Fracture parameter-diffusion effect accurate characterization method based on DFN

Abstract

The invention discloses a fracture parameter-diffusion effect accurate characterization method based on DFN, and relates to the technical field of oil and gas exploitation and CO2 geological sequestration, and the method comprises the steps: 1, carrying out dynamic diffusion coefficient coupling; step 2, data acquisition and processing; step 3, constructing a multi-scale DFN model under multi-source constraint and designing a sensitivity scheme; step 4, performing fitting matching; 5, diffusion effect-crack parameter relation quantitative characterization is carried out; according to the crack parameter-diffusion effect accurate characterization method based on the DFN, the problems of inaccurate diffusion behavior characterization, large diffusion effect prediction error and the like in the prior art are solved.

Description

Technical Field

[0001] This invention relates to the field of oil and gas extraction and CO2 geological storage technology, and in particular to a precise characterization method for fracture parameters-diffusion effect based on DFN. Background Technology

[0002] In fields such as CO2 geological storage and unconventional oil and gas reservoir development (e.g., shale oil, tight sandstone gas), higher demands are placed on the precise characterization of fluid transport patterns. Diffusion, as the core mechanism of fluid mass transfer at the micro-to-macro scale, directly affects reservoir utilization efficiency and storage security. In fractured dual-porosity media, fractures act as diffusion channels, and the matrix acts as diffusion storage space. The synergistic effect of fractures and matrix determines key engineering indicators such as CO2 retention rate and oil displacement efficiency. Therefore, accurately quantifying the impact of fracture parameters (density, aperture, etc.) on diffusion effects has become a core technical challenge for the industry.

[0003] However, existing technologies have several bottlenecks. First, there is a serious inaccuracy in characterizing diffusion behavior. This is because the dual-medium model equates the fracture system to a continuous, homogeneous, high-permeability region, ignoring the discrete distribution characteristics and high heterogeneity of natural fractures. This leads to miscalculation of the fracture-matrix contact area and fails to reflect the cross-scale coupling effect between micro-scale and macro-scale diffusion. Second, the dual-porosity media model relies on grid discretization to solve the continuity equation. However, the traditional grid scale is much larger than the micro-diffusion influence range, causing distortion of the matrix interior and resulting in errors. Consequently, the prediction error of diffusion effects is relatively large. Summary of the Invention

[0004] The purpose of this invention is to provide a precise characterization method for crack parameters and diffusion effects based on DFN, which solves the problems of inaccurate diffusion behavior characterization and large diffusion effect prediction errors in existing technologies.

[0005] To achieve the above objectives, this invention provides a precise characterization method for crack parameters and diffusion effects based on DFN, comprising the following steps: Step 1: Perform dynamic diffusion coefficient coupling; Step 2: Data Acquisition and Processing; Step 3: Construction of multi-scale DFN model and design of sensitivity scheme under multi-source constraints; Step 4: Perform fitting and matching; Step 5: Quantitative characterization of the relationship between diffusion effect and crack parameters.

[0006] Preferably, in step 1: S11. Solve for the diffusion coefficient using the Sigmund method and the collision integral formula; S12. Construct the seepage equation.

[0007] Preferably, in S11: The basic diffusion coefficient was calculated using the Sigmund method. The expression is: ; In the formula, This represents the molar density of the mixture under standard conditions, expressed in gmole / m³. 3 ; The binary diffusion coefficient of a mixture under standard conditions, expressed in cm. 2 / s; This represents the molar density of the mixture at the temperature and pressure of the reservoir, expressed in gmole / m³. 3 ; This represents the relative density of a mixture under standard conditions; it is dimensionless. Bulk density, unit: kg / m³ 3 ; satisfy: ; In the formula, Indicates the number of fluid phases (oil, gas, water), dimensionless; Indicates components mole fraction; Indicates components Critical molar volume, in m³ 3 / gmole; The thermodynamic temperature of a fluid is expressed in Kelvin (K). express , The diameter of the colliding molecules between components, in cm; express , The diffusion-collision integral of the components is dimensionless. This represents the universal gas constant, which is dimensionless. and Each represents a component and components The molecular weight is expressed in g / gmole. The dimensionless collision integral under reservoir conditions is obtained through the collision integral formula, and its expression is: ; In the formula, express , Intermolecular collision integral; denoted as the first fitting constant, matching the power-law variation trend of the collision integral in the high-temperature range; This represents the power-law constant of temperature and its power-law dependence with temperature. This represents the second fitting constant, which matches the exponential decay trend of the collision integral in the low-temperature range; The third fitting constant represents the change in collision integrals under strong intermolecular interactions in the low-temperature range; The first natural exponent term is used to fit the nonlinear change of intermolecular "short-range repulsion" with temperature. The second natural exponent term represents the nonlinear change of intermolecular "long-range attraction" (such as van der Waals forces) with temperature. Obtaining the components in the oil and gas phase system exist The expression for the diffusion coefficient in the phase is: ; In the formula, Indicates components The diffusion coefficient.

[0008] Preferably, in S12: The governing equation for the migration of hydrocarbon components within the seepage unit in the matrix is: ; In the formula, Indicates porosity; Indicates components Fraction in the oil phase; This is the oil phase density, measured in gmole / cm³. 3 ; and These represent oil phase saturation and gas phase saturation, respectively. It is a component Fraction in the gas phase; This is the gas phase density, measured in gmole / cm³. 3 ; and These represent the seepage velocities of the oil phase and the gas phase, respectively, both in cm / s. This is the diffusion term in the matrix flow equation, representing the transport flux of the target component through molecular diffusion and mechanical dispersion, with units of gmole / (cm). 2 ·s); For source and sink terms, this represents the injection / production rate of the target component, expressed in gmole / (cm³). 3 ·s); Incorporating the diffusion term into the component transport equation of the DFN crack, the components... The equation for the total diffusion flux is expressed as follows: ; In the formula, Indicates components In all phases Total mass transport in (oil, gas, water), unit gmole / (cm2·s); Indicates different phases; This represents phase density, with units of gmole / cm3; express Phase saturation; Indicates components exist The diffusion coefficient of a phase, expressed in cm² / s; Indicates components exist mole fraction of the phase; This indicates the derivative with respect to length, with units of 1 / cm; Mass conservation equation for the matrix region: ; In the formula, Darcy flux within fractures in the matrix system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the cracks in the matrix system, measured in gmole / (cm²). 2 ·s); This represents the matrix-crack channeling source term, in gmole / (cm). 3 ·s); Indicates the composition per unit pore volume Total number of moles, in gmole / cm³ 3 ; Indicates time, in seconds; Represents the matrix system; Mass conservation equation for the crack region: ; In the formula, Darcy flux within a fracture system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the fracture system, measured in gmole / (cm²). 2 ·s); Indicates source and sink terms, unit gmole / (cm) 3 ·s); Indicates a crack system; Effective diffusion coefficient equation: ; In the formula, This represents the effective diffusion coefficient after considering the rock curvature correction, in cm. 2 / s; This represents the total diffusion coefficient, expressed in cm.2 / s; Indicates curvature, dimensionless; Decomposition of matrix-crack total transport term: ; In the formula, This represents the Darcy flow term driven by pressure difference, in gmole / (cm²). 3 ·s); This represents the diffusion crossflow term driven by the concentration difference of components, in gmole / (cm³). 3 ·s); Matrix-fracture transfer term under Darcy action: ; In the formula, This represents the shape factor, with units of 1 / cm. 2 ; This represents the volume of the entire grid block, in cm. 3 ; This represents absolute penetration rate, expressed in mD. Indicates the first The molar density of the phase, in gmole / cm³ 3 ; Indicates components In the The mole fraction of a phase, dimensionless; Indicates the first The viscosity of a phase, expressed in mPa·s; This represents matrix pressure, in kPa. This represents the crack pressure, measured in kPa. Matrix-crack transfer term under diffusion: ; In the formula, This represents the diffusion shape factor, with units of 1 / cm. 2 ; Indicates components Mole fraction in the matrix, Indicates components Mole fraction in the crack; The Peng-Robinson state equation is expressed as follows: ; In the formula, The thermodynamic pressure of a fluid is expressed in Pa. This represents the universal gas constant, with units of J / (gmole·K); The thermodynamic temperature of a fluid is expressed in Kelvin (K). The molar volume of a fluid is expressed in cm³. 3 / gmole; This parameter, which is temperature-dependent, represents the strength of intermolecular attraction and is measured in Pa·(cm²). 3 / gmole) 2 ; The covolute (molar volume of the molecule itself) is expressed in cm³. 3 / gmole; The component constraint equation is: ; ; ; ; In the formula, Indicates the component's serial number; Indicates the number of components contained in the system; Indicates the first in the overall system Mole fraction of each component; Indicates the first phase of the oil phase Mole fraction of each component; Indicates the first gas phase Mole fraction of each component; and These represent phase distribution correction factors in the oil and gas phases, respectively, both of which are dimensionless. The saturation constraint equation is: ; In the formula, Indicates the water phase saturation.

[0009] Preferably, in step 2: S21, Multi-dimensional data collection; S22, Data Preprocessing.

[0010] Preferably, in S21: Collect geological parameters of the target block, including porosity, permeability, and fluid properties; fluid properties include viscosity and density. Extract the actual value range of the key crack parameters, which include crack density, aperture, and orientation.

[0011] Preferably, in S22: Geological parameters and key fracture parameters are preprocessed. The preprocessing operations include data cleaning, data denoising and standardization to eliminate outliers and data redundancy.

[0012] Preferably, in step 3: S31. Fusion and statistical regularity analysis of multi-source geological data; Based on the pre-processed core description, FMI imaging logging and 3D seismic attribute volume, the fracture development characteristics are quantitatively characterized. The attitude and linear density of micro-fractures around the well are identified using FMI logging data. The statistical distribution function of fracture development (such as power law distribution or log-normal distribution) is fitted, and a large-scale fracture network is extracted by combining ant-tracking or coherent seismic attributes. A statistical probability model including fracture length, aperture, azimuth, and intensity is established to provide seed points and constraints for stochastic modeling. The attitude of the fractures includes strike and dip. S32, "Deterministic + Random" Multi-Scale DFN Network Fusion; A multi-scale discrete fracture network (DFN) model is constructed using a hierarchical modeling strategy: Large-scale (deterministic modeling): For earthquake-visible faults and large cracks, deterministic modeling methods are used to accurately locate their spatial positions and extension trajectories, construct the main fluid seepage channel framework, and obtain deterministic large cracks; Small-scale (random modeling): For microcracks developed inside the matrix, based on the above statistical laws, random microcrack patches that conform to geological laws are generated in three-dimensional space through fractal geometry theory and Monte Carlo simulation algorithm. By fusing deterministic large fractures with random micro fractures, a multi-scale DFN model is obtained, which reproduces the full-scale geological features from macroscopic fractures to microscopic seepage networks. S33, Design of a crack parameter sensitivity comparison scheme; The "standard benchmark model" that reproduces the actual block characteristics (1:1 replica of crack strength, average aperture and dominant direction) is used as the standard scheme, and the standard scheme is used as the control group. Multiple parameter sensitivity test schemes are designed, including crack density sensitivity scheme, crack aperture sensitivity scheme and crack direction sensitivity scheme. Fracture density sensitivity scheme: Adjust the fracture surface density parameter to construct DFN models with low, medium and high development levels, thereby changing the shape factor of the matrix rock block and directly related to the contact area of ​​diffusion mass transfer; Crack aperture sensitivity scheme: Set the range of crack aperture distribution with stepwise variation, simulate the difference in crack conductivity under different closure pressures, and analyze the control effect of closure pressure on the diffusion rate of concentration field. Crack orientation sensitivity scheme: change the dispersion of the dominant azimuth angle of the crack, construct a comparative model of isotropic and strong anisotropic cracks, and analyze the influence of the crack network connectivity direction on the diffusion range. S34, Attribute coarsening and dual-pore medium parameter field generation; Based on the constructed multi-scale DFN model, the discrete fracture network properties are coarsened into the geological grid using the Oda algorithm or the streamline-based permeability tensor calculation method. The focus is on calculating and outputting the equivalent fracture porosity, equivalent fracture permeability tensor, and matrix-fracture shape factor three-dimensional parameter field, providing a high-precision static model foundation for subsequent numerical simulation of diffusion effects.

[0013] Preferably, in step 4: S41. The simulation results of the standard scheme are systematically fitted and matched with the multi-dimensional historical production data of the target block. By fine-tuning the key parameters of the geological model in the standard scheme, the adaptability of the geological model to the actual engineering scenario is optimized. The core comparison indicators in the fitting process include cumulative CO2 injection, daily crude oil production, bottom hole pressure change curve, CO2 mole fraction at the production end, and water cut evolution. S42. Quantitatively evaluate the fit: Combine the root mean square error (RMSE) and the coefficient of determination (R²). 2 () is used as a quantitative evaluation index for the fitting process to quantitatively evaluate the fitting effect; S43. Determine if the fit meets the requirements; if the fit is not less than 85%, proceed to the next step; if the fit is not greater than 85%, return to step 3.

[0014] Preferably, in step 5: S51. Conduct numerical simulation experiments on multiple sets of key fracture parameters to obtain a diffusion effect dataset; the diffusion effect dataset includes CO2 diffusion range, leading edge advance velocity, matrix swept volume, CO2 sequestration rate, and crude oil production increase, etc. S52. Add a traditional dual-medium model as a control group to compare the diffusion effect prediction results of the traditional dual-medium model and the DFN fracture network model, including the diffusion range and diffusion-enhanced extraction degree. S53. The relationship between crack parameters and diffusion effect is quantified through methods such as sensitivity analysis and correlation fitting.

[0015] Therefore, the present invention employs the above-mentioned precise characterization method of crack parameters-diffusion effect based on DFN, which has the following beneficial effects: (1) Significantly improved accuracy of cross-scale diffusion characterization: This method overcomes the accuracy bottleneck caused by the homogenization assumption and grid scale mismatch in existing dual-porosity media models, reducing the prediction of cross-scale diffusion effects; taking a certain block as an example, the comprehensive fit between the prediction results of this method for CO2 diffusion range and sequestration rate and actual engineering data reaches 89%, of which the coefficient of determination for crude oil production (R) is significantly higher. 2The sealing rate is 0.91, and the fitting error is only 4.2%, which is far better than the industry-recognized reliability threshold of 85%, accurately capturing the diffusion behavior of non-mainstream channels such as micropores, blind ends, and small pore throats; (2) Clarify the quantitative correlation between crack parameters and diffusion effect: For the first time, a quantitative mapping relationship between key crack parameters (density, aperture, orientation) and key diffusion indicators (range, rate, sweep efficiency) under the DFN framework is established, and the sensitivity law of each parameter to the diffusion effect is clarified. For example, in a certain block application, the crack parameter-diffusion effect can be accurately quantified by this method, which solves the industry pain point that traditional technology cannot quantify the impact of crack parameters and provides a clear quantitative basis for parameter control. (3) Significantly enhanced engineering adaptability: By restoring the actual geological conditions and natural fracture characteristics of the block at a 1:1 ratio, and combining historical production data to reverse-correct the model parameters, the consistency between the diffusion behavior simulation and the actual engineering is improved; after optimizing the fracture modification scheme using the method of this invention, the matrix sweep efficiency, unconventional oil and gas reservoir recovery rate and CO2 geological sequestration rate of the target block are all improved; at the same time, the universal design for different blocks can be quickly adapted to various fractured reservoirs such as shale oil, tight sandstone gas, and high sulfur gas reservoirs, and can be implemented without significant adjustments to the core logic; (4) Expanded the application scenarios of DFN technology: Filled the gap in the existing DFN technology which focuses on fluid convection and mechanical property analysis and lacks a special evaluation system for diffusion effect, forming a complete technical chain of "data acquisition-model construction-coupled calculation-fit verification-law quantification"; This method can not only accurately characterize diffusion behavior, but also simultaneously output engineering results such as crack modification optimization suggestions and CO2 injection parameter adjustment schemes, so that DFN technology can be upgraded from "theoretical simulation tool" to "engineering decision support system", providing full-process technical support for low-carbon development scenarios such as CO2 oil recovery and geological storage.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is an overall flowchart of the method for accurately characterizing crack parameters and diffusion effects based on DFN according to the present invention; Figure 2 This is a roadmap of the DFN multi-scale crack hierarchical modeling technology according to an embodiment of the present invention; Figure 3 This is a graph showing the fitting effect between a standard block scheme and historical data in an embodiment of the present invention. Figure 4 This is a quantitative relationship diagram of crack contact area and diffusion effect in a standard block scheme according to an embodiment of the present invention; Figure 5This is a comparison chart of the diffusion effect accuracy between the DFN model and the traditional dual-medium model in this embodiment of the invention. Detailed Implementation

[0018] The following detailed description of embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0019] Please see Figure 1-5 A precise characterization method for crack parameters and diffusion effects based on DFN includes the following steps: Step 1: Perform dynamic diffusion coefficient coupling; S11. Solve for the diffusion coefficient using the Sigmund method and the collision integral formula, specifically as follows: The basic diffusion coefficient was calculated using the Sigmund method. The expression is: ; In the formula, This represents the molar density of a mixture under standard conditions, expressed in gmole / cm³. 3 ; The binary diffusion coefficient of a mixture under standard conditions, expressed in cm. 2 / s; This represents the molar density of the mixture at the temperature and pressure of the reservoir, expressed in gmole / cm³. 3 ; This represents the relative density of a mixture under standard conditions; it is dimensionless. Bulk density, unit: kg / m³ 3 ;in satisfy: ; In the formula, Indicates the number of fluid phases (oil, gas, water), dimensionless; Indicates components mole fraction; Indicates components Critical molar volume, in cm³ 3 / gmole; The thermodynamic temperature of a fluid is expressed in Kelvin (K). express , The diameter of the colliding molecules between components, in cm; express , The diffusion-collision integral of the components is dimensionless. This represents the universal gas constant, which is dimensionless. and Each represents a component and components The molecular weight is expressed in g / gmole. The dimensionless collision integral under reservoir conditions is obtained through the collision integral formula, and its expression is: ; In the formula, express , Intermolecular collision integral; denoted as the first fitting constant, matching the power-law variation trend of the collision integral in the high-temperature range; This represents the power-law constant of temperature and its power-law dependence with temperature. This represents the second fitting constant, which matches the exponential decay trend of the collision integral in the low-temperature range; The third fitting constant represents the change in collision integrals under strong intermolecular interactions in the low-temperature range; The first natural exponent term is used to fit the nonlinear change of intermolecular "short-range repulsion" with temperature. The second natural exponent term represents the nonlinear change of intermolecular "long-range attraction" (such as van der Waals forces) with temperature. The above formulas can be used to obtain the components in the oil-gas phase system. exist The expression for the diffusion coefficient in the phase is: ; In the formula, Indicates components The diffusion coefficient; S12. Construct the seepage equation; The governing equation for the migration of hydrocarbon components within the seepage unit in the matrix is: ; In the formula, Indicates porosity; Indicates components Fraction in the oil phase; This is the oil phase density, measured in gmole / cm³. 3 ; and These represent oil phase saturation and gas phase saturation, respectively. It is a component Fraction in the gas phase; This is the gas phase density, measured in gmole / cm³. 3 ; and These represent the seepage velocities of the oil phase and the gas phase, respectively, both in cm / s. This is the diffusion term in the matrix flow equation, representing the transport flux of the target component through molecular diffusion and mechanical dispersion, with units of gmole / (cm). 2 ·s); For source and sink terms, this represents the injection / production rate of the target component, expressed in gmole / (cm³). 3 ·s); Incorporating the diffusion term into the component transport equation of the DFN crack, the components... The equation for the total diffusion flux is expressed as follows: ; In the formula, Indicates components In all phases Total mass transport in (oil, gas, water), unit gmole / (cm³) 2 ·s); Indicates different phases; Phase density is expressed in gmole / cm³. 3 ; express Phase saturation; Indicates components exist The diffusion coefficient of a phase, in cm. 2 / s; Indicates components exist mole fraction of the phase; This indicates the derivative with respect to length, with units of 1 / cm; Mass conservation equation for the matrix region: ; In the formula, Darcy flux within fractures in the matrix system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the cracks in the matrix system, measured in gmole / (cm²). 2 ·s); This represents the matrix-crack channeling source term, in gmole / (cm). 3 ·s); Indicates the composition per unit pore volume Total number of moles, in gmole / cm³ 3 ; Indicates time, in seconds; Represents the matrix system; Mass conservation equation for the crack region: ; In the formula, Darcy flux within a fracture system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the fracture system, measured in gmole / (cm²). 2 ·s); Indicates source and sink terms, unit gmole / (cm) 3 ·s); Indicates a crack system; Effective diffusion coefficient equation: ; In the formula, This represents the effective diffusion coefficient after considering the rock curvature correction, in cm. 2 / s; This represents the total diffusion coefficient, expressed in cm. 2 / s; Indicates curvature, dimensionless; Decomposition of matrix-crack total transport term: ; In the formula, This represents the Darcy flow term driven by pressure difference, in gmole / (cm²). 3 ·s); This represents the diffusion crossflow term driven by the concentration difference of components, in gmole / (cm³). 3 ·s); Matrix-fracture transfer term under Darcy action: ; In the formula, This represents the shape factor, with units of 1 / cm. 2 ; This represents the volume of the entire grid block, in cm. 3 ; This represents absolute penetration rate, expressed in mD. Indicates the first The molar density of the phase, in gmole / cm³ 3 ; Indicates components In the The mole fraction of a phase, dimensionless; Indicates the first The viscosity of a phase, expressed in mPa·s; This represents matrix pressure, in kPa. This represents the crack pressure, measured in kPa. Matrix-crack transfer term under diffusion: ; In the formula, This represents the diffusion shape factor, with units of 1 / cm. 2 ; Indicates components Mole fraction in the matrix, Indicates components Mole fraction in the crack; The Peng-Robinson state equation is expressed as follows: ; In the formula, The thermodynamic pressure of a fluid is expressed in Pa. This represents the universal gas constant, with units of J / (gmole·K); The thermodynamic temperature of a fluid is expressed in Kelvin (K). The molar volume of a fluid is expressed in cm³. 3 / gmole; This parameter, which is temperature-dependent, represents the strength of intermolecular attraction and is measured in Pa·(cm²). 3 / gmole) 2 ; The covolute (molar volume of the molecule itself) is expressed in cm³. 3 / gmole; The component constraint equation is: ; ; ; ; In the formula, Indicates the component's serial number; Indicates the number of components contained in the system; Indicates the first in the overall system Mole fraction of each component; Indicates the first phase of the oil phase Mole fraction of each component; Indicates the first gas phase Mole fraction of each component; and These represent phase distribution correction factors in the oil and gas phases, respectively, both of which are dimensionless. The saturation constraint equation is: ; In the formula, Indicates the water phase saturation.

[0020] Step 2: Data Acquisition and Processing; S21, Multi-dimensional data collection; Collect geological parameters of the target block, including porosity, permeability, and fluid properties; fluid properties include viscosity and density. Extract the actual value range of key crack parameters, which include crack density, aperture, and orientation. S22, Data preprocessing; Geological parameters and key fracture parameters are preprocessed. The preprocessing operations include data cleaning, data denoising and standardization to eliminate outliers and data redundancy.

[0021] Step 3: Construction of multi-scale DFN model and design of sensitivity scheme under multi-source constraints; S31. Fusion and statistical regularity analysis of multi-source geological data; Based on the pre-processed core description, FMI imaging logging and 3D seismic attribute volume, the fracture development characteristics are quantitatively characterized. The attitude and linear density of micro-fractures around the well are identified using FMI logging data. The statistical distribution function of fracture development (such as power law distribution or log-normal distribution) is fitted, and a large-scale fracture network is extracted by combining ant-tracking or coherent seismic attributes. A statistical probability model including fracture length, aperture, azimuth, and intensity is established to provide seed points and constraints for stochastic modeling. The attitude of the fractures includes strike and dip. S32, "Deterministic + Random" Multi-Scale DFN Network Fusion; A multi-scale Discrete Fracture Network (DFN) model is constructed using a hierarchical modeling strategy: Large-scale (deterministic modeling): For earthquake-visible faults and large cracks, deterministic modeling methods are used to accurately locate their spatial positions and extension trajectories, construct the main fluid seepage channel framework, and obtain deterministic large cracks; Small-scale (random modeling): For microcracks developed inside the matrix, based on the above statistical laws, random microcrack patches that conform to geological laws are generated in three-dimensional space through fractal geometry theory and Monte Carlo simulation algorithm. By fusing deterministic large fractures with random micro fractures, a multi-scale DFN model is obtained, which reproduces the full-scale geological features from macroscopic fractures to microscopic seepage networks. S33, Design of a crack parameter sensitivity comparison scheme; The "standard benchmark model" that reproduces the actual block characteristics (1:1 replica of crack strength, average aperture and dominant direction) is used as the standard scheme, and the standard scheme is used as the control group. Multiple parameter sensitivity test schemes are designed, including crack density sensitivity scheme, crack aperture sensitivity scheme and crack direction sensitivity scheme. Fracture density sensitivity scheme: Adjust the fracture surface density parameter to construct DFN models with low, medium and high development levels, thereby changing the shape factor of the matrix rock block and directly related to the contact area of ​​diffusion mass transfer; Crack aperture sensitivity scheme: Set the range of crack aperture distribution with stepwise variation, simulate the difference in crack conductivity under different closure pressures, and analyze the control effect of closure pressure on the diffusion rate of concentration field. Crack orientation sensitivity scheme: change the dispersion of the dominant azimuth angle of the crack, construct a comparative model of isotropic and strong anisotropic cracks, and analyze the influence of the crack network connectivity direction on the diffusion range. S34, Attribute coarsening and dual-pore medium parameter field generation; Based on the constructed multi-scale DFN model, the discrete fracture network properties are coarsened into the geological grid using the Oda algorithm or the streamline-based permeability tensor calculation method. The focus is on calculating and outputting the equivalent fracture porosity, equivalent fracture permeability tensor, and matrix-fracture shape factor three-dimensional parameter field, providing a high-precision static model foundation for subsequent numerical simulation of diffusion effects.

[0022] Step 4: Fitting and matching; S41. The simulation results of the standard scheme are systematically fitted and matched with the multi-dimensional historical production data of the target block. By fine-tuning the key parameters of the geological model in the standard scheme, the adaptability of the geological model to the actual engineering scenario is optimized. The core comparison indicators in the fitting process include cumulative CO2 injection, daily crude oil production, bottom hole pressure change curve, CO2 mole fraction at the production end, and water cut evolution. S42. Quantitatively evaluate the fit: Combine the root mean square error (RMSE) and the coefficient of determination (R²). 2 () is used as a quantitative evaluation index for the fitting process to quantitatively evaluate the fitting effect; S43. Determine if the fit meets the requirements; if the fit is not less than 85%, proceed to the next step; if the fit is not greater than 85%, return to step 3.

[0023] Taking a certain block as an example, after multiple rounds of parameter correction, the overall fit of the core production indicators reached 89%, among which the fit of crude oil production was R. 2With a fitting error of only 4.2% and a CO2 sequestration rate of 0.91, the model meets the industry-recognized reliability threshold (fit ≥ 85%). This fitting result verifies that the constructed DFN model can accurately reproduce the fluid transport and diffusion laws of the actual block, laying a reliable foundation for subsequent quantitative analysis of the influence of crack parameters on diffusion effect. Step 5: Quantitative characterization of the relationship between diffusion effect and crack parameters; S51. Conduct numerical simulation experiments on multiple sets of key fracture parameters to obtain a diffusion effect dataset; the diffusion effect dataset includes CO2 diffusion range, leading edge advance velocity, matrix swept volume, CO2 sequestration rate, and crude oil production increase, etc. S52. Add a traditional dual-medium model as a control group to compare the diffusion effect prediction results of the two models, including diffusion range and diffusion extraction degree; S53. The relationship between crack parameters and diffusion effect is quantified through methods such as sensitivity analysis and correlation fitting.

[0024] Taking a certain block as an example, fracture density is positively correlated with diffusion range. When the density increases to a critical value, the diffusion efficiency improvement slows down. Fracture aperture has a significant impact on diffusion rate. Increasing the aperture can reduce mass transfer resistance, but excessive increase can easily lead to gas channeling risk. The matching degree between fracture orientation and injection well pattern directly affects sweep uniformity. Optimizing orientation can improve matrix sweep efficiency by 10%-15%. The resulting quantitative relationships and influence patterns can provide a scientific basis for optimizing fracture stimulation schemes and adjusting CO2 injection parameters in actual blocks, helping to improve oil and gas recovery and storage safety.

[0025] Therefore, the present invention employs the above-mentioned precise characterization method of crack parameters-diffusion effect based on DFN, which has the following beneficial effects: (1) Significantly improved accuracy of cross-scale diffusion characterization: This method overcomes the accuracy bottleneck caused by the homogenization assumption and grid scale mismatch in existing dual-porosity media models, reducing the prediction of cross-scale diffusion effects; taking a certain block as an example, the comprehensive fit between the prediction results of this method for CO2 diffusion range and sequestration rate and actual engineering data reaches 89%, of which the coefficient of determination for crude oil production (R) is significantly higher. 2 The sealing rate is 0.91, and the fitting error is only 4.2%, which is far better than the industry-recognized reliability threshold of 85%, accurately capturing the diffusion behavior of non-mainstream channels such as micropores, blind ends, and small pore throats; (2) Clarify the quantitative correlation between crack parameters and diffusion effect: For the first time, a quantitative mapping relationship between key crack parameters (density, aperture, orientation) and key diffusion indicators (range, rate, sweep efficiency) under the DFN framework is established, and the sensitivity law of each parameter to the diffusion effect is clarified. For example, in a certain block application, the crack parameter-diffusion effect can be accurately quantified by this method, which solves the industry pain point that traditional technology cannot quantify the impact of crack parameters and provides a clear quantitative basis for parameter control. (3) Significantly enhanced engineering adaptability: By restoring the actual geological conditions and natural fracture characteristics of the block at a 1:1 ratio, and combining historical production data to reverse-correct the model parameters, the consistency between the diffusion behavior simulation and the actual engineering is improved; after optimizing the fracture modification scheme using the method of this invention, the matrix sweep efficiency, unconventional oil and gas reservoir recovery rate and CO2 geological sequestration rate of the target block are all improved; at the same time, the universal design for different blocks can be quickly adapted to various fractured reservoirs such as shale oil, tight sandstone gas, and high sulfur gas reservoirs, and can be implemented without significant adjustments to the core logic; (4) Expanded the application scenarios of DFN technology: Filled the gap in the existing DFN technology which focuses on fluid convection and mechanical property analysis and lacks a special evaluation system for diffusion effect, forming a complete technical chain of "data acquisition-model construction-coupled calculation-fit verification-law quantification"; This method can not only accurately characterize diffusion behavior, but also simultaneously output engineering results such as crack modification optimization suggestions and CO2 injection parameter adjustment schemes, so that DFN technology can be upgraded from "theoretical simulation tool" to "engineering decision support system", providing full-process technical support for low-carbon development scenarios such as CO2 oil recovery and geological storage.

[0026] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A precise characterization method for crack parameters and diffusion effects based on DFN, characterized in that, Includes the following steps: Step 1: Perform dynamic diffusion coefficient coupling; Step 2: Data Acquisition and Processing; Step 3: Construction of multi-scale DFN model and design of sensitivity scheme under multi-source constraints; Step 4: Perform fitting and matching; Step 5: Quantitative characterization of the relationship between diffusion effect and crack parameters.

2. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 1, characterized in that, In step 1: S11. Solve for the diffusion coefficient using the Sigmund method and the collision integral formula; S12. Construct the seepage equation.

3. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 2, characterized in that, In S11: The basic diffusion coefficient was calculated using the Sigmund method. The expression is: ； In the formula, This represents the molar density of a mixture under standard conditions, expressed in gmole / cm³. 3 ; The binary diffusion coefficient of a mixture under standard conditions, expressed in cm. 2 / s; This represents the molar density of the mixture at the temperature and pressure of the reservoir, expressed in gmole / cm³. 3 ; This represents the relative density of a mixture under standard conditions; it is dimensionless. Bulk density, unit: kg / m³ 3 ;in satisfy: ； In the formula, Represents the number of fluid phases; dimensionless. Indicates components mole fraction; Indicates components Critical molar volume, in cm³ 3 / gmole; The thermodynamic temperature of a fluid is expressed in Kelvin (K). express , The diameter of the colliding molecules between components, in cm; express , The diffusion-collision integral of the components is dimensionless. This represents the universal gas constant, which is dimensionless. and Each represents a component and components The molecular weight is expressed in g / gmole. The dimensionless collision integral under reservoir conditions is obtained through the collision integral formula, and its expression is: ； In the formula, express , Fitted expression for the intermolecular collision integral as a function of temperature; denoted as the first fitting constant, matching the power-law variation trend of the collision integral in the high-temperature range; represents the temperature power constant, and represents the power-law dependence of the collision integral on temperature; This represents the second fitting constant, which matches the exponential decay trend of the collision integral in the low-temperature range; The third fitting constant represents the change in collision integrals under strong intermolecular interactions in the low-temperature range; The first natural exponent term is used to fit the nonlinear change of intermolecular "short-range repulsion" with temperature. The second natural exponent term represents the nonlinear change of intermolecular "long-range attraction" with temperature. Obtaining the components in the oil and gas phase system exist The expression for the diffusion coefficient in the phase is: ； In the formula, Indicates components The diffusion coefficient.

4. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 3, characterized in that, In S12: The governing equation for the migration of hydrocarbon components within the seepage unit in the matrix is: ； In the formula, Indicates porosity; Indicates components Fraction in the oil phase; This is the oil phase density, measured in gmole / cm³. 3 ; and These represent oil phase saturation and gas phase saturation, respectively. It is a component Fraction in the gas phase; This is the gas phase density, measured in gmole / cm³. 3 ; and These represent the seepage velocities of the oil phase and the gas phase, respectively, both in cm / s. This is the diffusion term in the matrix flow equation, representing the transport flux of the target component through molecular diffusion and mechanical dispersion, with units of gmole / (cm). 2 ·s); For source and sink terms, this represents the injection / production rate of the target component, expressed in gmole / (cm³). 3 ·s); Incorporating the diffusion term into the component transport equation of the DFN crack, the components... The equation for the total diffusion flux is expressed as follows: ； In the formula, Indicates components In all phases Total mass transported in gmole / (cm³) 2 ·s); Indicates different phases; Phase density is expressed in gmole / cm³. 3 ; express Phase saturation; Indicates components exist The diffusion coefficient of a phase, in cm. 2 / s; Indicates components exist mole fraction of the phase; This indicates the derivative with respect to length, with units of 1 / cm; Mass conservation equation for the matrix region: ； In the formula, Darcy flux within fractures in the matrix system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the cracks in the matrix system, expressed in gmole / (cm²). 2 ·s); This represents the matrix-fracture channeling source term, in gmole / (cm). 3 ·s); Indicates the composition per unit pore volume Total number of moles, in gmole / cm³ 3 ; Indicates time, in seconds; Represents the matrix system; Mass conservation equation for the crack region: ； In the formula, Darcy flux within a fracture system, expressed in gmole / (cm²). 2 ·s); This represents the diffusion flux within the fracture system, measured in gmole / (cm²). 2 ·s); Indicates source and sink terms, unit gmole / (cm) 3 ·s); Indicates a crack system; Effective diffusion coefficient equation: ； In the formula, This represents the effective diffusion coefficient after considering the rock curvature correction, in cm. 2 / s; This represents the total diffusion coefficient, expressed in cm. 2 / s; Indicates curvature, dimensionless; Decomposition of matrix-crack total transport term: ； In the formula, This represents the Darcy flow term driven by pressure difference, in gmole / (cm²). 3 ·s); This represents the diffusion crossflow term driven by the concentration difference of components, in gmole / (cm³). 3 ·s); Matrix-fracture transfer term under Darcy action: ； In the formula, This represents the shape factor, with units of 1 / cm. 2 ; This represents the volume of the entire grid block, in cm. 3 ; This represents absolute penetration rate, expressed in mD. Indicates the first The molar density of the phase, in gmole / cm³ 3 ; Indicates components In the The mole fraction of a phase, dimensionless; Indicates the first The viscosity of a phase, expressed in mPa·s; This represents matrix pressure, in kPa. This represents the crack pressure, measured in kPa. Matrix-crack transfer term under diffusion: ； In the formula, This represents the diffusion shape factor, with units of 1 / cm. 2 ; Indicates components Mole fraction in the matrix, Indicates components Mole fraction in the crack; The Peng-Robinson state equation is expressed as follows: ； In the formula, The thermodynamic pressure of a fluid is expressed in Pa. This represents the universal gas constant, with units of J / (gmole·K); The thermodynamic temperature of a fluid is expressed in Kelvin (K). The molar volume of a fluid is expressed in cm³. 3 / gmole; This parameter, which is temperature-dependent, represents the strength of intermolecular attraction and is measured in Pa·(cm²). 3 / gmole) 2 ; The covolute (molar volume of the molecule itself) is expressed in cm³. 3 / gmole; The component constraint equation is: ； ； ； ； In the formula, Indicates the component's serial number; Indicates the number of components contained in the system; Indicates the first in the overall system Mole fraction of each component; Indicates the first phase of the oil phase Mole fraction of each component; Indicates the first gas phase Mole fraction of each component; and These represent phase distribution correction factors in the oil and gas phases, respectively, both of which are dimensionless. The saturation constraint equation is: ； In the formula, Indicates the water phase saturation.

5. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 4, characterized in that, In step 2: S21, Multi-dimensional data collection; S22, Data Preprocessing.

6. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 5, characterized in that, In S21: Collect geological parameters of the target block, including porosity, permeability, and fluid properties; fluid properties include viscosity and density. Extract the actual value range of the key crack parameters, which include crack density, aperture, and orientation.

7. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 6, characterized in that, In S22: Geological parameters and key fracture parameters are preprocessed, including data cleaning, data denoising, and standardization.

8. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 7, characterized in that, In step 3: S31. Fusion and statistical regularity analysis of multi-source geological data; Based on the pre-processed core description, FMI imaging logging and 3D seismic attribute volume, the fracture development characteristics are quantitatively characterized. The attitude and linear density of micro-fractures around the well are identified by FMI logging data, the statistical distribution function of fracture development is fitted, and a large-scale fracture network is extracted by combining the seismic attributes of ant bodies or coherent bodies; a statistical probability model including fracture length, aperture, azimuth and intensity is established to provide seed points and constraints for stochastic modeling; the attitude of fractures includes strike and dip. S32, "Deterministic + Random" Multi-Scale DFN Network Fusion; Constructing a multi-scale DFN model using a hierarchical modeling strategy: Deterministic modeling: For earthquake-visible faults and large cracks, a deterministic modeling method is used to lock the spatial location and extension trajectory, construct the fluid seepage channel framework, and obtain the deterministic large cracks; Random modeling: For microcracks developing inside the matrix, random microcrack patches that conform to geological laws are generated in three-dimensional space using fractal geometry theory and Monte Carlo simulation algorithm; By fusing deterministic large fractures with random micro fractures, a multi-scale DFN model is obtained, which reproduces the full-scale geological features from macroscopic fractures to microscopic seepage networks. S33, Design of a crack parameter sensitivity comparison scheme; Using the "standard benchmark model" that restores the actual block characteristics as the standard scheme and the standard scheme as the control group, we designed multiple parameter sensitivity test schemes, including crack density sensitivity scheme, crack aperture sensitivity scheme and crack orientation sensitivity scheme. Fracture density sensitivity scheme: Adjust the fracture surface density parameter, construct DFN models with low, medium and high development levels, change the shape factor of the matrix rock block, and directly correlate the contact area of ​​diffusion mass transfer; Crack aperture sensitivity scheme: Set the range of crack aperture distribution with stepwise variation, simulate the difference in crack conductivity under different closure pressures, and analyze the control effect of closure pressure on the diffusion rate of concentration field. Crack orientation sensitivity scheme: change the dispersion of the dominant azimuth angle of the crack, construct a comparative model of isotropic and strong anisotropic cracks, and analyze the influence of the crack network connectivity direction on the diffusion range. S34, Attribute coarsening and dual-pore medium parameter field generation; Based on the constructed multi-scale DFN model, the discrete fracture network properties are coarsened into the geological grid using the Oda algorithm or the streamline-based permeability tensor calculation method. The equivalent fracture porosity, equivalent fracture permeability tensor, and matrix-fracture shape factor three-dimensional parameter field are calculated and output.

9. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 8, characterized in that, In step 4: S41. Fit the simulation results of the standard scheme with the historical production data of the target block, and fine-tune the key parameters of the geological model in the standard scheme in reverse. The core comparison indicators in the fitting process include cumulative CO2 injection, daily crude oil production, bottom hole pressure change curve, CO2 mole fraction at the production end, and water cut evolution. S42. Quantitative evaluation of fitting effect: The root mean square error and the coefficient of determination are used as quantitative evaluation indicators of the fitting process to quantitatively evaluate the fitting effect. S43. Determine if the fit meets the requirements; if the fit is not less than 85%, proceed to the next step; if the fit is not greater than 85%, return to step 3.

10. The method for accurately characterizing crack parameters and diffusion effects based on DFN according to claim 9, characterized in that, In step 5: S51. Conduct numerical simulation experiments on multiple sets of key fracture parameters to obtain a diffusion effect dataset; the diffusion effect dataset includes CO2 diffusion range, leading edge advance velocity, matrix swept volume, CO2 sequestration rate, and crude oil production increase. S52. Add a traditional dual-medium model as a control group to compare the diffusion effect prediction results of the DFN model and the traditional dual-medium model, including the diffusion range and diffusion extraction degree. S53. The relationship between crack parameters and diffusion effect is quantified through sensitivity analysis and correlation fitting methods.

Images