Three-dimensional geological modeling method and system for gas reservoir of gas storage

Abstract

The invention discloses a three-dimensional geological modeling method and system for a gas reservoir of a gas storage. The method comprises the steps of modeling data preprocessing, wherein layering and well track errors are analyzed and corrected, a denudation area range is represented, stratigraphic structure data are processed, logging data and experimental data are processed and analyzed, model and constraint body data are determined, and a database is formed; building a construction model: building the construction model based on a plane gridding method, and carrying out local construction adjustment and verification; establishing an attribute model: simulating the physical property parameters based on the sedimentary microfacies type of the position of the physical property parameter point and forming the attribute model; and reserve calculation and uncertainty analysis: calculating the reserve of the gas reservoir based on the construction model and the attribute model, and performing cross check by using the average reservoir thickness and the attribute value of the model input data in comparison with the calculation value based on the planar graph. According to the method, a plurality of modeling methods are adopted for interactively processing the contact relation between the fault and the stratum, and the problems of high and steep recoil structures, thin reservoir denudation area characterization and the like can be effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of three-dimensional geological modeling, and in particular to a three-dimensional geological modeling method and system for a gas storage reservoir. Background Art

[0002] Three-dimensional geological modeling of gas reservoirs is both the best way to reflect the results of geological research and a key technical means for reservoir prediction. The accuracy of geological model construction is directly affected by the extent of previous geological research; the credibility of reservoir prediction is influenced both by the extent of previous research and by the modeling method. Factors such as reservoir characteristics and different types of oil and gas reservoirs lead to different modeling methods. The key to geological modeling is the effective combination of feasible modeling ideas and appropriate modeling techniques. Modeling techniques are comprehensive, combining multiple disciplines such as mathematics, geology, and statistics. Finding the optimal modeling technique suitable for the target reservoir based on traditional stochastic modeling techniques is a key research topic in three-dimensional reservoir geological modeling.

[0003] Currently used modeling methods fall into two main categories: deterministic modeling and stochastic modeling. Deterministic reservoir modeling provides deterministic predictions for unknown areas between wells. This means inferring deterministic, unique, and true reservoir parameters between wells from deterministic data. Stochastic modeling, based on known information and using random functions as its theory, applies stochastic simulation methods to generate multiple alternative, equally probabilistic reservoir models. Both approaches can be used in both microfacies simulation and physical property simulation.

[0004] Deterministic modeling starts from control points (such as well points) with deterministic data and infers the unique and definite reservoir parameters between points (such as between wells). The main deterministic modeling methods include reservoir seismology, reservoir sedimentology, and geostatistical kriging, all of which can be used alone or in combination. Deterministic modeling is a unique and definite representation of reservoir parameters between wells. However, underground reservoirs are complex. They are the result of the combined action of many complex geological processes, with complex spatial configurations of reservoir structures and spatial variations in reservoir parameters. It is relatively difficult to describe the distribution of underground reservoirs with very limited data, and there is bound to be uncertainty and randomness. Therefore, reservoir stochastic modeling technology came into being and is being used more and more widely. Summary of the Invention

[0005] To address the above-mentioned issues, the present invention proposes a three-dimensional geological modeling method and system for gas storage reservoirs. Multiple modeling methods are used to interactively process the contact relationships between faults and strata, effectively resolving difficult problems such as the characterization of steep thrust structures and thin reservoir denudation zones. Variation function analysis is performed using the attribute distribution of step-by-step well calibration, and random attribute simulation is performed with the aid of plane constraints, thereby improving the prediction accuracy of the attribute model.

[0006] The technical solution adopted in the present invention is as follows:

[0007] A three-dimensional geological modeling method for a gas storage reservoir, comprising:

[0008] Modeling data preprocessing: Analyze and correct stratification and well trajectory errors, characterize the extent of the denudation zone and process stratigraphic structural data, process and analyze logging data and experimental data, determine the model and constraint volume data, and form a database including geological recognition data, stratification data, drilling and logging interpretation and analysis data, oil and gas testing data, and seismic interpretation data;

[0009] Structural model establishment: Establishing a structural model based on a plane gridding method and performing local structural adjustments and verification. The structural model includes a fault model and a layer model.

[0010] Establishing a property model: Based on the sedimentary microfacies type at the location of the physical property parameter point, the physical property parameters are simulated and a property model is formed. The physical property parameters include porosity, permeability, water saturation and net-to-gross ratio;

[0011] Reserve calculation and uncertainty analysis: Calculate the reserves of the gas reservoir based on the structural and property models, using average reservoir thickness and property values ​​from the model input data to cross-check against the calculated values ​​based on the plan view; compare the volumes in the structural and property models with previous estimates from various sources to complete the uncertainty analysis of the reserves.

[0012] Furthermore, the modeling data preprocessing includes:

[0013] Layer and well trajectory error analysis and correction: In case of conflicting layer data in different files, the gas wells in the entire area are re-layered in combination with basic geological data of gas wells, lithologic combination characteristics and perforation production performance to provide layer data for the establishment of structural models;

[0014] Characterization of the erosion zone and stratigraphic structural data processing: Based on the original interpretation data and using well point data as hard data, the interpreted horizon data is corrected. The corrected data is then inspected and cleaned near the fault. Combined with previous research data, the scope of the erosion zone is characterized and stratigraphic structural data is processed. Based on the existing stratigraphic thickness map and with the drilled stratigraphic thickness as a constraint, a stratigraphic thickness map is generated in combination with the impact of the erosion zone.

[0015] Processing and analysis of well logging and experimental data: Based on the porosity plan map, the original map is initially corrected using well logging interpretation data. In combination with the distribution of denuded areas, the porosity data distribution characteristics are extracted and a corrected porosity distribution map is produced. Based on the processed data and map, the correlation between different parameters is explored to provide trends and parameter constraints for the establishment of attribute models.

[0016] Model and constraint body data selection: Extract the original attribute body plane attribute distribution trend constraints to establish a new attribute model.

[0017] Furthermore, the construction model establishment includes:

[0018] Fault model establishment: Corner point grid modeling and complex structure modeling are used to interactively process fault contact relationships and establish a fault model that characterizes the spatial distribution of faults;

[0019] Layer model establishment: Utilizing single-well layer data combined with previous structural interpretation results, an inter-well interpolation algorithm is used to form a data field between previously unrelated single-well layer data points. This data field drives the generation of surfaces to form a layer model.

[0020] Grid system establishment: The area and volume of the study area are simulated and projected into the 3D modeling software and the resolution is set to form a grid system;

[0021] Structural model verification: local structural adjustments and verification are performed based on the well trajectory and geological stratification. The latest perforation data is loaded in combination with the numerical model to perform local structural adjustments to ensure that the stratification is completely consistent with the structural surface.

[0022] Furthermore, during the establishment of the grid system, when constructing a simple research block model with undeveloped faults, trend lines are rarely used or not used; when constructing a complex research block model with developed faults, fault lines are used as trend lines or trend lines parallel / perpendicular to the fault lines are drawn to constrain the generation of the grid.

[0023] Furthermore, the attribute model establishment includes:

[0024] Curve discretization and data correction: Discretize the lithofacies classification curve, porosity, permeability, water saturation, and net-to-gross ratio curves; based on the physical property parameter plane map, use the well logging interpretation data to make a preliminary correction to the original map, and combine the distribution characteristics of the denudation zone to extract the distribution characteristics of the physical parameters and generate the corrected physical property parameter distribution map;

[0025] Variogram analysis: Based on the normally distributed data and the variogram extracted from the calibrated porosity plot, a variogram analysis is performed on the discretized porosity well data to determine relevant parameters including the major and minor ranges.

[0026] Establishment of lithofacies model: Based on the discretized data, variogram analysis is carried out on different lithofacies. Based on the time-sequential indicator simulation method, a lithofacies model is established using a stochastic simulation method.

[0027] Porosity model establishment: Using the normally distributed porosity data and variogram analysis results, a reservoir porosity model is established based on the sequential Gaussian random simulation method and the porosity plane distribution as a trend constraint;

[0028] Permeability model establishment: The well test permeability interpreted from existing wells is combined with the permeability interpreted from well logging. The well test permeability curve for each well is calculated based on the obtained relationship, and the well test permeability model is established using the sequential Gaussian random simulation method.

[0029] Water saturation model establishment: Based on the analysis of gas saturation data interpreted from well logging, combined with production performance and test data, referring to structural morphological characteristics, using the porosity model as a control, and the gas saturation plane distribution map as a trend constraint, the water saturation model is established using the sequential Gaussian simulation method;

[0030] Net-to-gross ratio model establishment: Based on the porosity model and gas saturation model, the net-to-gross ratio model is established in combination with production operation dynamics and calculation thresholds.

[0031] A three-dimensional geological modeling system for a gas storage reservoir, comprising:

[0032] The modeling data preprocessing module is configured to analyze and correct layer and well trajectory errors, characterize the extent of the erosion zone and process stratigraphic structural data, process and analyze well logging data and experimental data, determine the model and constraint volume data, and form a database including geological recognition data, layer data, drilling and logging interpretation and analysis data, oil and gas testing data, and seismic interpretation data;

[0033] a structural model building module configured to build a structural model based on a plane gridding method and perform local structural adjustment and verification, wherein the structural model includes a fault model and a layer model;

[0034] a property model building module configured to simulate physical property parameters based on the sedimentary microfacies type at the location of the physical property parameter point and form a property model, wherein the physical property parameters include porosity, permeability, water saturation and net-to-gross ratio;

[0035] The reserve calculation and uncertainty analysis module is configured to calculate the reserves of the gas reservoir based on the structural model and the attribute model, cross-check the calculated values ​​based on the plan view using the average reservoir thickness and the attribute values ​​of the model input data, and compare the volumes in the structural model and the attribute model with previous estimates from various sources to complete the uncertainty analysis of the reserves.

[0036] Furthermore, the modeling data preprocessing module includes:

[0037] The layer and well trajectory error analysis and correction unit is configured to address conflicts in layer data from different files. It combines basic geological data of gas wells, lithologic combination characteristics, and perforation production performance to re-stratify gas wells in the entire area and provide layer data for the establishment of the structural model.

[0038] The unit for characterizing the extent of the erosion zone and processing stratigraphic structure data is configured to use the original interpretation data as a basis and the well point data as hard data to correct the interpreted horizon data, check and clean the data near the fault on the corrected data, and characterize the extent of the erosion zone and process the stratigraphic structure data in combination with the previous research data; based on the existing stratigraphic thickness map and with the drilling stratigraphic thickness as a constraint, the stratigraphic thickness map is generated in combination with the influence of the erosion zone;

[0039] The logging and experimental data processing and analysis unit is configured to use the porosity plan map as a basis, utilize the logging interpretation data to perform preliminary corrections on the original map, combine the distribution of the denuded area, extract the distribution characteristics of the porosity data, and produce a corrected porosity distribution map. Based on the processed data and map, it explores the correlation between different parameters and provides trends and parameter constraints for the establishment of attribute models.

[0040] The model and constraint volume data selection unit is configured to extract the original attribute volume plane attribute distribution trend constraint to establish a new attribute model.

[0041] Furthermore, the construction model building module includes:

[0042] A fault model building unit is configured to interactively process fault contact relationships using corner point grid modeling and complex structure modeling to build a fault model that characterizes the spatial distribution of the fault;

[0043] The layer model building unit is configured to use the single-well layer data combined with the previous structural interpretation results, and use the inter-well interpolation algorithm to form a data field between the originally unrelated single-well layer data points. Under the influence of this data field, a surface is generated to form a layer model.

[0044] A grid system establishing unit is configured to simulate and project the area and volume size of the study block onto a three-dimensional modeling software and set the resolution, thereby forming a grid system;

[0045] The structural model verification unit is configured to perform local structural adjustments and verification based on the well trajectory and geological stratification, and to load the latest perforation data in combination with the numerical model to perform local structural adjustments so that the stratification is completely consistent with the structural surface.

[0046] Furthermore, in the grid system establishment unit, when constructing a simple research block model with undeveloped faults, trend lines are rarely used or not used; when constructing a complex research block model with developed faults, fault lines are used as trend lines or trend lines parallel / perpendicular to the fault lines are drawn to constrain the generation of the grid.

[0047] Furthermore, the attribute model building module includes:

[0048] The curve discretization and data correction unit is configured to discretize the lithofacies classification curve, porosity, permeability, water saturation, and net-to-gross ratio curves. Based on the physical property parameter plane map, the original map is preliminarily corrected using well logging interpretation data. In combination with the distribution characteristics of the denudation zone, the distribution characteristics of the physical parameters are extracted and a corrected physical property parameter distribution map is generated.

[0049] a variogram analysis unit configured to perform variogram analysis on the discretized porosity well data based on the normally distributed data and the variogram extracted according to the calibrated porosity plane map, and determine relevant parameters including a major range and a minor range;

[0050] The lithofacies model building unit is configured to conduct variogram analysis on different lithofacies based on discretized data, and to build a lithofacies model using a stochastic simulation method based on a time-sequential indicator simulation method;

[0051] The porosity model building unit is configured to call the normally distributed porosity data and the variogram analysis results, and build a reservoir porosity model based on the sequential Gaussian random simulation method and the porosity plane distribution as a trend constraint;

[0052] a permeability model building unit configured to perform intersection processing using the well test permeability interpreted by the existing wells and the permeability interpreted by the well logging, calculate the well test permeability curve of each well based on the obtained relationship, and establish the well test permeability model using a sequential Gaussian random simulation method;

[0053] The water saturation model building unit is configured to build a water saturation model based on the analysis of gas saturation data interpreted from well logging, combined with production performance and test data, with reference to structural morphological characteristics, using the porosity model as a control, and the gas saturation plane distribution map as a trend constraint, using the sequential Gaussian simulation method;

[0054] The net-to-gross ratio model establishing unit is configured to establish a net-to-gross ratio model based on the porosity model and the gas saturation model, combined with production operation dynamics and calculation thresholds.

[0055] The beneficial effects of the present invention are:

[0056] (1) The present invention provides a method for organizing and quality controlling drilling, seismic, geological and injection-production dynamic data, providing a reliable data source for geological models.

[0057] (2) Limited by the grid of the later numerical simulation and geomechanical model, the present invention adopts multiple modeling methods to interactively process the contact relationship between faults and strata, which can effectively solve the difficult problems of high-steep thrust structures and thin reservoir erosion zone characterization.

[0058] (3) The present invention uses the attribute distribution of step-by-step well calibration to perform variogram analysis and uses plane constraints to carry out attribute random simulation, which can improve the prediction accuracy of the attribute model.

[0059] (4) The present invention adopts a combination of dynamic and static methods to continuously correct model parameters to make the geological model more realistic. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 This is a flow chart of the three-dimensional geological modeling method for gas storage reservoirs according to Example 1 of the present invention.

[0061] Figure 2 This is a flow chart for improving the geological model of a gas storage reservoir according to Example 1 of the present invention.

[0062] Figure 3 It is a schematic diagram of selecting the structural trend reference surface when constructing the layer model. DETAILED DESCRIPTION

[0063] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, the specific embodiments of the present invention are now described. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention. That is, the embodiments described are only part of the embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present invention.

[0064] Example 1

[0065] like Figure 1As shown, this embodiment provides a three-dimensional geological modeling method for gas storage reservoirs, including the steps of modeling data preprocessing, structural model establishment, attribute model establishment, and reserve calculation and uncertainty analysis. The modeling data preprocessing includes: analyzing and correcting stratification and well trajectory errors, characterizing the scope of the erosion area and processing stratigraphic structural data, processing and analyzing logging data and experimental data, determining the model and constraint body data, and forming a database including geological recognition data, stratification data, drilling and logging interpretation and analysis data, oil and gas test data, and seismic interpretation data; the structural model establishment includes: establishing a structural model based on a plane gridding method and performing local structural analysis. Adjustment and verification, where the structural model includes a fault model and a bedding model; attribute model establishment includes: simulating the physical parameters based on the sedimentary microfacies type at the location of the physical parameter points and forming an attribute model, where the physical parameters include porosity, permeability, water saturation and net-to-gross ratio; reserve calculation and uncertainty analysis include: calculating the reserves of the gas reservoir based on the structural model and attribute model, using the average reservoir thickness and attribute values ​​of the model input data, and cross-checking against the calculated values ​​based on the plan view; comparing the volumes in the structural model and attribute model with previous estimates from various sources to complete the uncertainty analysis of the reserves.

[0066] Specifically, the three-dimensional geological modeling method of the gas storage reservoir is now described using the Carboniferous gas reservoir of the Xiangguo Temple Gas Storage in Chongqing as an example. This embodiment is based on comprehensive geological research, and a phase model is established by stochastic modeling. A porosity model is built based on phase control, and an initial water saturation field is built using the J function (the Xiangguo Temple Gas Storage reservoir is considered to be a two-phase gas-water gas reservoir, and the gas saturation field is thus obtained). The reserves are calculated and verified using a static method. The Carboniferous carbonate reservoir classification scheme of the Sichuan Basin is applied in attribute modeling, taking into account porosity, permeability, pore throat size, and pore and pore throat geometry. For porosity and other continuous attribute modeling, sequential Gaussian simulation is used as the modeling algorithm.

[0067] To perform reservoir modeling, two distinct permeability attributes were generated: matrix permeability derived from core experiments and total "effective" permeability interpreted from well tests. An empirical formula for calculating matrix permeability was established based on existing RCA / core test data. Wellpoint permeability interpolation was performed first, and the permeability field model was refined based on insights from seismic data, well tests, and well production performance. This completed the geological modeling process.

[0068] 1. Modeling Data Composition and Quality Control

[0069] Guided by multidisciplinary integration, we fully utilize data such as seismic, drilling, logging, well testing and experimental analysis to conduct comprehensive geological research, use interpreted data to carry out high-precision three-dimensional geological modeling of the Xiangguo Temple Carboniferous gas reservoir, and combine well and seismic data with the previously imported data, check for singular values ​​of structural surfaces, and inspect fault lines or sections. We remove outliers and process invalid and null values ​​to improve data quality. For new drilling data, we normalize it with the previously loaded data to improve data standardization.

[0070] 1.1 Data Preparation

[0071] The data required for geological modeling using 3D modeling software (such as Petrel software) include:

[0072] ① Well head data: includes well name, surface well location coordinates, and surface core elevation. There are 34 wells in total, including 26 gas storage production wells and monitoring wells, and 8 gas reservoir development wells (Facies 10, 12, 13, 14, 16, 18, 25, and 30).

[0073] Well deviation data (2): Well deviation data can be a combination of various data types that reflect the well deviation condition. For example, a combination of soundings, well deviation angle, and azimuth angle, or a combination of soundings, X-direction displacement, and Y-direction displacement can all be used as well deviation data. Well deviation data files are primarily used for well deviation correction. The well deviation data used in this modeling process uses a combination of soundings, well deviation angle, and azimuth angle.

[0074] ③Stratification data (well top): Stratification data include well name, layer number, layer type, single well stratification point sounding and other information. There are 287 wells involved in stratification in this area. The layer codes are S1, S21, S22, S23-0, S23-1, S23-2, and S23-3. The calibrated depth is the sounding, which is the bottom depth of each small layer code.

[0075] ④ Fault Data: Fault data can be generated by seismic tectonic interpretation programs or indirectly obtained from digital tectonic maps. After consultation with the Southwest Oil and Gas Field, to ensure consistency in the results, the fault data for this modeling was seismically interpreted faults, primarily including 82 faults. Furthermore, the interpretation of the F12 fault was supplemented to ensure the rationality of the fault and stratigraphic data.

[0076] ⑤ Stratum interpretation data: In the latest interpretation results, the Carboniferous System provides the top surface structure interpretation results due to the influence of seismic resolution. There are certain errors in the data processing of the contact relationship between the top and bottom surfaces of the Carboniferous System and the characterization of the bottom structure.

[0077] ⑥ Well logging curves and well logging interpretation data: This includes depth and well logging values. Depth is most commonly measured using depth logging, and well logging values ​​are the values ​​corresponding to different well logging curves at a specific depth. Well logging interpretation data includes the top and bottom depths of the interpreted interval, as well as interpreted data such as porosity, permeability, and gas saturation.

[0078] ⑦ Geological research data: 8 types of map data including top surface structure, reservoir thickness, energy storage coefficient, gas saturation, porosity distribution map, etc. are digitized and loaded into the software.

[0079] ⑧Production dynamic data: mainly includes perforation data of 27 wells.

[0080] 2.2 Data Quality Control

[0081] (1) Analysis and correction of stratification and well trajectory errors

[0082] Layering and well trajectory errors: After sorting out the basic geological data of 34 wells in the entire area (test well history, completion report, and completion handover book), it was found that there were conflicts in the layering data of different files. After loading the perforation dynamic data, it was inconsistent with the target layer of the previous layer. For example, there was a large error between the layering provided by the completion handover book of the Phase Reservoir 1 well and the geological layering of the test well history. Compared with the Phase Reservoir 6 and Phase Reservoir 9 wells, the reservoir section after the Phase Reservoir 1 well was above the target layer.

[0083] Cause of error: Some highly deviated wells were sidetracked multiple times (multiple stratified data for the same layer), and the stratified data in the model were not updated in a timely manner. In addition, due to multiple sets of well trajectory data generated by multiple lateral drillings, the well trajectory data used in the model were not updated in a timely manner.

[0084] Technical countermeasures: The well trajectories of 34 wells in the entire area were re-confirmed. Combined with the stratification data of the test oil and gas well history, handover documents, lithologic combination characteristics and perforation production dynamics, the entire area was re-stratified (such as the Phase Storage 1 well), providing accurate well stratification data for structural modeling.

[0085] Comparison of the well profile after the formation: Due to weathering and erosion, the actual drilling residual thickness is only 9m to 26.5m. The northern part of the formation is relatively thick, while the wing and southern sections are slightly thinner.

[0086] (2) Characterization of the erosion zone and data processing of the Carboniferous base structure

[0087] In 2021, Southwest Oil and Gas Field carried out the latest interpretation research on Xiangguo Temple, in which the Carboniferous System provided the top surface structural interpretation results due to the influence of seismic resolution. Because the target layer is a high-steep structure and a developed erosion area, there are certain errors in the characterization data processing of the top and bottom surface contact relationship of the Carboniferous System and the bottom structure.

[0088] In this study, the original interpretation data was used as the basis, and the well point data was used as the hard data to correct the interpreted layer data. The data near the fault was checked and cleaned for the corrected data. Combined with the previous research results of the Southwest Exploration and Research Institute, the scope of the erosion zone was characterized and the structural data of the top surface of the Carboniferous System was processed.

[0089] Based on a thorough analysis of previous research results, a Carboniferous stratum thickness map was generated using the existing stratum thickness map, with the drilled stratum thickness as a constraint and taking into account the influence of the erosion zone, providing a basis for the selection of stratigraphic modeling methods.

[0090] (3) Processing and analysis of logging data and experimental data

[0091] Well logging and attribute data: There are problems such as inconsistent units between previous data and new drilling data curves, outliers, errors between plane prediction data and well data, and unprocessed experimental abnormal data.

[0092] Data analysis and map correction: Based on the porosity plane map, the original map is preliminarily corrected using logging interpretation (core experiment) data. Combined with the distribution of the denuded area, the porosity data distribution characteristics are extracted, and a corrected porosity distribution map is produced.

[0093] Correlation analysis: Based on processed data and maps, explore the correlation between different parameters and provide trends and parameter constraints for subsequent attribute modeling.

[0094] (4) Selection of model and constraint data

[0095] In the early stage, the phase and attribute model established with the No. 1 and No. 3 regional boundary faults as the boundaries had structural errors. After resampling, the grid position and the new structural grid position attributes changed. In this study, we tried to use the new and old top and bottom surface structural constraints to perform velocity conversion to correct the original structure, so as to correct the original attribute model to the new structure. However, due to the increased velocity error in the high-steep structure, the correction effect was poor. Finally, we adopted the extraction of the original attribute body plane attribute distribution trend constraint to establish a new attribute model.

[0096] Based on the collation and quality control of the above data, a rich database of geological understanding, stratification, drilling and logging interpretation and analysis, oil and gas testing, and seismic interpretation has been formed (Table 1), laying a good foundation for geological modeling.

[0097] Table 1 - Statistics of geological model data collation and quality control work

[0098]

[0099]

[0100] 1.3 Model Improvement Process

[0101] Based on data quality control and preparation, the geological model was improved by reconstructing the structural model, attribute model, reserve calculation and uncertainty analysis (such as Figure 2 It mainly includes the following contents:

[0102] (1) For the fault model, the complex structural modeling was first used to generate the fault model for the selected 27 faults. The fault model data was exported for processing, and the target layer segment fault was intercepted for Pillar gridding to regenerate the fault model, and the fault morphology and grid quality were controlled.

[0103] (2) Integrating drilling, seismic interpretation, and structural map data, the corner grid is used to interactively process the horizon data with the complex structural model. A make zone is selected to establish a horizon model to control the distribution of strata and denudation zones. Manual adjustments to the horizon model are performed when necessary to ensure consistency between the horizon and the well points. Local structural adjustments are made based on well trajectories, geological stratification, and the latest perforation data to ensure the rationality of the reservoir sections and that the stratification is fully consistent with the structural surface.

[0104] (3) Based on the well spacing, reservoir characteristics, and numerical simulation requirements, the model is gridded (20*20 in the plane, and the reservoir is divided into 10 equal parts vertically).

[0105] (4) Normalize the previously loaded logging data and the newly drilled logging data, and import and coarsen the logging data.

[0106] (5) The coordinates and depth of the previous attribute model were corrected and resampled. The attribute model was established using a random simulation method with geological maps and the corrected attribute model as constraints.

[0107] (6) Check the model grid quality by checking the grid volume and other means, and perform reserve calculation and uncertainty assessment on the model.

[0108] 2. Establishment of the structural model

[0109] The structural model is the basis and carrier of the sedimentary microfacies model and physical property model. It includes fault model, layer model, model gridding, vertical subdivision layer and model quality inspection. The establishment of the structural model should pay attention to the following three aspects:

[0110] (1) The grid division should be as fine as possible both in the horizontal and vertical directions to lay the foundation for random simulation of reservoir heterogeneity.

[0111] (2) Ensure that the fault passes through the breakpoint to prevent the oil layer in the uplift plate from flowing to the downlift plate or vice versa, thus affecting the numerical model results.

[0112] (3) Ensure that the stratification in the model is consistent with the well stratification to avoid wrong layers when perforating in the digital model.

[0113] 2.1 Difficulties in structural modeling

[0114] The Xiangguosi Gas Storage is a narrow, fault-horst anticline controlled by a dipping reverse fault. Its primary structure is controlled by the F1, F5, F2, F4, and F3 faults. These reverse faults are wide at the top and narrow at the bottom, intersecting spatially at their bases, forming a Y-shaped planar layout. The quality of the fault mesh generated by the corner grids is poor. Furthermore, the Carboniferous reservoir and the overlying Liangshan Formation caprock are thin and steeply stratified, with erosion occurring in the northern, central, and southern portions of the formation. Characterizing the eroded zone using the modeling methods provided by Petrel software is challenging. Furthermore, most injection and production wells in the study area are highly deviated, making structural bedding control and well matching in the thin-bedded area challenging.

[0115] Based on this, complex structural modeling and Pillar Gridding were used to interactively process faults, and a method for processing the true thickness of high-steep strata and erosion surfaces by well calibration was explored. By comparing multiple methods and optimizing the Make zone method combined with the high-angle well trajectory correction method, a structural model of the cap rock-reservoir-support layer of the Xiangguo Temple high-steep thrust structure was established, and the stratigraphic characterization of the northern, central and southern erosion zones was completed.

[0116] Based on this, complex structural modeling and Pillar Gridding were used to interactively process faults, and a method for processing the true thickness of high-steep strata and erosion surfaces by well calibration was explored. By comparing multiple methods and optimizing the Make zone method combined with the high-angle well trajectory correction method, a structural model of the cap rock-reservoir-support layer of the Xiangguo Temple high-steep thrust structure was established, and the stratigraphic characterization of the northern, central and southern erosion zones was completed.

[0117] 2.2 Establishment of fault model

[0118] The fault relationship in the study area is complex, and there are many Y-shaped faults. This time, corner point grid modeling and complex structural modeling are used to interactively process the fault contact relationship and establish a structural fault model.

[0119] By examining the contact relationships between faults and stratigraphic data, the contact relationships between faults and stratigraphic layers in the deformation zone were analyzed. The missing fault f12 was reinterpreted and supplemented. Faults f1 and f3, f4 and f5, and f12 and f8 intersect to form a Y-shaped fault. Using the Structural Framework modeling method provided by Petrel software, the intersection relationships of 12 faults were analyzed to establish a complex structural fault model. Based on the model's longitudinal scope, the fault between the base of the Longtan Formation and the Silurian strata was intercepted and processed to provide fault data for pillar gridding. Through multiple adjustments and quality control, the fault model was established and refined.

[0120] By establishing a fault model, the characterization of the fault spatial distribution is completed, providing a good basis for fault stability evaluation.

[0121] 2.3 Establishment of the layer model

[0122] A layer model reflects the three-dimensional distribution of stratigraphic layers and provides an intuitive representation of the three-dimensional digitalization of geological stratification. Its principle is to combine single-well layer data with previous structural interpretation results and, through interwell interpolation algorithms, to form a data field between previously unconnected single-well layer data points. This data field then drives the generation of surfaces. When constructing a layer model using Petrel, it is important to provide as many structural trend reference surfaces as possible to obtain a layer model that best reflects geological understanding.

[0123] like Figure 3 The following are several cases of selecting structural trend reference surfaces. In actual work, the structural characteristics of the strata are diverse, and the selection of structural trend reference surfaces cannot be guided by a single reference surface for all layers. Different reference surfaces must be selected based on actual conditions. Establishing a layer model is a time-consuming task. The layers automatically generated by the software system are often uneven and have layers that jump up and down. While keeping the layer data unchanged as much as possible, the direction and trend of the contour lines in the layers are made to conform to natural laws and actual understanding. At the same time, the spatial position relationship between layers is taken into consideration, and the adjustment of areas without well point data is used as the main means to achieve the goal of no abnormal points (points) in the same layer and natural fluctuations and smooth transitions between layers.

[0124] Due to the influence of seismic resolution and the sedimentary strata in the study area, the seismic interpretation data for the Xiangguo Temple target layer only includes top surface interpretation data, while the bottom boundary data required for geological modeling are not provided. Based on this, this study first used the top surface interpretation data and formation thickness data to obtain bottom surface data. Based on the seismic prediction of formation thickness, combined with the plumb bob thickness calculated from actual drilling and inverse virtual well geological stratification as a constraint, a formation thickness map was generated. In combination with drilling data, structural sedimentary background, and geological understanding, the erosion zone polygon was delineated and the formation thickness within the erosion zone polygon was assigned a value of 0. This generated a formation thickness map containing the erosion zone. Using the formula bottom surface = top surface - formation thickness, the bottom surface data of the erosion formation was obtained.

[0125] This study explored the method of well calibration to process the true thickness of high-steep strata and process the erosion surface. By comparing multiple methods and optimizing the Makezone method combined with the high-angle well trajectory correction method, a structural model of the cap rock-reservoir-support layer of the Xiangguo Temple high-steep thrust structure was established, and the stratigraphic characterization of the northern, central and southern erosion zones was completed.

[0126] 2.4 Establishment of Skeleton Mesh System

[0127] After establishing the fault model and before establishing the layer model, it is necessary to define the horizontal and vertical identification range and resolution of the entire model. Pillar gridding is the process of projecting the area and volume of the study area into the Petrel software and assigning it a certain resolution. The grid system is the cornerstone of a geological model, just like the skeleton of an organism, playing a fundamental supporting role. The size of its resolution directly reflects the degree of subdivision of the sedimentary bodies in the geological model, which further affects the relevance of its processing and calculation by the simulation algorithm. In short, the higher the resolution, the greater the number of subdivided sedimentary bodies involved in the simulation calculation (the number of grids), and the higher the simulation accuracy. When establishing the grid system, the trend line constraint function should be used appropriately: when modeling work areas with simple structures and no fault development, try to use trend lines as little as possible or not at all. When modeling work areas with complex structures and developed faults, ① fault lines can be selectively used as trend lines; ② trend lines parallel or perpendicular to fault lines can be manually drawn to constrain grid generation. The best effect of the grid system is: the grid density can meet the requirements of modeling accuracy while being able to withstand the limitations of computer processing capabilities; the overall grid shape is smooth without grid singular points or singular areas (upward or twisted).

[0128] The modeling area extends 5.5 km west of the F1 fault, east to Well 28, south to the north of Well 11, and north to Well 19, encompassing an area of ​​327 km². Taking into account the main structural strike, fault orientation, and computational speed of subsequent numerical simulations, the grid orientation for this study was set along the main structural direction. Given the minimum spacing of gas storage wells of 398 m (Wells 4-9) and the minimum spacing of development wells less than 200 m, the horizontal grid size was 20 m x 20 m, totaling approximately 1.028 million horizontal grid cells. Vertical grid division primarily considers the vertical variation of the Huanglong Formation reservoir properties, dividing it into 10 layers. Furthermore, based on well logging characteristics and considering the needs of subsequent geomechanical evaluation of the caprock, the caprock was vertically subdivided into 10 layers. The Maokou-Qixia Formation and the underlying Silurian strata were each subdivided into 5 layers, for a total of 30 vertical subdivisions, totaling approximately 30.83 million grid cells in the study area.

[0129] 2.5 Construction model testing

[0130] Affected by the characteristics of highly deviated wells and high-steep thrust structures, the trajectories of some wells deviated from the original formations. In response, local structural adjustments and verifications were carried out in combination with the well trajectories and geological stratification to ensure the accuracy of the structural model. In combination with the numerical model, local structural adjustments were made after loading the latest perforation data to ensure the rationality of the reservoir section and complete consistency between the stratification and the structural surface.

[0131] 3. Establishment of attribute model

[0132] Different reservoir property parameters can reflect different reservoir characteristics from different perspectives. For example, permeability reflects the distribution characteristics of permeable layers in the reservoir. Its corresponding reservoir physical property parameter model can well reflect the geological conditions of the movement of formation fluids in the reservoir and reflect the three-dimensional macroscopic heterogeneity of the reservoir. Effective thickness reflects the distribution characteristics of oil and gas layers in the reservoir. Therefore, its corresponding reservoir physical property parameter model can be used to calculate the geological reserves of the reservoir. Therefore, different reservoir property models can be established to meet different research needs. In terms of reservoir property modeling, the more popular modeling method is "phase-controlled modeling", also known as "secondary modeling". That is, after the structural model is established, the sedimentary (micro) phase model is first established. Then, based on the sedimentary phase model, the phase belt is used to control the interpolation between wells. The interpolation of different phase belts is independent of each other, and different interpolation methods can also be used.

[0133] Phase-controlled physical property modeling essentially involves considering the sedimentary microfacies at the location of a physical property point when simulating physical properties such as porosity and permeability. From a statistical perspective, phase-controlled modeling involves dividing a sample area into several small blocks of varying sizes based on certain sample attributes. Random interpolation is then performed within each block using stochastic simulation to achieve the optimal simulated sample distribution. From a geological perspective, phase-controlled attribute modeling assumes that reservoir parameters in different facies zones have different expected values ​​and variances due to factors such as diagenesis, and therefore differ in their spatial correlations. Combining knowledge of geology and statistics, physical property interpolation requires simulation based on the sedimentary microfacies to which they belong.

[0134] Based on the results of previous geological understanding and combined with porosity and permeability data, the reservoirs in the study area are divided into four categories. During the modeling process, the porosity curve is mainly used as the basis to classify the reservoir sections vertically. Among them, the porosity of Class I reservoirs is Ф≧12.0%, the porosity of Class II reservoirs is 6.0%≦Ф<12.0%, the porosity of Class III reservoirs is 2.0%≦Ф<6.0%, and the porosity of Class IV reservoirs is Ф<2.0%.

[0135] Based on previous geological understanding, the reservoir rock is primarily breccia dolomite, with pores as the primary reservoir space. The reservoir permeability is fracture-porous, with individual well porosity ranging from 1% to 12%, matrix permeability from 0.01mD to 10mD, and well test-interpreted permeabilities from 19.74mD to 1200mD, generally exceeding 100mD. The study first normalized the porosity, permeability, and water saturation curves from 23 wells to ensure data consistency. Using reservoir classification standards, reservoir classification curves were generated using porosity thresholds, replacing Class IV reservoirs with a 2.5% cutoff value. A final classification was performed for all wells with porosity interpretation, and lithofacies classification curves were calculated.

[0136] 3.1 Curve discretization and data correction

[0137] Due to the accuracy of the model's vertical subdivision layers, the data sampled to the grid using the through cell discretization method provided by the Petrel software exhibits significant deviations when discretizing the lithofacies classification curve using the through cell method. Therefore, the neighbor cell method was selected for curve coarsening. When coarsening the reservoir attribute curves, based on the data characteristics, the Arithmetic method was selected for coarsening the porosity, water saturation, and NTG curves, and the Harmonic method was selected for coarsening the permeability curve.

[0138] SGS requires that the data must be normally distributed before modeling. The study found that the porosity data from well logging and cores tend to show skewed or non-normal distributions, which do not conform to the normal distribution. In carbonate rocks, this is due to the combination of different porosity categories (e.g., intragranular, loose) and the combined effects of diagenesis with sedimentation and diagenetic / burial structures. Therefore, before modeling, the input porosity data is subjected to the standard distribution transformation (NST). After the modeling is completed, the normalized values ​​are converted back to "true" porosity values. Based on this, based on the porosity isoplanar map, the original map is preliminarily corrected using well logging interpretation (core experiment) data, and combined with the distribution characteristics of the denudation zone, the distribution characteristics of porosity and other data are extracted, and the corrected porosity and other physical property distribution maps are generated, providing a basis for the subsequent variogram analysis.

[0139] 3.2 Variogram Analysis

[0140] The variogram extraction first used the corrected porosity plane map to extract the variogram. The intersection relationship between the discretized well data and the plane porosity distribution map was y = 0787835*x + 0.00598948, and the correlation coefficient was 0.61, which provided a basis for the correlation analysis of the second constraint variable of the attribute.

[0141] The stability of the variogram is inherently related to the azimuth of the ellipse's principal direction and the data. Based on normally distributed data and the extracted variogram, a variogram analysis was conducted on discretized porosity well data. Parameters such as the primary and secondary ranges were determined, providing a basis for stochastic attribute modeling.

[0142] 3.3 Lithofacies model

[0143] From the coarsening of the discrete curves to the reservoir grid characteristics, it can be seen that the reservoirs with better quality (grades I and II) are generally located in the middle of the formation. Similarly, the poorer grade IV reservoir quality is mainly found at the top and bottom of the formation. The facies modeling adopts the Sequential Indicator Simulation (SIS) algorithm. It allows different indicator variogram models to be assigned to each facies. In addition to hard data (wells), SIS also provides a variety of methods to adapt the simulation to trends in soft data or auxiliary data obtained from geological interpretation or geophysical measurements. Based on the discretized data, variogram analysis was carried out on different facies, and the Xiangguosi Huanglong Formation facies model was established using stochastic simulation.

[0144] 3.4 Porosity model

[0145] For modeling porosity and other continuous properties, Sequential Gaussian Simulation (SGS) is used as the modeling algorithm. SGS is a process that generates a Gaussian field using the kriging mean and variance. Input data must be normally distributed (hence, a normal distribution transformation is performed before modeling). When calculating values ​​for unsimulated grid cells, it uses both the input data and the simulated data.

[0146] This study used normally distributed porosity data and variogram analysis results, selected the sequential Gaussian random simulation method, and established a reservoir porosity model using the planar porosity distribution as a trend constraint (with a correlation coefficient of 0.61). Relatively high-value areas were XC19-XC11, XC3, XC4, XC8, and XC18, consistent with previous geological understanding.

[0147] 3.5 Permeability Model

[0148] For reservoir modeling purposes, two different permeability attributes are generated: matrix permeability derived from core experiments and total “effective” permeability interpreted from well tests.

[0149] Reservoir engineering analysis indicates that well test permeabilities are 1-2 orders of magnitude greater than core / log permeabilities. This suggests that the primary contribution to reservoir flow comes from excess permeability, typically connected fractures or macropores, that far exceeds what is explained by log and core data. Fracture permeability is generally independent of porosity, necessitating separate modeling of effective reservoir permeability using core, imaging, and well test data.

[0150] Modeling the permeability of the logs in the same manner as porosity modeling is not ideal for a number of reasons. The permeability distribution is log-normal, and coarsening it tends to constrain the data, resulting in missing data at the end. In the example shown here, the log permeability data ranges from 0–134 mD, while the coarsened data ranges from 0–14 mD. Therefore, a property calculator was used to generate the matrix permeability in the model. The calculation formula was borrowed from existing results (derived from RCA / core experimental data):

[0151] K=PER3*(Φ / POR3) 4.4

[0152] Where: PER3 is the basal pore permeability corresponding to the lower limit of pore volume (0.01 mD), and POR3 is the lower limit of effective porosity (2.5% for the Carboniferous system in eastern Sichuan).

[0153] Similar to many fractured carbonate reservoirs, fractures are key to improving productivity, but in reality, the impact of natural gas storage is minimal. Due to the large deviation between the permeability interpreted by well logging and the well test permeability, in actual research, permeability characterization focuses on integrating the well test permeability (effective permeability) with seismic and fracture attributes to generate a permeability model that accurately matches the production history. Based on this, the well test permeability interpreted by existing wells was combined with the permeability interpreted by well logging. Due to the influence of reservoir permeability heterogeneity, the well logging and well test permeability data of each single well were combined in this study. The well test permeability curve of each well was calculated based on the obtained relationship, and the well test permeability model was obtained using the sequential Gaussian random simulation algorithm.

[0154] 3.6 Water saturation model

[0155] Based on the analysis of gas saturation data interpreted by well logging, combined with production performance and test data research, with reference to structural morphological characteristics, using the porosity model as control, and the gas saturation plane distribution map as trend constraint, a water saturation model was established using the sequential Gaussian simulation algorithm.

[0156] 3.7 Net-to-Gross (NTG) Model

[0157] Based on the porosity and gas saturation model and combined with production and operation dynamics, a threshold value for calculating the static-to-gross ratio (NTG) was designed. When the porosity is ≥2.5% and the gas saturation is ≥75%, the NTG value is 1; otherwise, it is 0. Based on this, the NTG model for the Huanglong Formation was directly calculated.

[0158] IV. Reserve Calculation and Uncertainty Analysis

[0159] The study used reservoir attributes from the new 3D model to calculate reserves for the Xiangguosi Huanglong Formation gas reservoir. The reserves were then cross-checked against plan-based calculations using average reservoir thickness and attribute values ​​from the model input data. Formation volume factors were derived from previous studies and validated as part of a pre-history matching assessment for reservoir engineering. The modeled volumes were then compared with previous estimates from various sources to analyze the uncertainty of the reserves.

[0160] 4.1 Model reserves

[0161] The Xiangguosi gas reservoir's geological reserves are contained within a highly elongated ridge measuring approximately 24 km by 1.5 km. The northern end of the reservoir is closed by a dip, while the southern section is defined by faults and stratigraphic gaps. Previous research indicates that the gas-water contact is defined at -1986 m, but the gas saturation model uses two major fault blocks where wells have been drilled and confirmed to contain natural gas. For this study, geological reserves were calculated by dividing the study area into three zones: North, Central, and South, based on the reservoir's geological characteristics and the distribution of existing wells.

[0162] The GIIP for the Xiangguosi Huanglong Formation gas reservoir is based on the following elements:

[0163] The gas-water interface - reservoir thickness is above -1986m, and the range is within the structural core and fault block.

[0164] Net-to-gross ratio—calculated based on the porosity and gas saturation model.

[0165] Porosity—obtained from the porosity model by coarsening the effective porosity, with reservoir classification and porosity trend map as constraints.

[0166] Gas saturation - obtained by analyzing the water saturation model and the gas saturation data interpreted by well logging, and is taken as 82%.

[0167] Bg—take 0.00373 (1 / 268).

[0168] Based on the above parameter settings, the model was divided into three areas to calculate the GIIP of each grid cell in the Huanglong Formation reservoir. The calculated result was 44.83×10 8 m 3 , the total pore volume is 21×10 6 m 3 .

[0169] The GIIP distribution is derived by summing the K layers, and the reserves are mainly distributed in the XC21, XC18, XC15-XC8, XC1, XC4, XC3, XC2, and XC11 well areas.

[0170] 4.2 Uncertainty Analysis

[0171] The calculated reserves for the Xiangguosi Huanglong Formation reservoir are 4.483 billion cubic meters, consistent with the previous static reserves of 4.45 billion cubic meters (dynamic reserves of 4.39 billion cubic meters), with a fitting accuracy of >98%. No alternative thresholds were used for porosity or the Sw attribute in this study. Therefore, the only threshold used in the GIIP calculation was porosity for reservoir classification. Due to constraints such as porosity trends, the fitting results have some uncertainty. Based on this, the P10, P50, and P90 reserves were analyzed to be 3.50 billion cubic meters, 4.47 billion cubic meters, and 5.72 billion cubic meters, respectively. The differences in reserves are within reasonable limits, indicating that the reserve results are reliable.

[0172] Example 2

[0173] This embodiment provides a three-dimensional geological modeling system for a gas storage reservoir, including a modeling data preprocessing module, a structural model building module, an attribute model building module, and a reserve calculation and uncertainty analysis module, wherein:

[0174] The modeling data preprocessing module is configured to analyze and correct stratification and well trajectory errors, characterize the scope of the erosion zone and process stratigraphic structural data, process and analyze logging data and experimental data, determine the model and constraint body data, and form a database including geological recognition data, stratification data, drilling and logging interpretation and analysis data, oil and gas testing data, and seismic interpretation data.

[0175] The structural model building module is configured to build a structural model based on a plane gridding method and perform local structural adjustment and verification. The structural model includes a fault model and a layer model.

[0176] The property model building module is configured to simulate the physical property parameters based on the sedimentary microfacies type at the location of the physical property parameter point and form a property model. The physical property parameters include porosity, permeability, water saturation and net-to-gross ratio.

[0177] The Reserve Calculation and Uncertainty Analysis module is configured to calculate the reserves of a gas reservoir based on the structural model and the attribute model, using average reservoir thickness and attribute values ​​from the model input data to cross-check against the calculated values ​​based on the plan view; and compare the volumes in the structural model and the attribute model with previous estimates from various sources to complete the uncertainty analysis of the reserves.

[0178] Preferably, the modeling data preprocessing module includes:

[0179] The layer and well trajectory error analysis and correction unit is configured to address conflicts in layer data from different files. It combines basic geological data of gas wells, lithologic combination characteristics, and perforation production performance to re-stratify gas wells in the entire area and provide layer data for the establishment of the structural model.

[0180] The unit for characterizing the extent of the erosion zone and processing stratigraphic structure data is configured to use the original interpretation data as a basis and the well point data as hard data to correct the interpreted horizon data, check and clean the data near the fault on the corrected data, and characterize the extent of the erosion zone and process the stratigraphic structure data in combination with the previous research data; based on the existing stratigraphic thickness map and with the drilling stratigraphic thickness as a constraint, the stratigraphic thickness map is generated in combination with the influence of the erosion zone;

[0181] The logging and experimental data processing and analysis unit is configured to use the porosity plan map as a basis, utilize the logging interpretation data to perform preliminary corrections on the original map, combine the distribution of the denuded area, extract the distribution characteristics of the porosity data, and produce a corrected porosity distribution map. Based on the processed data and map, it explores the correlation between different parameters and provides trends and parameter constraints for the establishment of attribute models.

[0182] The model and constraint volume data selection unit is configured to extract the original attribute volume plane attribute distribution trend constraint to establish a new attribute model.

[0183] Preferably, the construction model building module includes:

[0184] A fault model building unit is configured to interactively process fault contact relationships using corner point grid modeling and complex structure modeling to build a fault model that characterizes the spatial distribution of the fault;

[0185] The layer model building unit is configured to use the single-well layer data combined with the previous structural interpretation results, and use the inter-well interpolation algorithm to form a data field between the originally unrelated single-well layer data points. Under the influence of this data field, a surface is generated to form a layer model.

[0186] A grid system establishing unit is configured to simulate and project the area and volume size of the study block onto a three-dimensional modeling software and set the resolution, thereby forming a grid system;

[0187] The structural model verification unit is configured to perform local structural adjustments and verification based on the well trajectory and geological stratification, and to load the latest perforation data in combination with the numerical model to perform local structural adjustments so that the stratification is completely consistent with the structural surface.

[0188] Preferably, in the grid system establishment unit, when constructing a simple research block model with undeveloped faults, trend lines are rarely used or not used; when constructing a complex research block model with developed faults, fault lines are used as trend lines or trend lines parallel / perpendicular to the fault lines are drawn to constrain the generation of the grid.

[0189] Preferably, the attribute model building module includes:

[0190] The curve discretization and data correction unit is configured to discretize the lithofacies classification curve, porosity, permeability, water saturation, and net-to-gross ratio curves. Based on the physical property parameter plane map, the original map is preliminarily corrected using well logging interpretation data. In combination with the distribution characteristics of the denudation zone, the distribution characteristics of the physical parameters are extracted and a corrected physical property parameter distribution map is generated.

[0191] a variogram analysis unit configured to perform variogram analysis on the discretized porosity well data based on the normally distributed data and the variogram extracted according to the calibrated porosity plane map, and determine relevant parameters including a major range and a minor range;

[0192] The lithofacies model building unit is configured to conduct variogram analysis on different lithofacies based on discretized data, and to build a lithofacies model using a stochastic simulation method based on a time-sequential indicator simulation method;

[0193] The porosity model building unit is configured to call the normally distributed porosity data and the variogram analysis results, and build a reservoir porosity model based on the sequential Gaussian random simulation method and the porosity plane distribution as a trend constraint;

[0194] a permeability model building unit configured to perform intersection processing using the well test permeability interpreted by the existing wells and the permeability interpreted by the well logging, calculate the well test permeability curve of each well based on the obtained relationship, and establish the well test permeability model using a sequential Gaussian random simulation method;

[0195] The water saturation model building unit is configured to build a water saturation model based on the analysis of gas saturation data interpreted from well logging, combined with production performance and test data, with reference to structural morphological characteristics, using the porosity model as a control, and the gas saturation plane distribution map as a trend constraint, using the sequential Gaussian simulation method;

[0196] The net-to-gross ratio model establishing unit is configured to establish a net-to-gross ratio model based on the porosity model and the gas saturation model, combined with production operation dynamics and calculation thresholds.

[0197] It should be noted that, for the sake of simplicity, the aforementioned method embodiments are described as a series of action combinations. However, those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

Claims

1. A three-dimensional geological modeling method for a gas storage reservoir, characterized in that: include: Modeling data preprocessing: Analyze and correct stratification and well trajectory errors, characterize the extent of the denudation zone and process stratigraphic structural data, process and analyze logging data and experimental data, determine the model and constraint volume data, and form a database including geological recognition data, stratification data, drilling and logging interpretation and analysis data, oil and gas testing data, and seismic interpretation data; Structural model establishment: Establishing a structural model based on a plane gridding method and performing local structural adjustments and verification. The structural model includes a fault model and a layer model. Establishing a property model: Based on the sedimentary microfacies type at the location of the physical property parameter point, the physical property parameters are simulated and a property model is formed. The physical property parameters include porosity, permeability, water saturation and net-to-gross ratio; Reserve calculation and uncertainty analysis: Calculate the reserves of the gas reservoir based on the structural and property models, using average reservoir thickness and property values ​​from the model input data to cross-check against the calculated values ​​based on the plan view; compare the volumes in the structural and property models with previous estimates from various sources to complete the uncertainty analysis of the reserves.

2. A three-dimensional geological modeling method for gas storage reservoirs according to claim 1, characterized in that: The modeling data preprocessing includes: Layer and well trajectory error analysis and correction: In case of conflicting layer data in different files, the gas wells in the entire area are re-layered in combination with basic geological data of gas wells, lithologic combination characteristics and perforation production performance to provide layer data for the establishment of structural models; Characterization of the erosion zone and stratigraphic structural data processing: Based on the original interpretation data and using well point data as hard data, the interpreted horizon data is corrected. The corrected data is then inspected and cleaned near the fault. Combined with previous research data, the scope of the erosion zone is characterized and stratigraphic structural data is processed. Based on the existing stratigraphic thickness map and with the drilled stratigraphic thickness as a constraint, a stratigraphic thickness map is generated in combination with the impact of the erosion zone. Processing and analysis of well logging and experimental data: Based on the porosity plan map, the original map is initially corrected using well logging interpretation data. In combination with the distribution of denuded areas, the porosity data distribution characteristics are extracted and a corrected porosity distribution map is produced. Based on the processed data and map, the correlation between different parameters is explored to provide trends and parameter constraints for the establishment of attribute models. Model and constraint body data selection: Extract the original attribute body plane attribute distribution trend constraints to establish a new attribute model.

3. The method for three-dimensional geological modeling of a gas storage reservoir according to claim 1, characterized in that: The construction model establishment includes: Fault model establishment: Corner point grid modeling and complex structure modeling are used to interactively process fault contact relationships and establish a fault model that characterizes the spatial distribution of faults; Layer model establishment: Utilizing single-well layer data combined with previous structural interpretation results, an inter-well interpolation algorithm is used to form a data field between previously unrelated single-well layer data points. This data field drives the generation of surfaces to form a layer model. Grid system establishment: The area and volume of the study area are simulated and projected into the 3D modeling software and the resolution is set to form a grid system; Structural model verification: local structural adjustments and verification are performed based on the well trajectory and geological stratification. The latest perforation data is loaded in combination with the numerical model to perform local structural adjustments to ensure that the stratification is completely consistent with the structural surface.

4. A three-dimensional geological modeling method for gas storage reservoirs according to claim 3, characterized in that: During the establishment of the grid system, when constructing a simple research block model with undeveloped faults, trend lines are rarely used or not used; when constructing a complex research block model with developed faults, fault lines are used as trend lines or trend lines parallel to / perpendicular to the fault lines are drawn to constrain the generation of the grid.

5. The method for three-dimensional geological modeling of a gas storage reservoir according to claim 1, characterized in that: The attribute model establishment includes: Curve discretization and data correction: Discretize the lithofacies classification curve, porosity, permeability, water saturation, and net-to-gross ratio curves; based on the physical property parameter plane map, use the well logging interpretation data to make a preliminary correction to the original map, and combine the distribution characteristics of the denudation zone to extract the distribution characteristics of the physical parameters and generate the corrected physical property parameter distribution map; Variogram analysis: Based on the normally distributed data and the variogram extracted from the calibrated porosity plot, a variogram analysis is performed on the discretized porosity well data to determine relevant parameters including the major and minor ranges. Establishment of lithofacies model: Based on the discretized data, variogram analysis is carried out on different lithofacies. Based on the time-sequential indicator simulation method, a lithofacies model is established using a stochastic simulation method. Porosity model establishment: Using the normally distributed porosity data and variogram analysis results, a reservoir porosity model is established based on the sequential Gaussian random simulation method and the porosity plane distribution as a trend constraint; Permeability model establishment: The well test permeability interpreted from existing wells is combined with the permeability interpreted from well logging. The well test permeability curve for each well is calculated based on the obtained relationship, and the well test permeability model is established using the sequential Gaussian random simulation method. Water saturation model establishment: Based on the analysis of gas saturation data interpreted from well logging, combined with production performance and test data, referring to structural morphological characteristics, using the porosity model as a control, and the gas saturation plane distribution map as a trend constraint, the water saturation model is established using the sequential Gaussian simulation method; Net-to-gross ratio model establishment: Based on the porosity model and gas saturation model, the net-to-gross ratio model is established in combination with production operation dynamics and calculation thresholds.

6. A three-dimensional geological modeling system for gas storage reservoirs, characterized in that: include: The modeling data preprocessing module is configured to analyze and correct layer and well trajectory errors, characterize the extent of the erosion zone and process stratigraphic structural data, process and analyze well logging data and experimental data, determine the model and constraint volume data, and form a database including geological recognition data, layer data, drilling and logging interpretation and analysis data, oil and gas testing data, and seismic interpretation data; a structural model building module configured to build a structural model based on a plane gridding method and perform local structural adjustment and verification, wherein the structural model includes a fault model and a layer model; a property model building module configured to simulate physical property parameters based on the sedimentary microfacies type at the location of the physical property parameter point and form a property model, wherein the physical property parameters include porosity, permeability, water saturation and net-to-gross ratio; The reserve calculation and uncertainty analysis module is configured to calculate the reserves of the gas reservoir based on the structural model and the attribute model, cross-check the calculated values ​​based on the plan view using the average reservoir thickness and the attribute values ​​of the model input data, and compare the volumes in the structural model and the attribute model with previous estimates from various sources to complete the uncertainty analysis of the reserves.

7. A gas storage reservoir 3D geological modeling system according to claim 6, characterized in that: The modeling data preprocessing module includes: The layer and well trajectory error analysis and correction unit is configured to address conflicts in layer data from different files. It combines basic geological data of gas wells, lithologic combination characteristics, and perforation production performance to re-stratify gas wells in the entire area and provide layer data for the establishment of the structural model. The unit for characterizing the extent of the erosion zone and processing stratigraphic structure data is configured to use the original interpretation data as a basis and the well point data as hard data to correct the interpreted horizon data, check and clean the data near the fault on the corrected data, and characterize the extent of the erosion zone and process the stratigraphic structure data in combination with the previous research data; based on the existing stratigraphic thickness map and with the drilling stratigraphic thickness as a constraint, the stratigraphic thickness map is generated in combination with the influence of the erosion zone; The logging and experimental data processing and analysis unit is configured to use the porosity plan map as a basis, utilize the logging interpretation data to perform preliminary corrections on the original map, combine the distribution of the denuded area, extract the distribution characteristics of the porosity data, and produce a corrected porosity distribution map. Based on the processed data and map, it explores the correlation between different parameters and provides trends and parameter constraints for the establishment of attribute models. The model and constraint volume data selection unit is configured to extract the original attribute volume plane attribute distribution trend constraint to establish a new attribute model.

8. The three-dimensional geological modeling system for gas storage reservoirs according to claim 6, characterized in that: The construction model building module includes: A fault model building unit is configured to interactively process fault contact relationships using corner point grid modeling and complex structure modeling to build a fault model that characterizes the spatial distribution of the fault; The layer model building unit is configured to use the single-well layer data combined with the previous structural interpretation results, and use the inter-well interpolation algorithm to form a data field between the originally unrelated single-well layer data points. Under the influence of this data field, a surface is generated to form a layer model. A grid system establishing unit is configured to simulate and project the area and volume size of the study block onto a three-dimensional modeling software and set the resolution, thereby forming a grid system; The structural model verification unit is configured to perform local structural adjustments and verification based on the well trajectory and geological stratification, and to load the latest perforation data in combination with the numerical model to perform local structural adjustments so that the stratification is completely consistent with the structural surface.

9. A gas storage reservoir 3D geological modeling system according to claim 8, characterized in that: In the grid system establishment unit, when constructing a simple research block model with undeveloped faults, trend lines are rarely used or not used; when constructing a complex research block model with developed faults, fault lines are used as trend lines or trend lines parallel / perpendicular to the fault lines are drawn to constrain the generation of the grid.

10. A three-dimensional geological modeling system for gas storage reservoirs according to claim 6, characterized in that: The attribute model building module includes: The curve discretization and data correction unit is configured to discretize the lithofacies classification curve, porosity, permeability, water saturation, and net-to-gross ratio curves. Based on the physical property parameter plane map, the original map is preliminarily corrected using well logging interpretation data. In combination with the distribution characteristics of the denudation zone, the distribution characteristics of the physical parameters are extracted and a corrected physical property parameter distribution map is generated. a variogram analysis unit configured to perform variogram analysis on the discretized porosity well data based on the normally distributed data and the variogram extracted according to the calibrated porosity plane map, and determine relevant parameters including a major range and a minor range; The lithofacies model building unit is configured to conduct variogram analysis on different lithofacies based on discretized data, and to build a lithofacies model using a stochastic simulation method based on a time-sequential indicator simulation method; The porosity model building unit is configured to call the normally distributed porosity data and the variogram analysis results, and build a reservoir porosity model based on the sequential Gaussian random simulation method and the porosity plane distribution as a trend constraint; a permeability model building unit configured to perform intersection processing using the well test permeability interpreted by the existing wells and the permeability interpreted by the well logging, calculate the well test permeability curve of each well based on the obtained relationship, and establish the well test permeability model using a sequential Gaussian random simulation method; The water saturation model building unit is configured to build a water saturation model based on the analysis of gas saturation data interpreted from well logging, combined with production performance and test data, with reference to structural morphological characteristics, using the porosity model as a control, and the gas saturation plane distribution map as a trend constraint, using the sequential Gaussian simulation method; The net-to-gross ratio model establishing unit is configured to establish a net-to-gross ratio model based on the porosity model and the gas saturation model, combined with production operation dynamics and calculation thresholds.

Images