A method and system for integrating geological modeling and stress field simulation

Abstract

This invention discloses a method and system for integrating geological modeling and stress field simulation, belonging to the field of geological exploration and development technology. The method includes: project setup and data loading, seismic interpretation and attribute analysis, structural modeling, discrete fracture network modeling, and stress field simulation. This invention uses a three-dimensional structural model as a unified carrier, taking stratigraphic boundaries, fault distribution, and stratigraphic stratification as the basic constraints for mechanical modeling, achieving complete matching of the geological framework and the mechanical model's mesh and spatial scope. It employs a coupling method of fault distance constraints and hierarchical control using variogram functions to distinguish between main faults, secondary faults, and surrounding rock zones, setting differentiated interpolation rules, and using lithology and tectonic deformation zones as zoning constraints to complete the spatial simulation of mechanical parameters under geological understanding constraints. The structural model provides spatial boundaries for mechanical attribute modeling, and the mechanical parameter inversion results feed back to correct the geological model, forming a two-way data transfer and collaborative constraint mechanism.

Description

Technical Field

[0001] This invention belongs to the field of geological exploration and development technology, specifically relating to a method and system for integrating geological modeling and stress field simulation. Background Technology

[0002] In the field of geological exploration and development, accurately constructing subsurface geological structure models and simulating the distribution of in-situ stress fields are crucial steps in oil and gas reservoir evaluation, drilling design, and geological hazard assessment. In traditional workflows, geological modeling and stress field simulation are typically conducted independently, lacking a systematic and standardized data transfer and collaboration mechanism. This leads to problems such as spatial framework mismatch and inconsistent attribute parameters between the model and mechanical analysis. Although geological modeling software such as Petrel is powerful and widely used in structural interpretation and attribute modeling, its native workflow does not integrate a stress field simulation module for the complex mechanical behavior of fault zones. Users often need to rely on third-party software to complete mechanical calculations, resulting in cumbersome data conversion and potential for errors.

[0003] Furthermore, existing methods for spatial modeling rock mechanical parameters (such as Young's modulus and Poisson's ratio) often employ single-attribute interpolation or simple seismic constraints, neglecting the rapid changes and strong anisotropy of parameters within fault zones. This leads to smoothed distortion of mechanical properties near fault fracture zones, failing to accurately characterize their heterogeneity. Simultaneously, the lack of a reverse correction mechanism for geological models in geostress simulation results hinders the improvement of fault zone and reservoir sweet spots identification accuracy. Therefore, there is an urgent need to establish an integrated technical solution for geological modeling and stress field simulation, achieving bidirectional coupling between the structural framework and the mechanical model to enhance the geological rationality and engineering application value of the simulation results. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a method for integrating geological modeling and stress field simulation, comprising the following steps: Project setup and data loading steps: Create or open a work area, set the coordinate system and unit system, and load the wellhead file, well trajectory data and 3D seismic data volume; The earthquake interpretation and attribute analysis steps involve constructing and smoothing the earthquake data volume, calculating the variance volume and generating ant volumes to identify fault and fracture development zones. The construction modeling steps are as follows: define the model plane distribution range according to the geological boundary or seismic interpretation boundary; generate structural surfaces using well layer data and seismic interpretation layer data; add fault constraints; select corner grids or simple grids to construct a three-dimensional grid skeleton; coarse the well logging curves; establish a three-dimensional lithofacies model; and establish a three-dimensional distribution model of Poisson's ratio and Young's modulus under the constraints of the lithofacies model. The discrete fracture network modeling steps are as follows: load the fracture data interpreted from the well, statistically analyze the fracture orientation, generate discrete fracture pieces based on the random simulation of the three-dimensional fracture strength volume, and construct a discrete fracture network model. The stress field simulation steps involve calculating Young's modulus and Poisson's ratio based on array acoustic logging data and well logging data, creating discrete values, setting vertical and planar principal and secondary variability functions, performing rock physics modeling, and outputting stress field simulation results.

[0005] Furthermore, in the process of structural modeling and stress field simulation, a three-dimensional structural model is used as a unified carrier, and stratigraphic boundaries, fault distribution, and stratigraphic stratification are used as the basic constraints for mechanical modeling to achieve complete matching between the geological framework and the mechanical model grid and spatial range. A coupling method of fault distance constraint and variogram hierarchical control is adopted to distinguish between main faults, secondary faults and surrounding rock areas, and differentiated interpolation rules are set. Lithology and tectonic deformation zones are used as zoning constraints to achieve spatial simulation of mechanical parameters under the constraints of geological understanding.

[0006] Furthermore, the structural modeling step also includes: transferring the fracture geometry, stratigraphic framework, and grid cell data output by the structural model to the stress field simulation to provide spatial boundaries for the property modeling of Poisson's ratio, Young's modulus, and horizontal maximum principal stress; and feeding back the mechanical parameter inversion results to the geological model to correct the spatial identification range of fracture zones and reservoir sweet spots, thus forming a two-way coupling mechanism for data transfer from the structural model to the stress field simulation and for mechanical parameters to be fed back to the geological model.

[0007] Furthermore, the coupling method of fracture distance constraint and variogram hierarchical control specifically includes: for the main fracture zone, the main direction range is set to 200-500 meters; for the secondary fracture zone, the main direction range is set to 500-1500 meters; for the surrounding rock area, the main direction range is set to 500-1500 meters; and the secondary direction range is set to one-third to one-half of the main direction range.

[0008] Furthermore, in the discrete crack network modeling step, rose diagrams and stereographic projection diagrams are used to statistically analyze the crack orientation, identify the dominant crack direction, and then perform random simulation based on the three-dimensional crack strength volume to generate discrete crack segments.

[0009] Furthermore, in the stress field simulation step, the range of the vertical variation function is set to one or two formation thicknesses, or set to 20-50 meters according to the vertical step size of the grid; if the vertical change of the attribute is drastic, it is set to 10-30 meters.

[0010] Furthermore, it also includes a model verification step: selecting actual drilling locations within the work area that were not involved in modeling, extracting model simulation data and comparing it with measured in-situ stress and rock mechanics parameters, calculating the coefficient of determination, mean absolute error and root mean square error, and evaluating the reliability of the model.

[0011] This invention also proposes a system for integrating geological modeling and stress field simulation to implement the above-mentioned method, comprising: The data loading module is used to perform project settings, coordinate system and unit selection, and load wellhead files, well trajectory data and 3D seismic data volumes; The earthquake interpretation module is used to perform structural smoothing, variance calculation, and ant volume generation on earthquake data volumes, and to identify fault and fracture development zones. The structural modeling module is used to define the model range based on geological boundaries, generate structural surfaces and add fault constraints, construct a three-dimensional mesh, coarse well logging curves, and establish a three-dimensional lithofacies model and a three-dimensional distribution model of Poisson's ratio and Young's modulus. The discrete crack network modeling module is used to load crack data, statistically analyze the orientation, and generate a discrete crack network model based on random simulation of crack strength volume. The stress field simulation module is used to calculate mechanical parameters based on sonic logging data, set vertical and planar variation functions, perform rock physics modeling, and output the stress field. The coupling control unit is used to take the three-dimensional structural model as a unified carrier, and take the stratigraphic boundaries, fault distribution and stratigraphic stratification as the basic constraints for mechanical modeling, so as to achieve complete matching of the geological framework and mechanical model mesh and spatial range. It adopts the coupling method of fault distance constraint and variogram hierarchical control to distinguish the main fault, secondary fault and surrounding rock zone, set differentiated interpolation rules, and take lithology and tectonic deformation zone as zoning constraints to realize the spatial simulation of mechanical parameters.

[0012] Furthermore, the coupling control unit is also used to: transmit the fracture geometry, stratigraphic framework, and grid cell data output by the structural model to the stress field simulation module as spatial boundaries, and feed back the mechanical parameters inverted by the stress field simulation module to the structural modeling module to correct the spatial identification range of the fracture zone and the sweet spot.

[0013] Furthermore, it also includes a model validation module, which is used to select actual drilling locations that were not involved in the modeling, compare the simulated data with the measured data, calculate the coefficient of determination, mean absolute error and root mean square error, and output a validation report.

[0014] Compared with the prior art, the present invention has the following beneficial technical effects: First, this invention achieves complete matching between geological and mechanical models on a spatial grid. By using a three-dimensional structural model (strata, faults, and grid skeleton) as a unified carrier, and taking stratigraphic boundaries and fault distribution as basic constraints for mechanical modeling, it ensures that the mechanical simulation grid and the geological model are completely consistent in spatial range and element division, eliminating the geometric deviations caused by model conversion in traditional methods and improving the physical consistency of the simulation.

[0015] Secondly, this invention establishes a coupled method of fracture distance constraint and variogram hierarchical control, which can distinguish the differential interpolation rules between the main fracture, secondary fracture, and surrounding rock zone. This mechanism effectively solves the problems of parameter smoothing distortion and inaccurate anomaly characterization near the fracture zone in conventional attribute modeling, and significantly improves the spatial simulation accuracy of mechanical parameters such as Young's modulus and Poisson's ratio in the fracture zone.

[0016] Furthermore, this invention enables bidirectional data transfer and feedback correction between the geological model and mechanical parameters. The structural model provides spatial boundaries for modeling rock physical properties, while the inversion results of mechanical parameters can be fed back to the geological model to correct the spatial identification range of fracture zones and reservoir sweet spots, forming data feedback and continuously improving model accuracy.

[0017] Finally, through case studies, the parameters such as the maximum horizontal principal stress and Young's modulus obtained by the method of this invention are highly correlated with the measured values, with a high coefficient of determination. The mean absolute error and root mean square error are significantly lower than those of traditional methods, providing a more reliable scientific basis for oil and gas reservoir exploration and development and geological hazard assessment. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for integrating geological modeling and stress field simulation; Figure 2 A three-dimensional distribution model was constructed. Figure 3 The diagram shows the three-dimensional spatial distribution model of Poisson's ratio and Young's modulus, where (a) is the Poisson's ratio diagram and (b) is the Young's modulus diagram. Figure 4 The diagram shows a comparison of the spatial distribution of Young's modulus for the main fracture and the secondary fracture, where (a) represents the Young's modulus of the main fracture and (b) represents the Young's modulus of the secondary fracture. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] like Figure 1 The diagram shows a flowchart of a method for integrating geological modeling and stress field simulation. The specific method is as follows: First, perform project setup and data loading. After creating or opening a Petrel work area, correctly select the coordinate system and unit system in the project settings interface. Enable the automatic save function and customize the time interval to prevent accidental data loss. Next, load the wellhead file, which must include key information such as well name, wellhead coordinates, core fill elevation, and completed well depth, and import the well trajectory data. Then, load the SEGY format 3D seismic data volume, paying attention to the distinction between time-domain and depth-domain data. Typically, the seismic volume analyzed by Realise can be used for work, saving work area storage space and facilitating simultaneous display from multiple cameras. After completing data loading, seismic interpretation can be initiated based on the 3D data volume.

[0021] In the seismic interpretation and attribute analysis stage, the original seismic data volume is first structurally smoothed to suppress random noise and highlight structural features. Based on the smoothed seismic volume, a variance volume is calculated to effectively identify discontinuities in seismic phase axes and highlight fault and fracture development zones. The process proceeds step-by-step from seismic volume to structurally smoothed volume, then to variance volume generation, and finally to the ant-shaped volume.

[0022] Structural modeling is the foundation for subsequent attribute modeling and mechanical analysis. First, the planar distribution range of the model is defined based on the geological or seismic interpretation boundaries of the study area. Using well-layer data and seismically interpreted bedding plane data, fine structural surfaces are generated through interpolation algorithms such as convergence or kriging, with fault constraints incorporated during generation to ensure correct contact between faults and bedding planes. The mesh type must be selected based on the structural complexity: for complex blocks with well-developed faults, corner meshes are used, defining the mesh framework through faults and boundaries to achieve reasonable mesh deformation at faults; for relatively simple areas, simple meshes can be used, directly generating the mesh under the control of the top and bottom surfaces and boundaries, making the modeling process more convenient. During vertical meshing, larger geological units are subdivided into thinner mesh layers according to certain rules to meet the resolution requirements of subsequent mechanical modeling and numerical simulation. Then, high-resolution logging curves need to be coarsened. According to the vertical resolution of the 3D mesh, well-logging values ​​are converted to each mesh cell. Coarsening methods can include arithmetic mean or volumetric weighted average. Subsequently, sequential indicator simulation and other methods were used to establish a three-dimensional lithofacies model under the constraint of coarsened lithofacies from the wellbore. Seismic attribute volumes could be introduced as trend constraints to improve the model's accuracy. Under the constraints of the lithofacies model, further methods such as sequential Gaussian simulation or Kriging interpolation were used to establish three-dimensional distribution models of continuous attributes such as Poisson's ratio and Young's modulus. Figure 2 As shown.

[0023] After completing the structural modeling, discrete fracture network modeling is carried out. First, the fracture data interpreted from the wellbore is loaded and displayed, typically presented as a tadpole diagram on the well profile. Statistical analysis of the fracture orientation is performed using rose diagrams and stereographic projections. Then, in the fracture modeling module, random simulations are performed based on the three-dimensional fracture strength volume to generate discrete fracture patches, constructing a discrete fracture network reflecting the spatial distribution of fractures. Fracture statistics are shown in Table 1. Table 1 Crack statistics Based on the fracture network model, stress field simulation is further performed. Using array sonic logging data and well logging data from the drilling, Young's modulus and Poisson's ratio are calculated according to relevant formulas, and the results are imported into the work area. In the data preparation module, discrete values ​​of Young's modulus, Poisson's ratio, and minimum horizontal principal stress are created using the well logging curve coarsening function. Next, data analysis is conducted by opening the data analysis function and adjusting the variability function. First, the vertical variability function is set, with its range typically set to one or two formation thicknesses. This range can also be adjusted according to the actual needs of the work area. For example, when the vertical grid step size is 10 meters, the vertical range can be set to 20 to 50 meters; if the vertical changes in the attribute are drastic, it can be set to 10 to 30 meters. Next, primary and secondary directions are set: the primary direction is input with the azimuth angle according to the fracture strike, and the variation function of the primary direction is fitted. Because the fracture zone's properties change rapidly and the correlation range is small, the range can be set to 200 to 500 meters; the surrounding rock changes relatively slowly and the correlation range is larger, so the range is set to 500 to 1500 meters. The secondary direction range is generally set to one-third to one-half of the primary direction range, which reflects the geological characteristics of good continuity along the fracture direction and rapid changes perpendicular to the fracture direction. The next step is rock physics modeling. In the rock physics options of the attribute modeling module, the previously completed data analysis is checked, and the extracted attribute volume is fitted to further improve the model accuracy. To verify the model's reliability, actual drilling locations within the work area are selected, and the model simulation data is compared and statistically analyzed with measured in-situ stress and rock mechanics parameters. The results show that the simulated values ​​of the horizontal maximum principal stress, Young's modulus, and Poisson's ratio have good correlation with the measured values, with high coefficients of determination, and the overall mean absolute error and root mean square error are small. Compared with conventional attribute modeling methods, this study adopted a coupled modeling approach that combines fault and strata constraints. This approach effectively mitigated the smoothing effect of fault zone parameters, significantly reduced numerical deviations, and improved both model fit and spatial characterization accuracy. This demonstrates that the coupled geological and mechanical modeling method is reliable and yields accurate results.

[0024] This invention also proposes a system for integrating geological modeling and stress field simulation, comprising: The data loading module is used to perform project settings, coordinate system and unit selection, and load wellhead files, well trajectory data and 3D seismic data volumes. The wellhead coordinate data loaded by this module is acquired by total station and GPS satellite positioning equipment in field drilling engineering. The well trajectory data is obtained by drilling while drilling survey instrument and downhole gyroscope survey instrument. The 3D seismic data volume is generated by controlled source excitation and geophone array reception in field seismic exploration, and processed by a professional seismic processing system. It is the hardware source foundation for all modeling data in the work area.

[0025] The seismic interpretation module is used to perform structural smoothing, variance volume calculation, and ant volume generation on seismic data volumes, and to identify fault and fracture development zones. This module relies on the original 3D seismic data acquired by seismic exploration and acquisition hardware to perform secondary analysis. The data source is a high-precision seismic acquisition instrument array in the field. It captures the wave impedance response characteristics of underground strata, faults, and fractures through seismic wave acquisition equipment, providing raw hardware acquisition data support for the calculation of fault and fracture sensitive attributes and the identification of structural anomalies.

[0026] The structural modeling module is used to define the model range based on geological boundaries, generate structural surfaces and add fault constraints, construct a three-dimensional mesh, coarse well logging curves, and establish a three-dimensional lithofacies model and a three-dimensional distribution model of Poisson's ratio and Young's modulus. The well logging curve data used in this module are collected from downhole measurements using a complete set of well logging equipment such as downhole sonic logging tools, density logging tools, and gamma logging tools. Structural boundary and fault data are obtained by interpreting the results of seismic acquisition hardware, realizing joint modeling of measured downhole instrument data and surface seismic exploration equipment data.

[0027] The discrete fracture network modeling module is used to load fracture data, statistically analyze the occurrence, and generate a discrete fracture network model based on the random simulation of fracture strength. The fracture occurrence and fracture development density data used in the module come from the measured imaging data of the downhole imaging logging tool (FMI) and the core observation data obtained by the drilling core sampling equipment. The module accurately depicts the development characteristics of underground fractures through downhole high-precision imaging equipment and core acquisition equipment, providing measured hardware data support for the random simulation of fracture networks.

[0028] The stress field simulation module is used to calculate mechanical parameters based on sonic logging data, set vertical and planar variation functions, perform rock physics modeling, and output the stress field. The core basic data is continuously obtained from downhole sonic time-of-flight logging instruments and density logging instruments. At the same time, it combines measured calibration data from uniaxial and triaxial mechanical experiments on core samples conducted by an indoor rock mechanics testing machine to correct the geostress calculation parameters and ensure that the stress field simulation results are consistent with the actual formation mechanical characteristics.

[0029] The coupling control unit is used to take the three-dimensional structural model as a unified carrier, and take the stratigraphic boundaries, fault distribution and stratigraphic stratification as the basic constraints for mechanical modeling, so as to achieve complete matching of the geological framework and mechanical model mesh and spatial range. It adopts the coupling method of fault distance constraint and variogram hierarchical control to distinguish the main fault, secondary fault and surrounding rock zone, set differentiated interpolation rules, and take lithology and tectonic deformation zone as zoning constraints to realize the spatial simulation of mechanical parameters.

[0030] The coupling control unit is also used to transmit the fracture geometry, stratigraphic framework, and grid cell data output by the structural model to the stress field simulation module as spatial boundaries, and to feed back the mechanical parameters inverted by the stress field simulation module to the structural modeling module to correct the spatial identification range of fracture zones and reservoir sweet spots.

[0031] The model validation module is used to select actual drilling locations that were not involved in the modeling, compare the simulated data with the measured data, calculate the coefficient of determination, mean absolute error and root mean square error, and output a validation report.

[0032] Example 1 This embodiment provides a geomechanical coupled modeling method, the specific implementation process of which is as follows: First, project setup and data loading were performed. After creating a new Petrel work area, the required coordinate system and unit system were correctly selected in the project settings interface. To prevent data loss due to unexpected software shutdown, the work area's automatic save function was enabled, and the save interval was set to 15 minutes. During the data loading phase, the wellhead file containing information such as well name, wellhead coordinates, core fill elevation, and completed well depth, along with the corresponding well trajectory data, was imported first. Then, a 3D seismic data volume in SEGY format was loaded, with strict separation between time and depth domain data. To save work area space and support multi-camera synchronous display during subsequent interpretation, seismic volumes processed by Realise analysis were used. After completing the above basic data preparation, seismic interpretation work was initiated based on the loaded 3D data volume.

[0033] The process then moves to seismic interpretation and attribute analysis. To suppress random noise in the raw data and more clearly highlight stratigraphic features, a structural smoothing process was first applied to the original seismic data volume. Next, a variance volume was calculated based on the obtained smoothed seismic volume. Variance volume analysis effectively identifies discontinuities in seismic phase axes, thereby highlighting fault and fracture development zones. The entire process follows a step-by-step workflow from the original seismic volume, structural smoothing volume, variance volume, to the final generation of the ant-shaped data volume, providing a reliable attribute basis for subsequent fault modeling.

[0034] After completing the seismic attribute analysis, structural modeling was initiated. First, the planar distribution range of the model was defined based on the actual geological boundaries of the study area and the boundaries delineated by seismic interpretation results. Using existing well-layer data and seismic interpretation bedding data, a fine-grained structural bedding layer was generated using the Kriging interpolation algorithm. During the interpolation process, faults were incorporated as constraints to ensure accurate contact relationships between faults and each bedding layer. Considering the complex structural characteristics of fault development in the work area, corner grids were selected for meshing. Faults and boundaries were used to define the mesh skeleton, ensuring reasonable deformation of the mesh near faults and accurately depicting fault morphology. During vertical meshing, large geological units were proportionally subdivided into thinner mesh layers to meet the stringent high-resolution requirements of subsequent mechanical modeling and numerical simulation. After mesh establishment, the high-resolution well logging curves were coarsened. Well logging values ​​near the wells were converted to each mesh cell according to the vertical resolution of the 3D mesh; the coarsening method used was the arithmetic mean method. After obtaining the coarsened lithofacies data, a three-dimensional lithofacies model constrained by the wellhead coarsened lithofacies was established using a sequential indicator simulation method. During this process, to further improve the accuracy of the model's lateral predictions, the previously generated seismic attribute volume was introduced as a trend constraint. Finally, under the constraints of the established lithofacies model framework, a three-dimensional spatial distribution model of two continuous attributes, Poisson's ratio and Young's modulus, was established using a sequential Gaussian simulation method, as shown below. Figure 3 and Figure 4 As shown, Figure 3 (a) in the diagram is the Poisson's ratio diagram. Figure 3 (b) is the Young's modulus diagram, which shows vertical stratification. As the depth increases, both Poisson's ratio and Young's modulus gradually increase. Poisson's ratio and Young's modulus show linear changes along the fault. Low Young's modulus and low Poisson's ratio indicate rock fracture zones and stress release areas, which are distributed in narrow bands and are segmented. Figure 4 In the figure, (a) represents the Young's modulus of the main fracture. Figure 4 (b) in the figure represents the Young's modulus of the secondary fracture. The figure shows that the strata in each area have obvious stratification, which confirms... Figure 3 The accuracy of the hierarchical model.

[0035] Next, discrete fracture network modeling is performed. First, fracture data obtained from wellbore interpretation is loaded into the work area, and fracture development is visually displayed on the well trajectory profile in the form of a tadpole diagram. Using this data, rose diagrams and stereographic projections are drawn to statistically analyze the main fracture development and orientation, clarifying the dominant fracture directions. Based on this, the fracture modeling module is entered, and random simulation calculations are performed based on the calculated 3D fracture strength volume to generate a series of discrete fracture slices. Finally, a discrete fracture network model reflecting the actual spatial distribution characteristics of fractures in the work area is constructed.

[0036] At this point, a complete three-dimensional attribute volume and fracture model have been obtained. Stress field simulation will then be performed to obtain geomechanical parameters. Based on array sonic logging data and well logging data provided by drilling in this work area, Young's modulus and Poisson's ratio for each well were calculated according to relevant petrophysical formulas, and these mechanical parameter curves were imported into the work area. In the well logging curve coarsening function of the data preparation module, discrete values ​​of Young's modulus, Poisson's ratio, and minimum horizontal principal stress along the well trajectory were created sequentially. Next, the data analysis and variogram adjustment steps were performed. In the data analysis interface, the vertical variogram function was first set. Given that the vertical step size of the grid in the work area was set to 10 meters, and considering the formation thickness, the vertical range was set to 30 meters. Then, the principal and secondary directional variogram functions on the plane were set. The azimuth angle of the principal direction was input as the fracture strike angle. After fitting, it was found that the fracture zone, due to its drastic attribute changes and small correlation range, had its principal direction range set to 300 meters; while the surrounding rock area had relatively gentle attribute changes and a larger correlation range, so the principal direction range was set to 800 meters. Finally, the secondary directional range is fitted and set to one-third of the primary directional range to reflect the anisotropy characteristics of good continuity along the fracture strike and rapid change in direction perpendicular to the fracture strike. After the variogram is adjusted, rock physics modeling is performed in the attribute modeling module. In the rock physics options, the pre-set variogram parameters are selected, and the extracted attribute volume data is used in the fitting calculation to enhance the accuracy of the model.

[0037] To verify the reliability of the method in this embodiment, accuracy and error statistics were performed. Actual drilling locations within the work area that were not included in the calculations were selected, and the simulated data at these well points were compared and analyzed with measured in-situ stress and rock mechanics parameters. Statistical results show that the simulated and measured values ​​of the maximum horizontal principal stress, Young's modulus, and Poisson's ratio exhibit a good correlation with each other, with high coefficients of determination. Simultaneously, the overall mean absolute error and root mean square error of each parameter are relatively small. In particular, compared to conventional attribute modeling methods, this embodiment, by employing a coupled modeling approach that combines fracture and strata constraints, effectively mitigates the parameter smoothing effect commonly found near fault zones, significantly reducing numerical deviations. Both the spatial fit of the model and the accuracy of characterizing heterogeneous bodies have been significantly improved, thus fully demonstrating that this geologically and mechanically coupled modeling method is stable and reliable, and the results obtained are accurate.

[0038] Example 2 Taking the Fuman work area in the Tarim Basin as an example, this embodiment provides a geological modeling method suitable for complex fault-controlled fractured caves. The specific implementation process is as follows.

[0039] First, project setup and data loading are performed. After creating a new work area, the correct coordinate system and unit system are selected for the Fuman work area in the project settings interface, and the automatic save function is enabled, with the save interval set to 30 minutes to prevent accidental data loss. Then, the wellhead files of all wells in the study area are loaded, including well name, wellhead coordinates, core filling elevation, and completed well depth, and the well trajectory data of each well are imported simultaneously. Based on this, a SEGY format 3D seismic data volume covering the work area is loaded, strictly distinguishing between time domain and depth domain data during loading. To optimize storage and support subsequent multi-camera collaborative display, this embodiment uses a seismic volume processed by Realise analysis. After data loading is completed, seismic interpretation work is started based on this 3D data volume.

[0040] The process then moves to seismic interpretation and attribute analysis. To address the random noise present in the original seismic data, structural smoothing is first applied to effectively suppress noise interference and clarify fault and stratigraphic features. After obtaining the smoothed seismic volume, its variance attribute volume is calculated. By detecting discontinuities in seismic phase axes, strike-slip fault zones and fracture development areas are highlighted. Subsequently, this embodiment follows a technical workflow from the original seismic volume to the structural smoothed volume, then to the variance volume, and finally to the extraction of ant-shaped data. Using the generated ant-shaped data attributes, combined with well-crossing seismic profiles, a detailed manual interpretation and combination of strike-slip faults and their associated fracture development zones within the work area is performed, providing an accurate three-dimensional fault model for structural modeling.

[0041] After obtaining detailed fault interpretation results, structural modeling was initiated. First, the planar distribution range of the model was defined based on the geological and seismic interpretation boundaries of the strike-slip fault zone in the Fuman work area. Using geological stratification data from each well and seismic interpretation bedding data, the Kriging interpolation algorithm was selected to generate key structural bedding planes. During the interpolation process, the previously interpreted fault model was incorporated as a constraint to ensure the rationality of the contact relationship between the fault and the bedding plane. Given the development of strike-slip faults and the complexity of the structure in the work area, this embodiment selected a corner grid as the three-dimensional grid framework. The grid skeleton was defined using faults and boundaries, allowing the grid to deform reasonably at fault locations, accurately reflecting the fault displacement and morphology. Vertically, the main target segment was subdivided into multiple thin grid layers according to a proportional or fixed thickness division method, achieving a vertical resolution of meters to meet the requirements of subsequent attribute modeling and mechanical analysis. After the grid model was established, the high-resolution logging curves were coarsened. The lithology and physical property curve values ​​at the well points were coarsened to each grid cell using an arithmetic mean method according to the vertical resolution of the three-dimensional grid. Next, a three-dimensional lithofacies model was established under the control of the coarsened lithofacies from the wellbore, using a sequential indicator simulation method. To enhance the accuracy of lateral prediction, the extracted seismic attribute volume was also introduced as a trend volume into the simulation process. Finally, within the framework constraints of the lithofacies model, a three-dimensional spatial distribution model of continuous attributes such as Poisson's ratio and Young's modulus was established using the sequential Gaussian simulation method, providing an attribute volume basis for subsequent analysis.

[0042] After completing the aforementioned attribute modeling, the process transitions to discrete fracture network modeling. First, imaging logging fracture interpretation data from a single well is loaded, and fracture development is displayed along the well trajectory profile in a tadpole diagram format. Rose diagrams and stereographic projections are generated from the fracture data, and the fracture orientation within the work area is statistically analyzed, clarifying the dominant fracture strike and dip angle. After clarifying the fracture development patterns, the fracture modeling module is entered. Based on the calculated 3D fracture strength volume, stochastic simulations are performed to discretize and generate a large number of fracture fragments conforming to statistical laws, thereby constructing a discrete fracture network model reflecting the spatial network distribution characteristics of fractures in the work area. This model provides an accurate fracture framework for subsequent fluid flow and geomechanical analysis.

[0043] After establishing the discrete fracture network model, stress field simulation was carried out. First, using array sonic logging data and well logging data from the drilling, the dynamic Young's modulus and dynamic Poisson's ratio were calculated according to rock physics formulas, and these mechanical parameter curves were imported into the work area.

[0044] In the formula, Δtp is the longitudinal wave time difference, in μ units / ft; Δt sρ is the transverse wave transit time, in μs / ft; ρ is the rock bulk density, in g / cm³. 3 Next, in the data preparation module, discrete values ​​for Young's modulus, Poisson's ratio, and minimum horizontal principal stress are created sequentially in the logging curve coarsening function to complete the gridding of parameters at the well point. Based on this, data analysis and variogram adjustment are performed. First, the vertical variogram is set, with its range generally taken as one or two formation thicknesses, depending on the actual conditions of the work area. For example, when the vertical grid step size is 10 meters, the vertical range can be set to 20 to 50 meters. If the vertical changes in attributes are drastic, it can be further reduced to 10 to 30 meters. Subsequently, the principal and secondary directions of the plane variogram were determined. The principal direction was input with the corresponding azimuth angle according to the fracture strike, and the variogram of the principal direction was fitted. Within the fracture zone, the properties change rapidly and the correlation range is small, so the range was set to 200 to 500 meters. In the surrounding rock area, the properties change slowly and the correlation range is large, so the range was set to 500 to 1500 meters. The range of the secondary direction was generally set to one-third to one-half of the range of the principal direction, thus reflecting the anisotropy characteristics of good continuity along the fracture strike and rapid changes perpendicular to the fracture strike. After completing the data analysis, the rock physics modeling stage of the attribute modeling module was entered. The previously set variogram analysis results were selected, and the extracted attribute volume was used in the fitting calculation to further improve the model accuracy. To verify the reliability of the model, actual drilling locations within the work area were selected, and the model simulation data was compared with the measured in-situ stress and rock mechanics parameters. The results showed that the simulated values ​​of the horizontal maximum principal stress, Young's modulus, and Poisson's ratio had good correlation with the measured values, with high coefficients of determination, and the overall mean absolute error and root mean square error were both small. Compared with conventional attribute modeling, the coupled modeling method using joint constraints of faults and strata effectively mitigates the smoothing effect of fault zone parameters, significantly reduces numerical deviation, and significantly improves model fit and spatial characterization accuracy. This proves that the geological and mechanical coupled modeling method is reliable and the results are accurate.

[0045] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0046] The above embodiments are merely illustrative examples and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for integrating geological modeling and stress field simulation, characterized in that, Includes the following steps: Project setup and data loading steps: Create or open a work area, set the coordinate system and unit system, and load the wellhead file, well trajectory data and 3D seismic data volume; The earthquake interpretation and attribute analysis steps involve constructing and smoothing the earthquake data volume, calculating the variance volume and generating ant volumes to identify fault and fracture development zones. The construction modeling steps are as follows: define the model plane distribution range according to the geological boundary or seismic interpretation boundary; generate structural surfaces using well layer data and seismic interpretation layer data; add fault constraints; select corner grids or simple grids to construct a three-dimensional grid skeleton; coarse the well logging curves; establish a three-dimensional lithofacies model; and establish a three-dimensional distribution model of Poisson's ratio and Young's modulus under the constraints of the lithofacies model. The structural modeling step also includes: transferring the fracture geometry, stratigraphic framework, and grid cell data output by the structural model to the stress field simulation to provide spatial boundaries for the property modeling of Poisson's ratio, Young's modulus, and horizontal maximum principal stress; and feeding back the mechanical parameter inversion results to the geological model to correct the spatial identification range of fracture zones and reservoir sweet spots, forming a two-way coupling mechanism for data transfer from the structural model to the stress field simulation and feedback of mechanical parameters to the geological model. The discrete fracture network modeling steps are as follows: load the fracture data interpreted from the well, statistically analyze the fracture orientation, generate discrete fracture pieces based on the random simulation of the three-dimensional fracture strength volume, and construct a discrete fracture network model. The stress field simulation steps involve calculating Young's modulus and Poisson's ratio based on array acoustic logging data and well logging data, creating discrete values, setting vertical and planar principal and secondary variability functions, performing rock physics modeling, and outputting stress field simulation results. In the process of structural modeling and stress field simulation, a three-dimensional structural model is used as a unified carrier. Stratigraphic boundaries, fault distribution, and stratigraphic stratification are used as the basic constraints for mechanical modeling to achieve a complete match between the geological framework and the mechanical model mesh and spatial range. Furthermore, a coupling method of fault distance constraint and variogram hierarchical control is adopted to distinguish between main faults, secondary faults, and surrounding rock areas. Differentiated interpolation rules are set, and lithology and tectonic deformation zones are used as zoning constraints to achieve spatial simulation of mechanical parameters under the constraints of geological understanding. The coupling method of fracture distance constraint and variogram hierarchical control specifically includes: for the main fracture zone, the main direction range is set to 200-500 meters; for the secondary fracture zone, the main direction range is set to 500-1500 meters; for the surrounding rock area, the main direction range is set to 500-1500 meters; and the secondary direction range is set to one-third to one-half of the main direction range.

2. The method according to claim 1, characterized in that, In the discrete crack network modeling step, rose diagrams and stereographic projection diagrams are used to statistically analyze the crack orientation, identify the dominant crack direction, and then random simulation is performed based on the three-dimensional crack strength volume to generate discrete crack segments.

3. The method according to claim 1, characterized in that, In the stress field simulation step, the range of the vertical variation function is set to one or two formation thicknesses, or to 20-50 meters according to the vertical step size of the grid; if the vertical change of the attribute is drastic, it is set to 10-30 meters.

4. The method according to claim 1, characterized in that, It also includes a model validation step: selecting actual drilling locations within the work area that were not involved in the modeling, extracting model simulation data and comparing it with measured ground stress and rock mechanics parameters, calculating the coefficient of determination, mean absolute error and root mean square error, and evaluating the reliability of the model.

5. A system for integrating geological modeling and stress field simulation, characterized in that, For implementing the method as described in any one of claims 1-4, comprising: The data loading module is used to perform project settings, coordinate system and unit selection, and load wellhead files, well trajectory data and 3D seismic data volumes; The earthquake interpretation module is used to perform structural smoothing, variance calculation, and ant volume generation on earthquake data volumes, and to identify fault and fracture development zones. The structural modeling module is used to define the model range based on geological boundaries, generate structural surfaces and add fault constraints, construct a three-dimensional mesh, coarse well logging curves, and establish a three-dimensional lithofacies model and a three-dimensional distribution model of Poisson's ratio and Young's modulus. The discrete crack network modeling module is used to load crack data, statistically analyze the orientation, and generate a discrete crack network model based on random simulation of crack strength volume. The stress field simulation module is used to calculate mechanical parameters based on sonic logging data, set vertical and planar variation functions, perform rock physics modeling, and output the stress field. The coupling control unit is used to take the three-dimensional structural model as a unified carrier, and take the stratigraphic boundaries, fault distribution and stratigraphic stratification as the basic constraints for mechanical modeling, so as to achieve complete matching of the geological framework and mechanical model mesh and spatial range. It adopts the coupling method of fault distance constraint and variogram hierarchical control to distinguish the main fault, secondary fault and surrounding rock zone, set differentiated interpolation rules, and take lithology and tectonic deformation zone as zoning constraints to realize the spatial simulation of mechanical parameters.

6. The system according to claim 5, characterized in that, The coupling control unit is also used to: transmit the fracture geometry, stratigraphic framework, and grid cell data output by the structural model to the stress field simulation module as spatial boundaries, and feed back the mechanical parameters inverted by the stress field simulation module to the structural modeling module to correct the spatial identification range of fracture zones and reservoir sweet spots.

7. The system according to claim 5, characterized in that, It also includes a model validation module, which is used to select actual drilling locations that were not involved in the modeling, compare the simulated data with the measured data, calculate the coefficient of determination, mean absolute error and root mean square error, and output a validation report.

Images