A method for intelligently screening key data of basin simulation based on VBA program

Abstract

The present application belongs to the technical field of basin simulation, and particularly relates to a method for intelligently screening key data of basin simulation based on VBA program.S1: starting the program and loading original logging data;S2: logging data preprocessing;S3: constructing mudstone section reference data set;S4: counting mudstone samples and extracting adaptive threshold;S5: mudstone identification and continuous layer section division of the whole well section;S6: optimizing mudstone acoustic travel time data;S7: outputting mudstone identification threshold and identification direction of each target lithology identification logging curve of the target single well, top boundary and bottom boundary of the continuous mudstone layer section, representative TVD and AC average value, acoustic travel time correction value and its dimensionless parameter to the Excel worksheet.The present application realizes integrated processing of mudstone sample counting, mudstone identification of the whole well section, continuous layer section division, parameter calculation and result output, and provides data basis for formation compaction trend analysis, formation denudation amount recovery, and recovery of burial history, thermal history and hydrocarbon generation history.

Description

Technical Field

[0001] This invention belongs to the field of basin simulation technology, specifically relating to an intelligent screening method for key basin simulation data based on VBA programs. Background Technology

[0002] In basin simulation, the recovery of formation erosion is a crucial foundation for establishing the stratigraphic framework and reconstructing burial, thermal, and hydrocarbon generation histories. In practical applications, formation erosion is typically recovered quantitatively based on formation compaction trends. Due to the complex and variable lithology of underground formations, accurate identification, unified classification, and effective screening of the lithology of the study well section are usually necessary to obtain fundamental data that effectively characterizes normal compaction trends. The accuracy of lithology identification and screening, as well as the rationality of the constraints imposed by normal compaction trends, directly affects the accuracy of formation erosion calculations and further impacts the reliability of basin burial evolution, thermal evolution, hydrocarbon generation evolution, and hydrocarbon accumulation analysis results. Drilling data, typically including logging and well logging data, is an important source of data for quantitatively obtaining basin simulation parameters. However, single-well study sections are often thousands of meters long, and directly usable known lithology labels often only cover a portion of the well section, making it difficult to provide a complete basis for lithology identification across the entire well section. Therefore, how to quickly identify mudstone in the entire well section using limited known mudstone samples and screen out key data that can better reflect the normal compaction trend of the formation has become an urgent problem to be solved in the process of efficiently obtaining basin simulation parameters.

[0003] Currently, when the known mudstone samples are incomplete, the screening of drilling lithology data mainly relies on manual judgment by technicians based on the characteristics of well logging curves. A common method is to identify mudstone intervals based on well logging data such as the natural gamma ray curve (GR), spontaneous potential curve (SP), and mudstone content curve (SH), combined with manually set empirical thresholds, and then further statistically analyze the corresponding mudstone parameters.

[0004] The above methods have the following main shortcomings: First, the setting of empirical thresholds relies on the subjective judgment of technical personnel, and different personnel may use different judgment standards, resulting in poor consistency and repeatability of lithology identification results. Second, single-well logging data is characterized by large well intervals and dense sampling points. Manually analyzing and dividing kilometer-level well intervals point by point and layer by layer is labor-intensive and inefficient, making it difficult to meet the needs of rapid processing of multi-well, large-volume data. Third, traditional manual processing methods can usually only complete the preliminary identification of mudstone intervals, and it is difficult to simultaneously achieve automatic division of continuous mudstone intervals and quantitative extraction of related parameters. Subsequent manual processing and calculation are still required, increasing the complexity of the data processing flow. Fourth, the sedimentary environment, lithological assemblage, and logging response characteristics vary in different study areas. Using fixed empirical thresholds for lithology identification is difficult to adapt to the specific geological conditions of different study areas.

[0005] Existing technologies have proposed methods for lithology identification and classification based on machine learning and multi-logging curve threshold fusion. However, some methods rely on dedicated programming environments, specialized geological software, or complex algorithm frameworks, requiring high levels of expertise from users and sophisticated hardware and software environments. Furthermore, they lack convenient data integration and a unified processing mechanism with the commonly used Excel logging data tables in exploration sites. While Excel is a common tool for storing, organizing, and conducting preliminary analysis of drilling and logging data, its built-in VBA functions offer batch calculation and automated workflow capabilities. However, current technologies lack an integrated method that can adaptively determine the lithology identification threshold and discrimination direction for a single well using limited mudstone samples, and complete mudstone identification across the entire well section, continuous mudstone strata division, and key basin simulation data screening within the Excel environment. Summary of the Invention

[0006] To address the problems of low efficiency in manual processing, poor inter-well adaptability of fixed thresholds, and insufficient consistency in mudstone interval identification under complex sandstone-mudstone interbedded conditions in current basin simulation key parameter acquisition processes, this invention provides an intelligent screening method for key basin simulation data based on a VBA program in Excel. This method is applicable to complex geological exploration conditions such as complex lithology, well-developed sandstone-mudstone interbedded layers, dense logging data, and well sections not covered by known lithology labels. It automatically identifies and screens key basin simulation data that characterizes formation compaction trends. The method uses a VBA program to statistically analyze the logging response differences between mudstone samples and full-well data, automatically determining the discrimination direction and wide boundary threshold for each logging curve. Finally, it uses an intersection judgment method that simultaneously satisfies multiple conditions to identify high-confidence mudstone data and divides continuous mudstone intervals according to a 0.125 m sampling interval, thereby achieving standardized, automated, and batch screening of single-well mudstone data.

[0007] The technical solution adopted in this invention is as follows: A method for intelligent screening of key data in basin simulation based on VBA program includes the following steps:

[0008] S1: Start the program and load raw logging data

[0009] Start the VBA program on the Excel platform, load the original logging data of the target single well arranged in ascending order of true vertical depth (TVD), the original logging data includes true vertical depth (TVD), sonic transit time (AC), and at least two valid target lithology discrimination logging curves, and the necessary data are complete. Select spontaneous potential (SP) curve, natural gamma ray (GR) curve, resistivity (RT) curve, and logging clay content (SH) curve as candidate target lithology discrimination logging curves; the true vertical depth (TVD) interval between adjacent sampling points is 0.125 m (STEP=0.125).

[0010] S2: Well logging data preprocessing

[0011] The raw logging data after loading is preprocessed, including but not limited to outlier detection and removal, removal of outlier data segments during logging instrument start and end, data standardization or normalization, curve smoothing filtering and noise removal, in order to reduce the impact of outlier data and random noise on subsequent threshold statistics and mudstone identification.

[0012] S3: Constructing a benchmark dataset for mudstone sections

[0013] Mudstone was selected as the target lithology to characterize the normal compaction trend of the formation. Considering the differences in the type and completeness of existing data from different wells, known mudstone samples were calibrated using at least one of the following methods based on the actual data available for the target well: well logging shale content (SH) curves, manual interpretation results, and cuttings logging lithology labels. When the well logging shale content (SH) curves had valid data, the logging points with SH curves greater than the preset shale content threshold were used as the initial mudstone samples, and supplemented or verified using manual interpretation results or cuttings logging lithology labels. When the judgment results from different data sources contradicted each other, the corresponding samples were removed. The true vertical depth (TVD), sonic transit time (AC), and target lithology discrimination logging curve data corresponding to the verified mudstone samples were collected to construct a standardized mudstone section benchmark dataset, providing standard sample support for the adaptive determination of subsequent single-well discrimination direction and wide boundary threshold.

[0014] S4: Statistical analysis of mudstone samples and extraction of adaptive thresholds

[0015] The VBA program automatically traverses the benchmark dataset of the mudstone section and the effective logging dataset of the entire well section, statistically analyzes the differences in logging responses between the mudstone samples and the data of the entire well section, and determines the discrimination direction and wide boundary threshold of the logging curves for each target lithology.

[0016] Specifically, the VBA program iterates through the benchmark dataset for the mudstone section. Based on the type and validity of the logging curves actually acquired from the target well, if both the spontaneous potential (SP) curve and the natural gamma ray (GR) curve have valid data among the spontaneous potential (SP) curve, natural gamma ray (GR) curve, resistivity (RT) curve, and mud content (SH) curve, then the SP curve and the GR curve are selected as the target lithology discrimination logging curves. Otherwise, at least two curves with valid data are selected from the four candidate target lithology discrimination logging curves as the target lithology discrimination logging curves. Differentiate logging curves; calculate the mean value of each target lithology discrimination logging curve in the mudstone sample and the whole well section data respectively, and determine the mudstone discrimination direction and wide boundary threshold for each curve. When the mean value of the mudstone sample is greater than or equal to the mean value of the whole well section, it is determined as the discrimination direction, and the minimum value of the mudstone sample is determined as the wide boundary threshold of the logging curve; when the mean value of the mudstone sample is less than the mean value of the whole well section, it is determined as the discrimination direction, and the maximum value of the mudstone sample is determined as the wide boundary threshold of the logging curve, thus forming adaptive identification parameters adapted to the target single well.

[0017] S5: Mudstone Identification and Continuous Stratum Division Throughout the Well Section

[0018] The VBA program automatically traverses all valid logging data in the well section. Based on the discrimination direction and wide boundary threshold determined in step S4, it uses the intersection rule that multiple target lithology discrimination logging curves simultaneously satisfy the conditions to identify mudstone points and divide continuous mudstone layers.

[0019] Specifically, the VBA program iterates through the logging data of the entire well section, determining whether each target lithology discrimination logging curve selected in step S4 meets the corresponding wide boundary threshold and discrimination direction. Only when the logging data point to be judged simultaneously meets the wide boundary threshold and corresponding discrimination direction of all selected target lithology discrimination logging curves is the logging data point marked as a mudstone point. For logging data arranged in ascending order of true vertical depth (TVD), when the current mudstone point and the previous mudstone point are adjacent sampling points and the difference in their true vertical depth (TVD) is 0.125 m, the two mudstone points are classified into the same continuous mudstone interval. When the difference in the true vertical depth (TVD) of adjacent mudstone points is greater than 0.125 m, or when there are non-mudstone sampling points between the two mudstone points, the previous continuous mudstone interval is terminated, and a new continuous mudstone interval is divided starting from the next mudstone point. Only intervals containing at least two continuous mudstone points are determined as continuous mudstone intervals; isolated single mudstone points are not included in subsequent calculations as continuous mudstone intervals.

[0020] S6: Optimize the processing of mudstone acoustic transit time data

[0021] The program automatically identifies mudstone strata, calculates the arithmetic mean of the acoustic transit time (AC) data for each stratum, and correlates the arithmetic mean with the true vertical depth represented by the stratum. This reduces the clustering effect of thin interbedded strata physical property data, enhances the linear characteristics of strata compaction trends, and provides data support for subsequent basin simulation analyses such as strata erosion recovery.

[0022] The true vertical depth (TVD) of the top boundary, the true vertical depth (TVD) of the bottom boundary, and the representative true vertical depth (TVD) of each continuous mudstone section were determined. The arithmetic mean of the acoustic transit time (AC) data within each continuous mudstone section was calculated, and the arithmetic mean was corrected using the acoustic transit time of the mudstone skeleton. According to Δt c =Δt mean -Δt ma Perform acoustic transit time correction on mudstone skeletons, where Δt mean The arithmetic mean of the acoustic transit time AC for the continuous mudstone section is given by Δt. ma To pre-determine the acoustic transit time of the mudstone skeleton, Δt c The acoustic transit time correction value for the mudstone skeleton is obtained by subtracting the preset acoustic transit time of the mudstone skeleton from the arithmetic mean of the acoustic transit time AC of the continuous mudstone section; when Δt c When it is greater than zero, according to Y=ln(Δt) c / Δt ref Perform a dimensionless logarithmic transformation, where Δt ref For Δt c Preset reference values ​​with the same dimensions, Δt mean Δt ma and Δt ref Using the same unit of acoustic transit time, Y is a dimensionless parameter; by assigning a representative data point to a continuous mudstone interval, the clustering effect of high-density logging data is reduced, and key data for characterizing formation compaction trends are obtained.

[0023] S7: Output to Excel results worksheet

[0024] The discrimination direction and wide boundary threshold of the target lithology discrimination logging curves of each target well, the true vertical depth (TVD) of the top and bottom boundaries of the continuous mudstone interval, the representative true vertical depth (TVD), the arithmetic mean of the acoustic transit time (AC), the acoustic transit time correction value of the mudstone skeleton, and the dimensionless parameter Y of the acoustic transit time correction value of the mudstone skeleton are output to the Excel results worksheet to complete the screening of key data for single-well basin simulation.

[0025] Furthermore, the logging data used to characterize lithology are at least two of the following: spontaneous potential (SP), spontaneous gamma ray (GR), resistivity (RT), and logging clay content (SH).

[0026] Furthermore, outlier detection and removal are used to remove jump values, missing values, and data that exceed the preset effective range; outlier data segment removal during the start and end of logging instrument testing is used to remove data generated when the instrument's working state is unstable; data standardization or normalization is used to unify the numerical scale of different logging curves; curve smoothing filtering and noise removal are used to reduce the impact of random noise on the continuity of logging curves; when using standardization or normalization processing, mudstone sample data and full well section data use the same processing parameters.

[0027] Furthermore, the initial value of the preset mud content threshold is 50%, and it can be adjusted according to the geological conditions of the target study area.

[0028] Furthermore, the true vertical depth (TVD) represents the arithmetic mean of the true vertical depth (TVD) of each sampling point within the continuous mudstone section, and the arithmetic mean of the acoustic transit time (AC) represents the arithmetic mean of the acoustic transit time of each sampling point within the continuous mudstone section.

[0029] Furthermore, the results in Excel format output in step S7 are used for basin simulation analysis.

[0030] The beneficial effects of this invention: To solve the technical problems of this application, this invention provides an intelligent screening method for key basin simulation data based on a VBA program in Excel. It calibrates mudstone samples using at least one of the following: mud content, manual interpretation results, and available logging lithology labels. The VBA program statistically analyzes the differences in logging responses between mudstone samples and whole-well data, automatically determining the discrimination direction and wide boundary threshold for each logging curve. Finally, it identifies high-confidence mudstone data using an intersection judgment method that simultaneously satisfies multiple conditions, and divides continuous mudstone layers according to a 0.125 m sampling interval, thereby achieving standardized, automated, and batch screening of single-well mudstone data. Its main advantages are as follows:

[0031] (1) Adaptive statistics of single-well identification parameters: By utilizing the difference between known mudstone samples and the logging response of the whole well section, the discrimination direction and wide boundary threshold of no less than two target lithology discrimination logging curves are automatically determined, reducing the inter-well identification deviation caused by uniform fixed thresholds.

[0032] (2) Improve data processing efficiency by using VBA programs to automate mudstone sample statistics, full well section traversal, continuous layer division and parameter output, thereby reducing the workload of manual point-by-point interpretation and statistics.

[0033] (3) Taking into account both the reliability and practicality of identification, based on the wide range of single threshold values, the intersection judgment method that satisfies multiple conditions at the same time is adopted to screen high confidence mudstone data; at the same time, relying on the Excel platform, no additional professional geological software is required.

[0034] (4) Form a data optimization process for basin simulation compaction trend analysis, compress each continuous mudstone segment into a data point representing the true vertical depth (TVD) and the average value of acoustic transit time (AC), and perform skeleton acoustic transit time correction and dimensionless logarithmic transformation to reduce the data clustering effect and make the formation compaction trend easier to identify. Attached Figure Description

[0035] Figure 1 This is a flowchart of a method for intelligent screening of key data in basin simulation based on a VBA program.

[0036] Figure 2 A comparison chart of the original data, mudstone screening data, and optimized data of the mean of continuous mudstone intervals.

[0037] Figure 3 This is a schematic diagram for fitting the compaction trend of different strata and identifying erosion surfaces. Detailed Implementation

[0038] Example 1

[0039] First, a VBA program is launched in Excel to load the true vertical depth (TVD) of the target well and corresponding logging data such as spontaneous potential, natural gamma, resistivity, sonic transit time, and shale content. Outliers and abnormal data segments are removed, and the data is standardized or normalized, smoothed, and noise is removed. Second, known mudstone samples are calibrated using at least one of the following: shale content curves, manually interpreted results, and available logging lithology labels. The judgment results from different sources are verified to construct a benchmark dataset for mudstone sections. Subsequently, the VBA program iterates through the benchmark dataset, statistically analyzing the multidimensional logging response characteristics of mudstone samples. The mean of the mudstone samples is compared with the mean of the entire well data to determine the discrimination direction of each logging curve. An adaptive identification threshold matching the target well is formed using the wide boundary value of the mudstone samples. Using this identification threshold, the logging data of the entire well section is iterated. Only data points that simultaneously meet the discrimination conditions of each logging curve are marked as mudstone points. Adjacent mudstone points with a sampling interval of 0.125 m along the TVD direction and continuously marked are merged into a continuous mudstone section. Then, the representative true vertical depth (TVD) and the arithmetic mean of sonic transit time for each continuous mudstone interval are calculated. These average values ​​are then correlated with the corresponding representative TVD, and sonic transit time correction and dimensionless logarithmic transformation are performed on the mudstone skeleton to reduce the clustering effect of logging data under thin-layer and sandstone-mudstone interbedded conditions, obtaining key data characterizing the normal compaction trend of the formation. Finally, the mudstone identification threshold and discrimination direction, the range of continuous mudstone intervals, the representative TVD, the average sonic transit time, the correction value, and its logarithmic value are output to an Excel results worksheet.

[0040] The intelligent screening method for key basin simulation data based on VBA programs is applicable to complex conditions such as complex lithology, interbedded sandstone and mudstone, and well sections not covered by known lithology labels. Well logging curves typically employ a high-density sampling method of 0.125 m / point, with single-well sections reaching thousands of meters in length. Manually dividing mudstone sections and extracting data point by point is a significant workload. Furthermore, spontaneous potential (SP), natural gamma ray (GR), resistivity (RT), and mudstone content (SH) logging curves are relative measurement parameters. Different single wells may have different logging baselines, numerical ranges, and mudstone response directions. Using a uniform fixed threshold can easily lead to misidentification or omission. This method can adaptively determine identification parameters based on the logging baselines and numerical ranges of different single wells, reducing misidentification and omission caused by fixed thresholds. It achieves integrated processing of mudstone sample statistics, full-section mudstone identification, continuous section division, parameter calculation, and result output, providing a data foundation for formation compaction trend analysis, formation erosion recovery, and the recovery of burial history, thermal history, and hydrocarbon generation history.

[0041] The specific implementation method is as follows:

[0042] like Figure 1 As shown, the intelligent screening method for key basin simulation data based on VBA programs includes the following steps:

[0043] Step 1: In an Excel workbook, set up a full-well data worksheet, a mudstone sample worksheet, and a results worksheet, and start the VBA processing program. Import the target single-well logging data into the full-well data worksheet, including the true vertical depth (TVD), sonic transit time (AC), spontaneous potential (SP), natural gamma ray (GR), and mud content (SH) curves. Each row corresponds to a logging sampling point, and they are arranged in ascending order of true vertical depth (TVD) from shallowest to deepest. The true vertical depth interval between adjacent logging sampling points is 0.125 m. First, check if the data columns for true vertical depth, sonic transit time, and target lithology discrimination logging curves exist; if true vertical depth, sonic transit time, or fewer than two valid target lithology discrimination logging curves are missing, stop processing and prompt for supplementary data.

[0044] Step 2: Preprocess the loaded data. Using a VBA program, check the entire well data worksheet line by line, identifying and removing null values, non-numerical values, outliers exceeding the effective range of the corresponding logging curves, and abnormal data segments formed during the start and end phases of the logging instrument. Based on the data quality of the target well, perform standardization or normalization processing on logging curves of different dimensions. Perform smoothing filtering or noise removal processing on curves with significant random fluctuations. When using standardization or normalization, the mudstone sample data and the entire well data use the same processing parameters. After preprocessing, match the true vertical depth, sonic transit time, and target lithology discrimination logging curves line by line according to the same sampling points to form an effective logging dataset for mudstone sample statistics and mudstone identification throughout the well.

[0045] Step 3: Select mudstone as the target lithology to characterize the normal compaction trend of the formation. Based on the actual data available in the target well, calibrate known mudstone samples using one or more of the following: mudstone content curve (SH), manual interpretation results, and cuttings logging lithology labels. When the SH curve has valid data, logging sampling points with SH values ​​greater than the preset mudstone content threshold are identified as initial mudstone samples. Layers manually interpreted as mudstone and layers interpreted as mudstone by cuttings logging are mapped to logging sampling points within their corresponding true vertical depth ranges. Mudstone identification results from different sources are verified point by point. When the same logging sampling point is consistently identified as mudstone by multiple data sources, it is identified as a valid mudstone sample. When only one valid data source exists, the mudstone sample is determined according to the identification result of that source. When the identification results from different data sources contradict each other, the corresponding logging sampling point is removed from the mudstone sample. The true vertical depth, sonic transit time, and target lithology discrimination logging data of the calibrated mudstone samples are written into the mudstone sample worksheet, and mudstone sample identifiers are set to construct a standardized mudstone section benchmark dataset. In this embodiment, the initial value of the preset mud content threshold is 50%.

[0046] Step 4: Using a VBA program, traverse the benchmark dataset for the mudstone section and the effective logging dataset for the entire well section, calculating the arithmetic mean of the target lithology discrimination logging curves for each mudstone sample and the entire well section data. For any target lithology discrimination logging curve, if the mean value of the mudstone sample is greater than or equal to the mean value of the curve for the entire well section, it will be determined as the mudstone discrimination direction of the curve, and the minimum value of the curve in the mudstone sample will be determined as the wide boundary threshold; if the mean value of the mudstone sample is less than the mean value of the curve for the entire well section, it will be determined as the mudstone discrimination direction of the curve, and the maximum value of the curve in the mudstone sample will be determined as the wide boundary threshold. Determine the discrimination direction and wide boundary threshold for all target lithology discrimination logging curves one by one using the above method, and combine the curve name, discrimination direction, and wide boundary threshold into an adaptive identification parameter group adapted to the target single well.

[0047] Step 5: Perform mudstone identification and continuous mudstone interval division throughout the well. Using a VBA program, traverse the valid logging data line by line in ascending order of true vertical depth. For target lithology discrimination logging curves with a discrimination direction greater than or equal to the target lithology, determine if the curve value at the current sampling point is greater than or equal to the corresponding wide boundary threshold. For target lithology discrimination logging curves with a discrimination direction less than or equal to the target lithology, determine if the curve value at the current sampling point is less than or equal to the corresponding wide boundary threshold. Only when the current logging sampling point simultaneously satisfies the discrimination direction and wide boundary threshold of all target lithology discrimination logging curves is the mudstone identifier of that sampling point set to 1; otherwise, its mudstone identifier is set to 0. Read the mudstone identifiers sequentially in ascending order of true vertical depth. When the current mudstone point and the previous mudstone point are adjacent sampling points, and the difference in their true vertical depth is 0.125 m, the two mudstone points are classified into the same continuous mudstone interval. When the difference in true vertical depth between adjacent mudstone points is greater than 0.125 m, or when there is a sampling point with a mudstone identifier of 0 between the two mudstone points, the previous continuous mudstone interval is terminated, and a new continuous mudstone interval is established starting from the next mudstone point. Only intervals containing at least two continuous mudstone points are identified as continuous mudstone intervals; isolated single mudstone points are not included in subsequent calculations as continuous mudstone intervals.

[0048] Step Six: Optimize the acoustic transit time data of continuous mudstone sections. For each continuous mudstone section obtained in Step Five, determine the true vertical depth of the first mudstone point in the section as the top true vertical depth, and the true vertical depth of the last mudstone point as the bottom true vertical depth. Calculate the arithmetic mean of the true vertical depths of all mudstone points within the section, and use this as the representative true vertical depth. Calculate the arithmetic mean of the acoustic transit time data of all mudstone points within the continuous mudstone section, and assign one representative true vertical depth and one average acoustic transit time data point to each continuous mudstone section. According to Δt... c =Δt mean -Δt ma The average acoustic transit time of the layer is corrected for the acoustic transit time of the mudstone skeleton, where Δt mean The arithmetic mean of the acoustic transit time AC for the continuous mudstone section is given by Δt. ma To pre-determine the acoustic transit time of the mudstone skeleton, Δt c The acoustic transit time correction value for the mudstone skeleton is obtained by subtracting the preset acoustic transit time of the mudstone skeleton from the arithmetic mean of the acoustic transit time AC of the continuous mudstone layer. When Δt c When it is greater than zero, according to Y = ln(Δt) c / Δt ref Perform a dimensionless logarithmic transformation, where Δt ref For Δt c Preset reference values ​​with the same dimensions; Δt mean Δt ma and Δt refUsing the same unit of acoustic time difference, Y is a dimensionless parameter; when Δt c When the value is less than or equal to zero, its dimensionless logarithmic value is not calculated, and the data point is marked as invalid correction data. By processing the mean of continuous mudstone intervals, correcting the sonic transit time of the mudstone skeleton, and performing dimensionless logarithmic transformation, the clustering effect of high-density logging data is reduced, and key basin simulation data for characterizing the normal compaction trend of the formation is obtained.

[0049] Step 7: Output key basin simulation data. Write the target well name, the name of each target lithology discrimination logging curve, discrimination direction and wide boundary threshold, and the top and bottom true vertical depths, representative true vertical depths, arithmetic mean of sonic transit time, sonic transit time correction values ​​for the mudstone skeleton, and dimensionless parameters of the sonic transit time correction values ​​for the mudstone skeleton, in ascending order of true vertical depth, into an Excel results worksheet. Simultaneously retain the mudstone identifier in the full well data worksheet for mudstone identification result verification. After completing the results writing, end the program. The obtained results are used for formation compaction trend fitting, formation erosion recovery, and analysis of basin burial history, thermal history, and hydrocarbon generation history.

[0050] like Figure 2 As shown, the left figure is the original AC-TVD relationship diagram of a single well formation; the middle figure is the AC-TVD relationship diagram after mudstone screening using an adaptive threshold and multi-condition intersection judgment method; and the right figure is the AC-TVD relationship diagram after further mean processing according to continuous mudstone intervals. Each continuous mudstone interval is represented by the arithmetic mean of its representative true vertical depth (TVD) and sonic transit time (AC). The number of data points is significantly reduced, and the compaction trend is more concentrated. To approximate the degree of optimization shown in the diagram, Figure 2 The absolute value of the correlation coefficient |R| of the original data is approximately 0.18, which is approximately 0.23 after mudstone screening, and approximately 0.65 after optimization of the mean of continuous mudstone intervals; the corresponding coefficient of determination R² increases from approximately 0.03 to 0.42. In practical engineering applications, the regression calculation results of the original numerical data should be used as the standard.

[0051] like Figure 3 As shown, the key data obtained in step six are fitted with compaction trends according to different strata to obtain the corresponding compaction trend lines. After screening and layer averaging, the data points are distributed along their respective trend lines. The compaction trends of adjacent strata show an identifiable discontinuity at the erosion surface, which can help determine the location of the erosion surface and provide data basis for the recovery of stratum erosion.

[0052] This embodiment utilizes the Excel platform and its VBA functions to complete mudstone sample statistics, single-well threshold and discrimination direction extraction, mudstone identification across the entire well section, continuous layer division, and key parameter calculation. Mud content curves or manually interpreted results can be used independently to establish mudstone samples; when logging lithology labels are available, they can be used as supplementary sample sources. Since each individual threshold uses wide boundary values ​​from the mudstone samples, the final identification employs an intersection rule that satisfies multiple conditions simultaneously to improve the confidence level of the screened mudstone data. This method can adapt to differences in logging baselines, numerical ranges, and response directions across different single wells, and improves the automation and consistency of mudstone data screening under high-density sampling and frequent sand-mudstone interbedded conditions.

Claims

1. A method for intelligent screening of key data in basin simulation based on VBA program, characterized in that: The intelligent screening method includes the following steps: S1: Start the program and load raw logging data Launch the VBA program on the Excel platform and load the raw logging data of the target single well, sorted in ascending order by true vertical depth (TVD). The raw logging data includes true vertical depth (TVD), sonic transit time (AC), and at least two target lithology discrimination logging curves. Select spontaneous potential (SP) curve, natural gamma ray (GR) curve, resistivity (RT) curve, and logging clay content (SH) curve as candidate target lithology discrimination logging curves. The true vertical depth (TVD) interval between adjacent sampling points is 0.125 m. S2: Well logging data preprocessing The raw logging data after loading is preprocessed, including outlier detection and removal, removal of outlier data segments during the start and end of logging instrument testing, data standardization or normalization, curve smoothing filtering, and noise removal. S3: Constructing a benchmark dataset for mudstone sections Mudstone was selected as the target lithology to characterize the normal compaction trend of the formation. Known mudstone samples were calibrated using at least one of the following: well logging shale content (SH) curves, manual interpretation results, and cuttings logging lithology labels. When the well logging shale content (SH) curves had valid data, well logging points with SH curves greater than the preset shale content threshold were used as the initial mudstone samples, and supplemented or verified using manual interpretation results or cuttings logging lithology labels. When the judgment results from different data sources contradicted each other, the corresponding samples were removed. The true vertical depth (TVD), sonic transit time (AC), and target lithology discrimination well logging curve data corresponding to the verified mudstone samples were collected to construct a standardized mudstone section benchmark dataset, providing standard sample support for the adaptive determination of subsequent single-well discrimination direction and wide boundary threshold. S4: Statistical analysis of mudstone samples and extraction of adaptive thresholds The VBA program iterates through the benchmark dataset for mudstone sections. Based on the actual logging curve types and data validity of the target wells, if both the spontaneous potential (SP) curve and the natural gamma ray (GR) curve are valid data from the spontaneous potential (SP) curve, natural gamma ray (GR) curve, resistivity (RT) curve, and mud content (SH) curve, then the SP and GR curves are selected as the target lithology discrimination logging curves. Otherwise, at least two curves with valid data are selected from the four candidate target lithology discrimination logging curves as the target lithology discrimination logging curves. Well logging curves: Calculate the mean value of each target lithology discrimination logging curve in the mudstone sample and the entire well section data respectively, and determine the mudstone discrimination direction and wide boundary threshold for each curve. When the mean value of the mudstone sample is greater than or equal to the mean value of the entire well section, it is determined as the discrimination direction, and the minimum value of the mudstone sample is determined as the wide boundary threshold of the logging curve. When the mean value of the mudstone sample is less than the mean value of the entire well section, it is determined as the discrimination direction, and the maximum value of the mudstone sample is determined as the wide boundary threshold of the logging curve, thus forming adaptive identification parameters adapted to the target single well. S5: Mudstone Identification and Continuous Stratum Division Throughout the Well Section The VBA program iterates through the logging data of the entire well section, determining whether each target lithology discrimination logging curve selected in step S4 meets the corresponding wide boundary threshold and discrimination direction. Only when the logging data point to be judged simultaneously meets the wide boundary threshold and corresponding discrimination direction of all selected target lithology discrimination logging curves is the logging data point marked as a mudstone point. For logging data arranged in ascending order of true vertical depth (TVD), when the current mudstone point and the previous mudstone point are adjacent sampling points and the difference in their true vertical depth (TVD) is 0.125 m, the two mudstone points are classified into the same continuous mudstone interval. When the difference in the true vertical depth (TVD) of adjacent mudstone points is greater than 0.125 m, or when there are non-mudstone sampling points between the two mudstone points, the previous continuous mudstone interval is terminated, and a new continuous mudstone interval is divided starting from the next mudstone point. Only intervals containing at least two continuous mudstone points are determined as continuous mudstone intervals; isolated single mudstone points are not included in subsequent calculations as continuous mudstone intervals. S6: Optimize the processing of mudstone acoustic transit time data The true vertical depth (TVD) of the top boundary, the true vertical depth (TVD) of the bottom boundary, and the representative true vertical depth (TVD) of each continuous mudstone section were determined. The arithmetic mean of the acoustic transit time (AC) data within each continuous mudstone section was calculated, and the arithmetic mean was corrected using the acoustic transit time of the mudstone skeleton. According to Δt c =Δt mean -Δt ma Perform acoustic transit time correction on mudstone skeletons, where Δt c The acoustic transit time correction value for the mudstone skeleton is Δt. mean The arithmetic mean of the acoustic transit time AC for the continuous mudstone section is given by Δt. ma To preset the acoustic transit time of the mudstone skeleton; when Δtc is greater than zero, according to Y=ln(Δt c / Δt ref Perform a dimensionless logarithmic transformation, where Δt ref For Δt c Preset reference values ​​with the same dimensions, Δt mean , Δt ma and Δt ref Using the same unit of acoustic transit time, Y is a dimensionless parameter; by assigning a representative data point to a continuous mudstone interval, the clustering effect of high-density logging data is reduced, and key data for characterizing formation compaction trends are obtained. S7: Output to Excel results worksheet The discrimination direction and wide boundary threshold of the target lithology discrimination logging curves of each target well, the true vertical depth (TVD) of the top and bottom boundaries of the continuous mudstone interval, the representative true vertical depth (TVD), the arithmetic mean of the acoustic transit time (AC), the acoustic transit time correction value of the mudstone skeleton, and the dimensionless parameter Y of the acoustic transit time correction value of the mudstone skeleton are output to the Excel results worksheet to complete the screening of key data for single-well basin simulation.

2. The intelligent screening method for key basin simulation data based on VBA program according to claim 1, characterized in that: Outlier detection and removal are used to remove jump values, missing values, and data that exceed the preset effective range; outlier data segment removal during the start and end of logging instrument testing is used to remove data generated when the instrument's working state is unstable; data standardization or normalization is used to unify the numerical scale of different logging curves; curve smoothing filtering and noise removal are used to reduce the impact of random noise on the continuity of logging curves; when using standardization or normalization processing, mudstone sample data and full well section data use the same processing parameters.

3. The intelligent screening method for key basin simulation data based on VBA program according to claim 1, characterized in that: The initial value of the preset mud content threshold is 50%, and it can be adjusted according to the geological conditions of the target study area.

4. The intelligent screening method for key basin simulation data based on VBA program according to claim 1, characterized in that: The true vertical depth (TVD) represents the arithmetic mean of the true vertical depth (TVD) of each sampling point within the continuous mudstone section, while the arithmetic mean of the acoustic transit time (AC) represents the arithmetic mean of the acoustic transit time (AC) of each sampling point within the continuous mudstone section.

5. The intelligent screening method for key basin simulation data based on VBA program according to claim 1, characterized in that: The results in Excel format output in step S7 are used for basin simulation analysis.

Images