain parameters and determine the sampling range; S3: Complete the first iteration of the workflow operation. This workflow is combined with the basic case and generates N initial samples using the Latin Hypercube Sampling (LHS) method. S4: The formed sample is substituted into the crack propagation model to simulate the pump pressure response; S5: The XGBoost model serves as a surrogate model for crack propagation simulation, used to train these N input / output combinations; S6: Using MCMC sampling, surrogate model and error benchmark objective function (weighted mean square error), extract the next M optimal samples and provide them to the crack propagation model to complete the second iteration of the workflow; S7: Repeat the above process until the surrogate model converges or the maximum number of iterations is reached.

[0050] In one implementation, establishing a reservoir simulation proxy model based on machine learning algorithms can include: automatically fitting actual production data and inverting reservoir and fracture characteristic parameters based on the reservoir simulation proxy model. Specifically, an initial reservoir model based on matrix, natural fractures, and hydraulic fracture coupling can be established, and production data can be fitted by correcting some parameters. The parameters to be corrected are fracture parameters and geological parameters. Among them, fracture parameters can include: effective hydraulic fracture length, effective hydraulic fracture height, hydraulic fracture conductivity, hydraulic fracture water saturation, hydraulic fracture width, and natural fracture conductivity. Geological parameters can include: rock compressibility coefficient, relative permeability curve matrix permeability, porosity, water saturation, etc. The production data to be fitted can include: gas production, water production, and bottom hole flowing pressure. Other parameters can include: number of iterations, number of initial iterations, number of subsequent iterations, maximum simulation duration, and number of single runs.

[0051] Based on the aforementioned surrogate model, an automatic historical data fitting workflow for production capacity can be formed. Specifically, an automatic historical data fitting workflow can be established using a neural network surrogate model and a Markov chain-Monte Carlo algorithm, and may include the following steps: S1: The conductivity between NNCs is calculated using the EDFM preprocessor, and the flow of gas and water phases in the matrix and fractures is calculated using the reservoir simulator. S2: Select a suitable range of values ​​for the uncertain parameters, use the LHS method to sample within the value space of the uncertain parameters, and obtain initial iterative samples. Then, use a reservoir numerical simulator to obtain the simulation results of the initial samples, and train the surrogate model to minimize the number of simulation runs; S3: The MCMC algorithm based on the surrogate model is used for sampling. The historical fitting candidate solution set is obtained with the global error as the target. The obtained candidate solution set is then brought into the reservoir simulator for verification, and the simulation results are used as the training dataset to further train the surrogate model. S4: Increase the number of computational cases and run iteratively until convergence or the maximum number of iterations is reached.

[0052] Step 104: Couple the fracture propagation model with the reservoir simulation model based on the embedded discrete fracture model (EDFM) to obtain a reservoir simulation model containing fractures.

[0053] That is, the input to this step is the fracture morphology corrected in step 103 and the three-dimensional geomechanical model established in step 101, and the output is a reservoir simulation model including fractures. Specifically, the matrix grid and fracture grid can be connected by non-adjacent connection pairs (NNC). The matrix uses a structured grid, and the fractures are simulated as rectangular plates embedded in the matrix grid. The matrix domain and fracture domain are meshed separately. This method not only has high operating efficiency and easy mesh generation, but also allows for flexible handling of non-planar complex fractures and fractures with varying conductivity (or varying fracture width). Furthermore, users can customize the permeability and aperture of hydraulic fractures, inactive natural fractures, and activated natural fractures.

[0054] Step 105: Predict the simulated production dynamics data and reservoir stimulation volume using the above-mentioned reservoir simulation model containing fractures.

[0055] In other words, the input to this step is a reservoir simulation model including fractures and the production regime (e.g., bottomhole flowing pressure, production rate, etc.), and the output is simulated production dynamics data (e.g., production rate, pressure, etc.) and reservoir stimulation volume. Specifically, production prediction models based on gas-water two-phase flow, adsorption-desorption (Langmuir isotherm adsorption model), molecular diffusion (Fick's law), and stress-sensitive effects can be considered to calculate the flow of gas-water two phases in the matrix and fractures. Input parameters include reservoir permeability, porosity, matrix / fracture compressibility, fluid PVT, relative permeability curve, fluid density, Langmuir volume / pressure, reservoir fracturing, well control data, etc. Furthermore, historical well data can be imported for historical fitting; the simulation time length, frequency, and step size (day, month, year) can also be customized.

[0056] Step 106: Establish a reservoir simulation proxy model based on machine learning algorithms to invert and obtain reservoir and fracture characteristic parameters.

[0057] That is, the input to this step is simulated production dynamic data and actual production history data, and the output is the reservoir and fracture characteristic parameters after a second correction. This output further improves the parameter accuracy through a second correction. Specifically, a neural network-Markov chain Monte Carlo (NN–MCMC) algorithm can be used. A suitable range of values ​​for the uncertain parameters is selected, and the LHS method is used to sample in the space of uncertain parameter values ​​to obtain initial iterative samples. The MCMC algorithm is then used to sample, and a candidate solution set for history fitting is obtained with the global error as the target. The cases are iteratively run until convergence, ultimately achieving automatic history fitting, inverting reservoir and fracture characteristic parameters, and predicting long-term production capacity and final recoverable reserves.

[0058] Specifically, such as Figure 3As shown, using daily gas production, daily water production, or bottom hole flowing pressure data as constraints, and minimizing the error of the production capacity prediction model as the objective function, the automatically retrieved reservoir parameters and fracture geometry parameters can include: S201: The initial sample set is generated using the Latin hypercube sampling method. Specifically, the initial sample set can be generated based on the distribution range of prior parameters, where prior parameters may include, but are not limited to, key parameters such as reservoir porosity, permeability, fracture conductivity, and fracture half-length. Latin hypercube sampling ensures a uniform distribution of sampling points in the multidimensional parameter space, providing sufficient sample data for subsequent surrogate model training.

[0059] S202: Generating simulation training data through an embedded discrete fracture model. Specifically, the generated initial sample set is input into the embedded discrete fracture model and the reservoir simulation model to calculate the corresponding production dynamic response, which includes bottom hole pressure, gas production, and water production. By running numerical simulations in batches, a simulation training dataset of parameter combinations and production responses is constructed, providing training samples for the establishment of a neural network proxy model.

[0060] S203: Constructing a multi-response parameter surrogate model based on a neural network algorithm. Specifically, using the simulated training data generated above, a multi-response parameter surrogate model is constructed using a neural network algorithm. This surrogate model can quickly predict the dynamic production response under a given combination of parameters, replacing the computationally intensive numerical model. The neural network structure includes an input layer, a hidden layer, and an output layer. The input layer corresponds to reservoir and fracture parameters, and the output layer corresponds to dynamic production indicators.

[0061] S204: An automatic history fitting surrogate model is established using the NN-MCMC algorithm. Specifically, the neural network surrogate model is combined with the Markov chain Monte Carlo algorithm to establish the NN-MCMC automatic history fitting model. This model uses actual observed production history data as constraints and searches the parameter space through MCMC sampling to find the parameter combination with the highest matching degree to the measured data.

[0062] S205: Obtain the next generation of sampling examples based on convergence criteria. Specifically, based on the convergence status of the NN-MCMC algorithm, determine whether sampling needs to continue based on preset convergence criteria (such as acceptance rate, change in likelihood function value, etc.). If the convergence criteria are not met, obtain the next generation of sampling examples and return to S202 for a new round of simulation calculation and surrogate model update.

[0063] S206: Update the training dataset for the neural network surrogate model. Specifically, add the results of the latest iteration of the simulation examples to the training dataset of the neural network surrogate model, retrain and optimize the surrogate model, and improve its prediction accuracy. Through iterative updates, the surrogate model gradually approximates the true parameter-response relationship.

[0064] S207: Determine if the surrogate model has converged. That is, evaluate the convergence status of the surrogate model and determine whether the prediction error of the surrogate model is less than the preset threshold. If the convergence condition is met, proceed to S208; otherwise, return to S205 to continue iterative optimization.

[0065] S208: Output the optimal solution set for historical fitting. That is, when the surrogate model reaches the convergence criterion, the optimal solution set for historical fitting is output, including: posterior parameter distribution characteristics, optimal parameter combination and its uncertainty range, and historical fitting quality evaluation index. This optimal solution set provides reliable input parameters for subsequent long-term capacity prediction and effective modification volume assessment.

[0066] Step 107: Predict long-term productivity and effective stimulation volume based on the second revised fracture and reservoir characteristic parameters.

[0067] That is, the input to this step is the reservoir and fracture characteristic parameters obtained from the inversion in step 106, and the output is the long-term production capacity prediction result and the effective stimulation volume quantification result, where the effective stimulation rate is quantified by the ratio of fluid discharge volume to reservoir stimulation volume. Specifically, the fracture and reservoir characteristic parameters after the second correction are input into the reservoir simulation model to predict long-term production capacity (e.g., cumulative gas production, cumulative water production, bottom hole flowing pressure, etc.) and fluid discharge volume. In implementation, a solution set can also be optimized based on minimizing the global error as the objective function, and the optimized solution set can be input into the reservoir simulation model to predict long-term production capacity and stimulation volume. For example, P10, P50, P90, and the most representative daily production, cumulative production, and bottom hole flowing pressure data can be obtained.

[0068] The above scheme achieves dual automatic historical correction for fracturing and production capacity, significantly improving the accuracy and reliability of post-fracturing effect evaluation. The core of the dual intelligent fitting correction of fracturing and production capacity is to obtain the most representative effective hydraulic fracture geometry and parameters, and the optimal combination scheme, by achieving the minimum global error. Specifically, steps 103 and 106 achieve dual correction through a fracture propagation proxy model and a reservoir simulation proxy model, respectively, effectively reducing model uncertainty. Finally, step 107 quantifies the effective stimulation rate, providing a scientific basis for fracturing effect evaluation; furthermore, it can provide predictive effects for different drainage schemes and provide fracturing design optimization guidance for other fracturing wells on the same platform.

[0069] This invention also provides a post-pressure evaluation system based on the integration of geology, engineering, and intelligence, such as... Figure 4 As shown, the system may include: Data collection module 301 is used to collect data from the target block and establish an integrated geological engineering shared database; Geological model building module 302 is used to build the fine three-dimensional geomechanical model; Crack propagation module 303 is used to establish the mine-scale crack propagation model; The first correction module 304 is used to establish a crack propagation proxy model. Fracture embedding module 305 is used to couple the fracture propagation model with the reservoir simulation model; Reservoir simulation module 306 is used to calculate the flow of gas-water two-phase flow in the matrix and the first-corrected fractures; The second correction module 307 is used to establish a reservoir simulation proxy model; The results visualization module 308 is used to display long-term capacity forecasts and effective modification volume results.

[0070] In the above example, a rapid post-fracturing effect evaluation method and system based on geology-engineering-intelligence integration is provided. This method fully utilizes a large amount of existing geological and engineering data to achieve rapid and accurate evaluation of post-fracturing effects. Based on dual automatic historical correction of fracturing and production capacity, this method accurately quantifies the effective stimulation volume, making the post-fracturing effect evaluation more realistic and accurate. Compared with existing prediction and evaluation methods, this example leverages the data mining and learning capabilities of artificial intelligence to quickly establish fracture propagation and reservoir models based on geological, engineering, and drainage data of the production fracturing wells. Through dual automatic correction of the models using construction pump pressure fitting and production data, the computationally intensive simulation runs are greatly reduced, as are the complex procedures of manual parameter adjustment, providing a new method for rapid and accurate prediction. Furthermore, the evaluation method provided in this embodiment does not require a large number of parameters to complete the modeling, reducing dependence on parameters and avoiding the inability to form a model due to insufficient parameters. Once the model is established, it can be quickly applied to other wells in the same block, requiring only updates to the input parameters, providing a reliable basis for fracturing scheme optimization, EUR prediction, and stimulation evaluation. Therefore, the evaluation method described in this embodiment has strong generalization ability and cross-block applicability.

[0071] Figure 5This is a flowchart of one embodiment of the method for determining the fracturing effect of dual intelligent correction provided in this application. Although this application provides method operation steps or device structures as shown in the following embodiments or figures, more or fewer operation steps or module units may be included in the method or device based on conventional or non-inventive effort. In steps or structures where there is no logically necessary causal relationship, the execution order of these steps or the module structure of the device is not limited to the execution order or module structure described in the embodiments and figures of this application. When the method or module structure is applied in actual devices or end products, it can be executed sequentially or in parallel according to the method or module structure shown in the embodiments or figures (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed processing environment).

[0072] Specifically, such as Figure 5 As shown, the method for determining the fracturing effect of the above-mentioned dual intelligent correction may include the following steps: Step 501: Collect multi-source data of the target block, calculate mechanical parameters through the multi-source data, and obtain discrete well point data of the target block; Step 502: Interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; Step 503: Based on the dual medium model of pores and fractures, using the fracturing construction parameters of the target block, natural fractures are implanted into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing construction parameters, a numerical model of fracture propagation at the mine scale is established to output the simulated fracture geometry. Step 504: Based on the simulated crack geometry and the actual pumping pressure data, find the optimal parameter combination in the high-dimensional uncertainty space, and automatically fit the pumping pressure curve to obtain the crack geometry after one correction. Step 505: Couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. Step 506: Using the target block reservoir numerical simulation model and production regime, simulate the production dynamics data and reservoir stimulation volume; Step 507: Based on the simulated production dynamic data and the actual production history data, establish a reservoir simulation proxy model, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. Step 508: Based on reservoir characteristic parameters and fracture geometry after secondary correction, predict the long-term productivity and effective stimulation volume of the target block.

[0073] In actual implementation, step 507 above establishes a reservoir simulation proxy model based on the simulated production dynamic data and actual production historical data, inverts reservoir characteristic parameters and fracture geometry, and uses the inverted fracture geometry as the fracture geometry after secondary correction. This may include: S1: Based on the distribution range of prior parameters, an initial sample set is generated using the Latin hypercube sampling method, wherein the prior parameters include at least one of the following: reservoir porosity, permeability, fracture conductivity, and fracture half-length; S2: Input the initial sample set into a mesh model containing fractures, calculate the corresponding production dynamic data, and perform batch numerical simulations on the production dynamic data to construct simulation training data of parameter combinations and production response. The production dynamic data includes at least one of the following: bottom hole flowing pressure, daily gas production, daily water production, etc. S3: Using the simulated training data, a multi-response parameter surrogate model is constructed using a neural network algorithm, wherein the multi-response parameter surrogate model is used to predict production dynamic data under a given combination of parameters; S4: Combine the multi-response parameter surrogate model with Markov chain Monte Carlo to establish an automatic history fitting model. In the automatic history fitting model, the actual production history data is used as a constraint. The parameter space is searched through a preset sampling method to find the parameter combination with the highest matching degree with the actual production history data. S4: Based on the preset convergence criteria, determine whether sampling needs to continue. If the convergence criteria are not met, re-sample to perform the next round of simulation calculation and proxy model update. If the convergence criteria are met, the parameter combination that best matches the actual production historical data is taken as the optimal historical fitting solution set. The optimal historical fitting solution set includes at least one of the following: posterior parameter distribution characteristics, optimal parameter combination and its uncertainty range, and historical fitting quality evaluation index. S5: By fitting the optimal solution set in history, establish a reservoir simulation proxy model to invert reservoir characteristic parameters and fracture geometry.

[0074] In one embodiment, multi-source data of the target block is collected, and mechanical parameters are calculated using the multi-source data to obtain discrete well point data of the target block. This may include: collecting geological data, logging data, drilling data, and completion data of the target block, wherein the geological data includes at least one of the following: regional structure, reservoir geological description, stratigraphic information, and fault / natural fracture distribution; the logging data includes at least one of the following: in-situ stress and geological parameters; the drilling data includes at least one of the following: wellhead coordinates, wellbore trajectory, and casing program; the completion data includes at least one of the following: perforation information, fracturing design, construction curve, flowback data, and gas testing data; and the geological data, logging data, drilling data, and completion data are used as multi-source data to calculate mechanical parameters, wherein the mechanical parameters include at least one of the following: vertical depth, depth sounding, Young's modulus, Poisson's ratio, and in-situ stress.

[0075] In practical implementation, based on the fracturing construction parameters of the target block and the dual-medium model of pores and fractures, natural fractures are embedded into the three-dimensional geomechanical model through a discrete fracture network model to obtain a field-scale fracture propagation numerical model. This can include: generating a discrete fracture network model that conforms to statistical laws using the Monte Carlo method; embedding the generated discrete fracture network model into the three-dimensional geomechanical model to obtain a field-scale fracture propagation numerical model; wherein, in the field-scale fracture propagation numerical model, natural fractures are regarded as friction interfaces without flow conductivity, the hydraulic fracture propagation mode under the interference of natural fractures is determined based on the fracture intersection criterion, and the extension direction of the hydraulic fracture after intersecting with the natural fracture is determined by analyzing the stress state near the tip of the hydraulic fracture. If the maximum tensile stress on one side of the natural fracture is greater than or equal to the tensile strength of the rock, the hydraulic fracture will pass through the natural fracture and continue to propagate. If the shear stress and normal stress acting on the natural fracture satisfy the linear friction law, the hydraulic fracture is captured by the natural fracture and deflected into the natural fracture.

[0076] Furthermore, based on the fracturing construction parameters of the target block and the dual-medium model of pores and fractures, natural fractures are embedded into the three-dimensional geomechanical model through a discrete fracture network model to obtain a field-scale fracture propagation numerical model. This model can include: iteratively coupling the rock deformation and fluid flow at each time step to obtain the fluid pressure distribution; repeating the following steps until the model converges to make a preliminary estimate of the fluid pressure within the fracture; using the rock elastic deformation theory to describe the displacement discontinuity, and using the estimated pressure to calculate the displacement discontinuity; after obtaining the displacement discontinuity, using the Newton-Raphson iteration method to calculate the fluid pressure distribution, comparing the calculated pressure with the preliminary estimated pressure to verify convergence; using the mass balance equation to ensure that the total injected fluid volume is equal to the sum of the fluid volume in the fracture and the filtration volume; if the time step does not meet the calculation requirements, continue to estimate the fluid pressure within the fracture and perform the calculation for the next time step; if the time step meets the calculation requirements, the iterative coupling solution ends.

[0077] Specifically, a preliminary estimate of the fluid pressure within the crevice can include: S1: Calculate the pressure drop within a single crack using the following formula:

[0078] in, This represents the pressure drop within a single crack. Indicates internal pressure. Indicates the distance along the crack. Indicates the consistency index. Represents the power-law exponent of fluid. Indicates the crack height. Indicates the crack width. Indicates fluid flow rate; S2: Calculate the total injection displacement based on the injection displacement of each fracture using the following formula:

[0079] in, Indicates the total injection displacement. Let N represent the injection displacement of the i-th fracture, and N represent the number of fractures. S3: Calculate the crack inlet pressure using the following formula:

[0080] in, The inlet pressure of the crack. Indicates wellhead pressure, For the perforation friction voltage drop, For wellbore friction pressure drop, Let be the perforation friction pressure drop at the i-th crack. For the density of the proppant slurry, The density of water, Let be the number of perforations at the i-th crack. Let be the diameter of the perforation at the i-th crack. Let be the wellbore friction pressure drop at the i-th fracture. Indicates the wellbore friction coefficient. Indicates the location of the crack. Represents the power-law exponent of fluid. This indicates the flow rate in the current section of the wellbore. S4: Based on the pressure drop within the single fracture, the total injection volume, and the fracture inlet pressure, the pressure drop along the fracture is calculated using lubrication theory, and then fully coupled with rock deformation for iterative solution to obtain the fluid pressure within the fracture.

[0081] The above calculation of the displacement discontinuity surface using the preliminary estimated fluid pressure within the fracture can include: dividing the boundary into multiple planar elements; calculating the stress generated by the displacement discontinuity in each planar element; and linearly combining the stresses generated by the displacement discontinuities in each planar element to obtain the stress at any point. Specifically, the displacement discontinuity obtained using the displacement discontinuity method determines the fracture opening width, providing a geometric channel for the fluid. Simultaneously, the pressure within the fracture obtained by solving the nonlinear fluid equations using the Newton numerical method acts as an internal boundary load on the fracture surface. Since these two are mutually causal, the model uses Picard iteration for iterative calculations at each time step. That is, the updated pressure is used to correct the current displacement, and then the internal pressure is recalculated based on the new displacement until the entire rock mechanics and fluid system reaches convergence equilibrium.

[0082] Specifically, the stress generated by the displacement discontinuities of each planar element in multiple planar elements is linearly combined to obtain the stress at any point. This can be achieved by calculating the stress generated by the displacement discontinuity at point η on the boundary at any point within the region using the following formula:

[0083] in, Let η be the displacement discontinuity at point η on the boundary, which generates stress at point ξ within the region. For displacement, This is the tensor field representing the effect of concentrated force on displacement. Here, j and k represent the component coordinate indices of the stress tensor, which are the integration boundaries.

[0084] The above calculation of the stress generated by the displacement discontinuity of each planar element can include: Under given boundary conditions, the stresses generated by the normal displacement discontinuity and the shear displacement discontinuity of each planar element are obtained by solving the following system of equations:

[0085] in,

[0086] in, The total normal stress is generated by the surface displacement discontinuity of planar element j. The total shear stress generated by the surface displacement discontinuity of planar element j. This represents the discontinuity in the normal displacement of planar element j, corresponding to the crack opening. The shear displacement discontinuity of planar element j represents the relative slippage between crack surfaces, and N is the number of planar elements. Let be the elastic influence coefficient matrix under plane strain. Let be the normal stress response induced at plane element i by the discontinuity of shear displacement in plane element j. Let be the distance between the center of crack element i and the center of crack element j. The height of the crack. and These are empirical constants determined to match the actual 3D crack solution.

[0087] The crack element is a discretized planar element. The displacement discontinuity of each planar element is transformed into a stress contribution in space through the Green's function (analytical solution). Finally, the contributions of all elements are added together to form a complex crack mesh stress field.

[0088] After calculating the stress generated by the displacement discontinuity of each planar element, the relative net pressure can also be calculated as the stress shadowing effect according to the following formula, wherein the stress shadowing effect is used to characterize the induced stress compression of the internal crack by the external cracks on both sides:

[0089] in, Indicates relative net pressure. This indicates the net pressure inside the crack. This represents the minimum horizontal principal stress. This represents the maximum horizontal principal stress. Then, the stress generated by the displacement discontinuity of each planar element is optimized through the stress shadowing effect.

[0090] This example also provides specific calculation and improvement formulas for long-term production capacity and effective modification volume. Specifically, it can be...

[0091] Calculate the effective modification volume using the following formula:

[0092] in, In order to effectively modify the volume, Let be the effective permeability at (x, y, z) in three-dimensional space. For the effective permeability of the crack, For the effective permeability of the matrix, The effective porosity at (x, y, z) in three-dimensional space. Initial porosity, The stress sensitivity coefficient, This represents the effective stress variation.

[0093] After predicting the long-term productivity and effective stimulation volume of the target block based on the reservoir fracture morphology after secondary correction, the corrected long-term productivity can be calculated using the following formula based on the predicted long-term productivity:

[0094] in, This represents the long-term production capacity after adjustment at time t. Forecast long-term capacity, This is the crack attenuation correction factor. Let be the crack conductivity at time t. Let be the matrix permeability at time t. Let be the average water saturation at time t.

[0095] The methods and embodiments provided in the above-described embodiments of this application can be executed in a mobile terminal, computer terminal, or similar computing device. Taking operation on an electronic device as an example... Figure 6 This is a hardware structure block diagram of an electronic device for a dual intelligent correction method for determining the fracturing effect provided in this application. (See diagram for example.) Figure 6As shown, the electronic device 10 may include one or more (only one is shown in the figure) processors 02 (processors 02 may include, but are not limited to, processing devices such as microprocessors (MCUs) or programmable logic devices (FPGAs), a memory 04 for storing data, and a transmission module 06 for communication functions. Those skilled in the art will understand that... Figure 6 The structure shown is for illustrative purposes only and does not limit the structure of the electronic device described above. For example, electronic device 10 may also include... Figure 6 The more or fewer components shown, or having the same Figure 6 The different configurations shown.

[0096] The memory 04 can be used to store software programs and modules of application software, such as the program instructions / modules corresponding to the dual intelligent correction fracturing modification effect determination method in this embodiment. The processor 02 executes various functional applications and data processing by running the software programs and modules stored in the memory 04, thereby realizing the dual intelligent correction fracturing modification effect determination method of the above-mentioned application. The memory 04 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 04 may further include memory remotely located relative to the processor 02, and these remote memories can be connected to the electronic device 10 via a network. Examples of the above-mentioned networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0097] The transmission module 06 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the electronic device 10. In one example, the transmission module 06 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission module 06 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0098] At the software level, the aforementioned dual-intelligent correction fracturing effect determination device can, as... Figure 7 As shown, it includes: The acquisition module 701 is used to acquire multi-source data of the target block and calculate mechanical parameters through the multi-source data to obtain discrete well point data of the target block; Interpolation module 702 is used to interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; The output module 703 is used to, based on the dual medium model of pores and fractures, embed natural fractures into the three-dimensional geomechanical model through a discrete fracture network model using the fracturing construction parameters of the target block, and establish a field-scale fracture propagation numerical model in combination with the fracturing construction parameters, so as to output the simulated fracture geometry. The primary correction module 704 is used to find the optimal parameter combination in a high-dimensional uncertainty space based on the simulated crack geometry and the actual pumping pressure data, and to automatically fit the pumping pressure curve to obtain the crack geometry after primary correction. Module 705 is established to couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model, thereby establishing a numerical simulation model of the target block reservoir. The simulation module 706 is used to simulate production dynamics data and reservoir stimulation volume through the target block reservoir numerical simulation model and production regime. The secondary correction module 707 is used to establish a reservoir simulation proxy model based on the simulated production dynamic data and the actual production history data, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. The prediction module 708 is used to predict the long-term productivity and effective stimulation volume of the target block based on reservoir characteristic parameters and fracture geometry after secondary correction.

[0099] In one implementation, the secondary correction module 707 can specifically generate an initial sample set based on the distribution range of prior parameters using the Latin hypercube sampling method. The prior parameters include at least one of the following: reservoir porosity, permeability, fracture conductivity, and fracture half-length. The initial sample set is input into a grid model containing fractures to calculate the corresponding production dynamic data. Batch numerical simulations are then performed on the production dynamic data to construct simulation training data for parameter combinations and production responses. The production dynamic data includes at least one of the following: bottomhole flowing pressure, daily gas production, and daily water production. Using the simulation training data, a multi-response parameter surrogate model is constructed using a neural network algorithm. This multi-response parameter surrogate model is used to predict production dynamic data under a given parameter combination. The multi-response parameter surrogate model is then compared with Markov chain Monte Carlo simulation. An automatic history fitting model is established by combining these methods. This model uses actual production history data as constraints and searches the parameter space through a preset sampling method to find the parameter combination that best matches the actual production history data. Based on a preset convergence criterion, it determines whether further sampling is needed. If the convergence criterion is not met, sampling is repeated for the next round of simulation calculations and surrogate model updates. If the convergence criterion is met, the parameter combination that best matches the actual production history data is taken as the optimal historical fitting solution set. The optimal historical fitting solution set includes at least one of the following: posterior parameter distribution characteristics, optimal parameter combination and its uncertainty range, and historical fitting quality evaluation index. Using the optimal historical fitting solution set, a reservoir simulation surrogate model is established to invert reservoir characteristic parameters and fracture geometry.

[0100] In one embodiment, the acquisition module 701 can specifically acquire geological data, logging data, drilling data, and completion data of the target block. The geological data includes at least one of the following: regional structure, reservoir geological description, stratigraphic information, and fault / natural fracture distribution. The logging data includes at least one of the following: in-situ stress and geological parameters. The drilling data includes at least one of the following: wellhead coordinates, wellbore trajectory, and casing program. The completion data includes at least one of the following: perforation information, fracturing design, construction curves, flowback data, and gas testing data. The geological data, logging data, drilling data, and completion data are used as multi-source data to calculate mechanical parameters, which include at least one of the following: vertical depth, depth sounding, Young's modulus, Poisson's ratio, and in-situ stress.

[0101] In one implementation, the output module 703 can specifically use the Monte Carlo method to generate a discrete fracture network model that conforms to statistical laws; embed the generated discrete fracture network model into a three-dimensional geomechanical model to obtain a numerical model of fracture propagation at the mine scale; wherein, in the numerical model of fracture propagation at the mine scale, natural fractures are regarded as frictional interfaces without flow conductivity, the hydraulic fracture propagation mode under the interference of natural fractures is determined based on the fracture intersection criterion, and the extension direction of the hydraulic fracture after intersecting with the natural fracture is determined by analyzing the stress state near the tip of the hydraulic fracture. If the maximum tensile stress on one side of the natural fracture is greater than or equal to the tensile strength of the rock, the hydraulic fracture will pass through the natural fracture and continue to propagate. If the shear stress and normal stress acting on the natural fracture satisfy the linear friction law, the hydraulic fracture is captured by the natural fracture and deflected into the natural fracture.

[0102] In one implementation, the output module 703 can specifically perform iterative coupling solution of rock deformation and fluid flow at each time step in the following manner to obtain the fluid pressure distribution: repeat the following steps until the model converges: make a preliminary estimate of the fluid pressure in the fracture; use the rock elastic deformation theory to describe the displacement discontinuity, and use the preliminary estimated fluid pressure in the fracture to calculate the displacement discontinuity; after obtaining the displacement discontinuity, use the Newton iteration method to calculate the fluid pressure distribution, compare the calculated pressure with the preliminary estimated pressure, and verify the convergence; use the mass balance equation to ensure that the total injected fluid volume is equal to the sum of the fluid volume in the fracture and the filtration volume; if the time step does not meet the calculation requirements, continue to estimate the fluid pressure in the fracture and perform the calculation for the next time step; if the time step meets the calculation requirements, the iterative coupling solution ends.

[0103] In one implementation, a preliminary estimate of the fluid pressure within the crevice may include: Calculate the pressure drop within a single crack using the following formula:

[0104] in, This represents the pressure drop within a single crack. Indicates internal pressure. Indicates the distance along the crack. Indicates the consistency index. Represents the power-law exponent of fluid. Indicates the crack height. Indicates the crack width. Indicates fluid flow rate; Based on the injection displacement of each fracture, the total injection displacement is calculated using the following formula:

[0105] in, Indicates the total injection displacement. Let N represent the injection displacement of the i-th fracture, and N represent the number of fractures. Calculate the fracture inlet pressure using the following formula:

[0106] in, The inlet pressure of the crack. Indicates wellhead pressure, For the perforation friction voltage drop, For wellbore friction pressure drop, Let be the perforation friction pressure drop at the i-th crack. For the density of the proppant slurry, The density of water, Let be the number of perforations at the i-th crack. Let be the diameter of the perforation at the i-th crack. Let be the wellbore friction pressure drop at the i-th fracture. Indicates the wellbore friction coefficient. Indicates the location of the crack. Represents the power-law exponent of fluid. This indicates the flow rate in the current section of the wellbore. Based on the pressure drop within the single fracture, the total injection volume, and the fracture inlet pressure, the pressure drop along the fracture is calculated using lubrication theory, and then fully coupled with rock deformation for iterative solution to obtain the fluid pressure within the fracture.

[0107] In one embodiment, calculating the displacement discontinuity surface using a preliminary estimate of the fluid pressure within the gap may include: dividing the boundary into multiple planar elements; calculating the stress generated by the displacement discontinuity in each planar element; and linearly combining the stress generated by the displacement discontinuity in each of the multiple planar elements to obtain the stress at any point.

[0108] In one implementation, linearly combining the stresses generated by the displacement discontinuities of various planar elements in a plurality of planar elements to obtain the stress at any point may include: Calculate the stress generated at any point within the region due to the displacement discontinuity at point η on the boundary using the following formula:

[0109] in, Let η be the displacement discontinuity at point η on the boundary, which generates stress at point ξ within the region. For displacement, This is the tensor field representing the effect of concentrated force on displacement. Here, j and k represent the component coordinate indices of the stress tensor, which are the integration boundaries.

[0110] In one implementation, calculating the stress generated by the displacement discontinuity of each planar element may include: Under given boundary conditions, the stresses generated by the normal displacement discontinuity and the shear displacement discontinuity of each planar element are obtained by solving the following system of equations:

[0111] in,

[0112] in, Let be the total normal stress generated by the surface displacement discontinuity of planar element j. Let J be the total shear stress generated by the surface displacement discontinuity of planar element j. This represents the discontinuity in the normal displacement of planar element j, corresponding to the crack opening. The shear displacement discontinuity of planar element j represents the relative slippage between crack surfaces, and N is the number of planar elements. Let be the elastic influence coefficient matrix under plane strain. Let be the normal stress response induced at plane element i by the discontinuity of shear displacement in plane element j. Let be the distance between the center of crack element i and the center of crack element j. The height of the crack. and These are empirical constants determined to match the actual 3D crack solution.

[0113] In one implementation, after calculating the stress generated by the displacement discontinuity of each planar element, the relative net pressure can also be calculated as a stress shadowing effect according to the following formula, wherein the stress shadowing effect is used to characterize the induced stress compression of the internal crack by the external cracks on both sides:

[0114] in, Indicates relative net pressure. This indicates the net pressure inside the crack. This represents the minimum horizontal principal stress. This represents the maximum horizontal principal stress. The stress shadowing effect optimizes the stress generated by the displacement discontinuity of each planar element.

[0115] This application also provides a specific implementation of an electronic device capable of implementing all steps in the dual intelligent correction fracturing effect determination method described in the above embodiments. The electronic device specifically includes: a processor, a memory, a communication interface, and a bus; wherein the processor, memory, and communication interface communicate with each other via the bus; the processor is used to call a computer program in the memory, and when the processor executes the computer program, it implements all steps in the dual intelligent correction fracturing effect determination method described in the above embodiments. For example, when the processor executes the computer program, it implements the following steps: Step 1: Collect multi-source data of the target block, calculate mechanical parameters using the multi-source data, and obtain discrete well point data of the target block; Step 2: Interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; Step 3: Based on the dual-medium model of pores and fractures, using the fracturing construction parameters of the target block, natural fractures are implanted into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing construction parameters, a numerical model of fracture propagation at the mine scale is established to output the simulated fracture geometry. Step 4: Based on the simulated crack geometry and the actual pumping pressure data, find the optimal parameter combination in the high-dimensional uncertainty space, and automatically fit the pumping pressure curve to obtain the crack geometry after one correction. Step 5: Couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. Step 6: Using the target block reservoir numerical simulation model and production regime, simulate the production dynamics data and reservoir stimulation volume; Step 7: Based on the simulated production dynamic data and the actual production history data, establish a reservoir simulation proxy model, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. Step 8: Based on reservoir characteristic parameters and fracture geometry after secondary correction, predict the long-term productivity and effective stimulation volume of the target block.

[0116] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the fracturing effect determination method with dual intelligent correction in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the fracturing effect determination method with dual intelligent correction in the above embodiments. For example, when the processor executes the computer program, it implements the following steps: Step 1: Collect multi-source data of the target block, calculate mechanical parameters using the multi-source data, and obtain discrete well point data of the target block; Step 2: Interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; Step 3: Based on the dual-medium model of pores and fractures, using the fracturing construction parameters of the target block, natural fractures are implanted into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing construction parameters, a numerical model of fracture propagation at the mine scale is established to output the simulated fracture geometry. Step 4: Based on the simulated crack geometry and the actual pumping pressure data, find the optimal parameter combination in the high-dimensional uncertainty space, and automatically fit the pumping pressure curve to obtain the crack geometry after one correction. Step 5: Couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. Step 6: Using the target block reservoir numerical simulation model and production regime, simulate the production dynamics data and reservoir stimulation volume; Step 7: Based on the simulated production dynamic data and the actual production history data, establish a reservoir simulation proxy model, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. Step 8: Based on reservoir characteristic parameters and fracture geometry after secondary correction, predict the long-term productivity and effective stimulation volume of the target block.

[0117] As can be seen from the above description, the embodiments of this application form a complete closed loop of geological modeling, fracture propagation, fracture correction, reservoir coupling, parameter inversion and production capacity assessment. The first correction of fracture morphology is performed by fitting pump pressure history, and the second correction of fracture parameters is performed by fitting production history. Through dual intelligent correction and full-process coupling, the accuracy of fracture morphology and production capacity prediction is significantly improved, the cost of manual intervention and calculation is reduced, and the fracturing effect can be evaluated quickly, accurately and quantitatively and the field fracturing optimization can be guided.

[0118] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. In particular, hardware + program embodiments are relatively simple in description because they are fundamentally similar to method embodiments; relevant parts can be referred to the descriptions in the method embodiments.

[0119] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0120] While this application provides the method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the methods shown in the embodiments or drawings can be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment).

[0121] While this specification provides method operation steps as described in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive means. The order of steps listed in the embodiments is merely one possible execution order among many and does not represent the only execution order. In actual device or end product execution, the methods shown in the embodiments or drawings may be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment, or even a distributed data processing environment). The terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, product, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, product, or apparatus. Without further limitations, the presence of other identical or equivalent elements in the process, method, product, or apparatus that includes said elements is not excluded.

[0122] For ease of description, the above devices are described in terms of function, divided into various modules. Of course, in implementing the embodiments of this specification, the functions of each module can be implemented in one or more software and / or hardware components, or a module that performs the same function can be implemented by a combination of multiple sub-modules or sub-units. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between devices or units, and may be electrical, mechanical, or other forms.

[0123] Those skilled in the art will also know that, besides implementing the controller using purely computer-readable program code, the same functions can be achieved by logically programming the method steps, making the controller function as logic gates, switches, application-specific integrated circuits (ASICs), programmable logic controllers (PLCs), and embedded microcontrollers. Therefore, such a controller can be considered a hardware component, and the devices within it used to implement various functions can also be considered structures within that hardware component. Alternatively, the devices used to implement various functions can be considered as both software modules implementing the method and structures within a hardware component.

[0124] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0125] Those skilled in the art will understand that the embodiments of this specification can be provided as methods, systems, or computer program products. Therefore, the embodiments of this specification can take the form of entirely hardware embodiments, entirely software embodiments, or embodiments combining software and hardware aspects. Furthermore, the embodiments of this specification can take the form of computer program products implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] The embodiments described in this specification can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform a specific task or implement a specific abstract data type. The embodiments of this specification can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0127] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, system embodiments are basically similar to method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments. In the description of this specification, the terms "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the embodiments in this specification. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0128] The above description is merely an embodiment of the present specification and is not intended to limit the embodiments of the present specification. For those skilled in the art, various modifications and variations can be made to the embodiments of the present specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the embodiments of the present specification should be included within the scope of the claims of the embodiments of the present specification.

Claims

1. A method for determining the fracturing effect with dual intelligent correction, characterized in that, The method includes: Collect multi-source data of the target block, calculate mechanical parameters using the multi-source data, and obtain discrete well point data of the target block; The discrete well point data is interpolated into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block; Based on the fracturing construction parameters of the target block and the dual medium model of pores and fractures, natural fractures are implanted into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing construction parameters, a numerical model of fracture propagation at the mine scale is established to output the simulated fracture geometry. Based on the simulated crack geometry and the actual pumping pressure data, the optimal parameter combination is searched in the high-dimensional uncertainty space, and the pumping pressure curve is automatically historically fitted to obtain the crack geometry after one correction. The corrected fracture geometry is coupled to the reservoir numerical model through an embedded discrete fracture model to establish a numerical simulation model of the target block reservoir. By using a numerical simulation model of the target block reservoir and the production regime, dynamic production data and reservoir stimulation volume were obtained through simulation. Based on the simulated production dynamics data and actual production history data, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry. The inverted fracture geometry is used as the fracture geometry after secondary correction. Based on reservoir characteristic parameters and fracture geometry after secondary correction, the long-term productivity and effective stimulation volume of the target block are predicted.

2. The method according to claim 1, characterized in that, Based on the simulated production dynamics data and actual production history data, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry, including: Based on the distribution range of prior parameters, an initial sample set is generated using the Latin hypercube sampling method. The prior parameters include at least one of the following: reservoir porosity, permeability, fracture conductivity, and fracture half-length. The initial sample set is input into a mesh model containing fractures, the corresponding production dynamic data is calculated, and batch numerical simulations are performed on the production dynamic data to construct simulation training data of parameter combinations and production response. The production dynamic data includes at least one of the following: bottom hole flowing pressure, daily gas production, and daily water production. Using the simulated training data, a multi-response parameter surrogate model is constructed using a neural network algorithm, wherein the multi-response parameter surrogate model is used to predict production dynamic data under a given combination of parameters; The multi-response parameter surrogate model is combined with Markov chain Monte Carlo to establish an automatic history fitting model. In the automatic history fitting model, the actual production history data is used as a constraint, and the parameter combination with the highest matching degree with the actual production history data is searched in the parameter space through a preset sampling method. Based on the preset convergence criteria, it is determined whether sampling needs to continue. If the convergence criteria are not met, sampling is performed again for the next round of simulation calculation and proxy model update. If the convergence criteria are met, the parameter combination that best matches the actual production historical data is taken as the optimal historical fitting solution set. The optimal historical fitting solution set includes at least one of the following: posterior parameter distribution characteristics, optimal parameter combination and its uncertainty range, and historical fitting quality evaluation index. By fitting the optimal solution set in history, a reservoir simulation proxy model is established to invert reservoir characteristic parameters and fracture geometry.

3. The method according to claim 1, characterized in that, Collect multi-source data of the target block, calculate mechanical parameters using the multi-source data, and obtain discrete well point data of the target block, including: Collect geological data, logging data, drilling data, and completion data for the target block. The geological data includes at least one of the following: regional structure, reservoir geological description, stratigraphic information, and fault / natural fracture distribution. The logging data includes at least one of the following: geostress and geological parameters. The drilling data includes at least one of the following: wellhead coordinates, wellbore trajectory, and casing program. The completion data includes at least one of the following: perforation information, fracturing design, construction curves, flowback data, and gas testing data. Geological data, well logging data, drilling data, and well completion data are used as multi-source data to calculate mechanical parameters, wherein the mechanical parameters include at least one of the following: vertical depth, sounding depth, Young's modulus, Poisson's ratio, and in-situ stress.

4. The method according to claim 1, characterized in that, Based on the fracturing operation parameters of the target block and the dual-medium model of pores and fractures, natural fractures are incorporated into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing operation parameters, a field-scale fracture propagation numerical model is established to output the simulated fracture geometry, including: A discrete crack network model conforming to statistical laws was generated using the Monte Carlo method. The generated discrete fracture network model is embedded into a three-dimensional geomechanical model to obtain a numerical model of fracture propagation at the mine scale. In the numerical model of crack propagation at the mine scale, natural cracks are regarded as frictional interfaces without flow conductivity. The propagation mode of hydraulic cracks under the interference of natural cracks is determined based on the crack intersection criterion. The extension direction of hydraulic cracks after intersecting with natural cracks is determined by analyzing the stress state near the tip of the hydraulic crack. If the maximum tensile stress on one side of the natural crack is greater than or equal to the tensile strength of the rock, the hydraulic crack will pass through the natural crack and continue to propagate. If the shear stress and normal stress acting on the natural crack satisfy the linear friction law, the hydraulic crack is captured by the natural crack and deflected into the natural crack.

5. The method according to claim 1, characterized in that, Based on the fracturing operation parameters of the target block and the dual-medium model of pores and fractures, natural fractures are incorporated into the three-dimensional geomechanical model through a discrete fracture network model. Combined with the fracturing operation parameters, a field-scale fracture propagation numerical model is established to output the simulated fracture geometry, including: The rock deformation and fluid flow at each time step are iteratively coupled and solved in the following manner to obtain the fluid pressure distribution: Repeat the following steps until the model converges: make a preliminary estimate of the fluid pressure inside the fracture; use the rock elastic deformation theory to describe the displacement discontinuity surface, and use the preliminary estimated fluid pressure inside the fracture to calculate the displacement discontinuity surface; after obtaining the displacement discontinuity surface, use the Newton-Raphson iteration method to calculate the fluid pressure distribution, and compare the calculated pressure with the preliminary estimated pressure to verify the convergence; The mass balance equation is used to ensure that the total injected fluid volume is equal to the sum of the fluid volume in the crack and the fluid loss volume; If the time step does not meet the calculation requirements, the fluid pressure inside the crack is estimated and the calculation is performed for the next time step. If the time step meets the calculation requirements, the iterative coupling solution ends.

6. The method according to claim 5, characterized in that, A preliminary estimate of the fluid pressure within the joint is made, including: Calculate the pressure drop within a single crack using the following formula: in, This represents the pressure drop within a single crack. Indicates internal pressure. Indicates the distance along the crack. Indicates the consistency index. Represents the power-law exponent of fluid. Indicates the crack height. Indicates the crack width. Indicates fluid flow rate; Based on the injection displacement of each fracture, the total injection displacement is calculated using the following formula: in, Indicates the total injection displacement. Let N represent the injection displacement of the i-th fracture, and N represent the number of fractures. Calculate the fracture inlet pressure using the following formula: in, The inlet pressure of the crack. Indicates wellhead pressure, For the perforation friction voltage drop, For wellbore friction pressure drop, Let be the perforation friction pressure drop at the i-th crack. For the density of the proppant slurry, The density of water, Let be the number of perforations at the i-th crack. Let be the diameter of the perforation at the i-th crack. Let be the wellbore friction pressure drop at the i-th fracture. Indicates the wellbore friction coefficient. Indicates the location of the crack. Represents the power-law exponent of fluid. This indicates the flow rate in the current section of the wellbore. Based on the pressure drop within the single fracture, the total injection volume, and the fracture inlet pressure, the pressure drop along the fracture is calculated using lubrication theory, and then fully coupled with rock deformation for iterative solution to obtain the fluid pressure within the fracture.

7. The method according to claim 5, characterized in that, The displacement discontinuity surface is calculated using a preliminary estimate of the fluid pressure within the joint, including: Divide the boundary into multiple planar units; Calculate the stress generated by the displacement discontinuity of each planar element; The stresses generated by the displacement discontinuities of each plane element in multiple plane elements are linearly combined to obtain the stress at any point.

8. The method according to claim 7, characterized in that, The stresses generated by the displacement discontinuities of various planar elements in multiple planar elements are linearly combined to obtain the stress at any point, including: Calculate the stress generated at any point within the region due to the displacement discontinuity at point η on the boundary using the following formula: in, Let η be the displacement discontinuity at point η on the boundary, which generates stress at point ξ within the region. For displacement, This is the tensor field representing the effect of concentrated force on displacement. Here, j and k represent the component coordinate indices of the stress tensor, which are the integration boundaries.

9. The method according to claim 7, characterized in that, Calculate the stress generated by the displacement discontinuity of each planar element, including: Under given boundary conditions, the stresses generated by the normal displacement discontinuity and the shear displacement discontinuity of each planar element are obtained by solving the following system of equations: in, in, Let be the total normal stress generated by the surface displacement discontinuity of planar element j. Let J be the total shear stress generated by the surface displacement discontinuity of planar element j. This represents the discontinuity in the normal displacement of planar element j, corresponding to the crack opening. The shear displacement discontinuity of planar element j represents the relative slippage between crack surfaces, and N is the number of planar elements. Let be the elastic influence coefficient matrix under plane strain. Let be the normal stress response induced at plane element i by the discontinuity of shear displacement in plane element j. Let be the distance between the center of crack element i and the center of crack element j. The height of the crack. and These are empirical constants determined to match the actual 3D crack solution.

10. A device for determining the fracturing effect with dual intelligent correction, characterized in that, include: The acquisition module is used to acquire multi-source data of the target block, calculate mechanical parameters through the multi-source data, and obtain discrete well point data of the target block. An interpolation module is used to interpolate the discrete well point data into continuous three-dimensional spatial data to form a three-dimensional geomechanical model of the target block. The output module is used to embed natural fractures into the three-dimensional geomechanical model through a discrete fracture network model based on the dual medium model of pores and fractures using the fracturing construction parameters of the target block, and to establish a field-scale fracture propagation numerical model in combination with the fracturing construction parameters, so as to output the simulated fracture geometry. The primary correction module is used to find the optimal parameter combination in a high-dimensional uncertainty space based on the simulated crack geometry and the actual pumping pressure data, and to automatically fit the pumping pressure curve to obtain the crack geometry after primary correction. A module is established to couple the corrected fracture geometry to the reservoir numerical model through an embedded discrete fracture model, thereby establishing a numerical simulation model of the target block reservoir. The simulation module is used to simulate production dynamics data and reservoir stimulation volume using a numerical simulation model of the target block reservoir and production regime. The secondary correction module is used to establish a reservoir simulation proxy model based on the simulated production dynamic data and the actual production history data, invert reservoir characteristic parameters and fracture geometry, and use the inverted fracture geometry as the fracture geometry after secondary correction. The prediction module is used to predict the long-term productivity and effective stimulation volume of the target block based on reservoir characteristic parameters and fracture geometry after secondary correction.

Images