Method for correcting logging sound wave velocity to VSP velocity trend based on automatic segmentation and application

Abstract

The invention provides a method for correcting logging sound wave velocity to a VSP velocity trend based on automatic segmentation and application, and relates to the technical field of oil-gas exploration. The method comprises the following steps: collecting logging sound wave velocity data, VSP data and geological stratification data, and carrying out preprocessing and calibration; analyzing the obtained VSP data, identifying a main speed change horizon or speed trend, and constructing a change rate of the VSP speed; according to the VSP speed trend and the stratum characteristics, logging sound wave speed data are automatically segmented through an automatic algorithm; in each segment, a mathematical function is used for fitting the logging sound wave velocity, and a correction factor or a correction curve is obtained; correcting logging sound wave speed, quality control and result output. According to the method, the exploration efficiency and accuracy are improved, the cost is reduced, the interpretability of data is enhanced, and powerful support is provided for oil-gas exploration and development.

Description

Technical Field

[0001] This invention relates to the field of oil and gas exploration technology, specifically to a method and application for correcting logging acoustic velocity to VSP velocity trend based on automatic segmentation. Background Technology

[0002] The Tarim Basin, located in Xinjiang, northwest China, is China's largest inland basin and one of the world's richest regions in oil and gas resources. While the basin is rich in oil and gas resources, exploration and development are extremely challenging, primarily due to the complex geological conditions and the prevalence of ultra-deep, ultra-high-temperature, and ultra-high-pressure oil and gas reservoirs.

[0003] Well logging is a downhole oil and gas exploration method used to discover oil and gas reservoirs and assess their reserves and production. It also has wide applications in oil and gas field development and drilling engineering. Well logging technology is an indispensable means of accurately discovering oil and gas layers and accurately describing oil and gas reservoirs. It is an important scientific basis for calculating oil and gas reserve parameters, assessing production capacity, and formulating and adjusting development plans.

[0004] There are numerous well logging methods, with electrical, acoustic, and radiometric methods being three basic ones. Acoustic velocity logging, based on the propagation speed of sound waves in the formation, measures the time difference Δt (the reciprocal of the formation's longitudinal wave velocity) of the slip wave to calculate and determine formation porosity, lithology, and pore fluid properties. It is the most commonly used well logging method (e.g., Chinese patents CN102866435A and CN1621860A). Vertical seismic profiling (VSP) technology achieves subsurface mapping through surface excitation and well reception. Because VSP technology offers more direct reservoir characterization (absorption attenuation, velocity, anisotropy, etc.), better time-depth calibration capabilities, and higher resolution than surface seismic logging, it is widely used for finely characterizing complex wellbore structures, reservoir and fluid spatial distribution. It serves as an indispensable bridge between surface seismic, well logging, and geological information, improving the imaging accuracy of reservoir targets. However, due to economic considerations, some exploration and appraisal wells have only undergone acoustic logging without VSP logging.

[0005] VSP (Volatile Seismic Flow Perimeter) observations ensure that seismic waves pass through the low-velocity zone at the surface only once, reducing energy attenuation, particularly the loss of high-frequency components. The geophone is placed along the wellbore, receiving seismic waves within the formation. This allows for more accurate positioning at the target depth, improving the precision of velocity analysis. The closer the VSP geophone is to the target layer, the less interference the received seismic signal experiences, resulting in less amplitude distortion and more reliable data. VSP and acoustic logging trends exhibit a high degree of similarity. Well-controlled anisotropic seismic data processing requires more well data, especially VSP velocities, to correct traditional acoustic logging velocities to VSP velocity trends, thus better achieving control points in well-controlled anisotropic seismic data processing.

[0006] In the field of oil exploration and development, correcting well logging sonic velocities to VSP (Vertical Seismic Profile) velocity trends is a crucial technical step. Traditional correction methods mainly rely on manual segmentation and adjustment, which is not only inefficient but also susceptible to human factors, making it difficult to guarantee the accuracy and reliability of the correction results. In recent years, with the continuous development of computer technology, automated correction methods such as neural networks and curve fitting have been proposed and applied. However, these methods still have certain limitations when dealing with complex geological conditions.

[0007] Therefore, developing a new logging acoustic velocity correction technology is of great significance for improving the efficiency and accuracy of oil exploration and development in the Tarim Basin. Summary of the Invention

[0008] To address the aforementioned problems, this invention provides a method for correcting logging acoustic velocity to VSP velocity trends based on automatic segmentation. By introducing an advanced automatic segmentation algorithm, it enables automatic analysis and processing of logging acoustic data, thereby accurately identifying the characteristics of different formations and lithologies.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0010] On one hand, the present invention provides a method for correcting logging acoustic velocity to VSP velocity trend based on automatic segmentation, comprising the following steps:

[0011] Step 1: Data Preparation: Collect well logging sonic velocity data, VSP data, and geological stratification data;

[0012] Step 2: Data Preprocessing: Preprocess the logging sonic velocity data and VSP data, and then perform calibration;

[0013] Step 3: Analyze VSP velocity trends: Analyze the VSP data obtained in Step 2, identify the main velocity change layers or velocity trends, and construct the VSP velocity change rate;

[0014] Step 4: Automatic segmentation: Based on the VSP velocity trend and formation characteristics, the logging sonic velocity data is automatically segmented using an automatic algorithm;

[0015] Step 5: Segmented Fitting: Within each segment, use mathematical functions to fit the logging sonic velocity to obtain a correction factor or correction curve;

[0016] Step 6: Correct the logging sonic velocity: Correct the original logging sonic velocity according to the correction factor or correction curve obtained from the fitting.

[0017] Step 7: Quality Control: Perform a quality check on the corrected logging sonic velocity to ensure that no unreasonable velocity changes or abnormal values ​​were introduced during the correction process;

[0018] Step 8: Output Results: Output the corrected data.

[0019] Preferably, in step 1, the logging acoustic velocity data includes the depth of the target well and adjacent wells, and the acoustic logging velocity.

[0020] Preferably, in step 1, the VSP data includes the depths of the target well and adjacent wells, and the VSP logging rate;

[0021] Preferably, in step 1, the geological stratification data includes well depth and stratigraphic information;

[0022] Preferably, in step 2, the preprocessing includes at least one of cleaning, filtering, denoising, and removing outliers, missing values, or erroneous data; more preferably, the preprocessing includes at least one of cleaning and removing outliers, missing values, or erroneous data.

[0023] Preferably, in step 2, the calibration includes depth and / or time calibration;

[0024] Preferably, step 3 specifically includes the following steps:

[0025] 3-1. Data Preparation: Ensure the accuracy and completeness of the data, and preprocess missing or abnormal data;

[0026] 3-2. Velocity variation analysis with depth: Identify the layers or intervals where the velocity changes significantly with depth;

[0027] 3-3. Identification of major velocity change strata or trends: Combine geological background data to analyze the geological significance of velocity change strata or trends;

[0028] 3-4. Outlier Check: Check for outliers in the speed data, and mark and record outliers.

[0029] 3-5. Calculation of velocity parameters: Calculate the average velocity, layer velocity, and root mean square velocity;

[0030] 3-6. Handling of outlier speed parameters: Check the obtained speed parameters for outliers, remove outliers and / or recalculate them;

[0031] 3-7. VSP Solution Speed ​​Selection: After outlier handling, the average speed is selected as the VSP solution speed for the next step.

[0032] Preferably, in steps 3-5, the formula for calculating the average speed is:

[0033]

[0034] in,

[0035] V a The average velocity represents the average propagation speed of sound waves within a specific well section.

[0036] Hd i The depth of the observation point (measured from the ground surface);

[0037] T i / 2 Vertical one-way travel time (from the ground surface).

[0038] Preferably, in steps 3-5, the formula for calculating the layer velocity is:

[0039]

[0040] in,

[0041] V ni For layer velocity;

[0042] Hd i The depth of the observation point (measured from the ground surface);

[0043] T i / 2 Vertical one-way time (from the ground surface);

[0044] Preferably, in steps 3-5, the root mean square velocity calculation formula is as follows:

[0045]

[0046] in,

[0047] It is the root mean square velocity;

[0048] T i It is the time value of the i-th measurement point;

[0049] V i It is the velocity value at the i-th measurement point;

[0050] n is the total number of measurement points.

[0051] Preferably, in step 4, the automatic algorithm includes at least one of cluster analysis, dynamic programming, and sliding window;

[0052] Preferably, step 4 specifically includes the following steps:

[0053] 4-1. Data Preparation:

[0054] Import the database and load the well logging sonic velocity data (y_data) and the corresponding depth or time data (x_data);

[0055] 4-2. Threshold setting:

[0056] Set an initial threshold to identify significant changes in the sound wave velocity data;

[0057] 4-3. Automatic segmentation algorithm:

[0058] Define a function find_breakpoints that takes the sound velocity data (y), the threshold, and the minimum length (min_length) as input;

[0059] Inside the function, the difference (diff) between adjacent velocity values ​​is calculated;

[0060] Indices whose absolute difference is greater than a threshold correspond to possible segmentation points;

[0061] Iterate through all possible segmentation points. If there is already a segmentation point before the current segmentation point or if the number of data points between two points is greater than the minimum length (min_length), then add the current segmentation point to the segmentation point list.

[0062] Ensure that the last segment point is set correctly;

[0063] If the distance between the last segment point and the end of the data is greater than the minimum length minus one, then the end of the data is taken as the segment point;

[0064] 4-4. Segmentation point filtering:

[0065] Initialize a new list (filtered_segments) to store the filtered segment points; the first point is always retained.

[0066] Iterate through the list of segment points and check if the difference between the current point and the previous point is greater than or equal to a preset value. If the condition is met and the depth of the current point is less than the specified value, add the current point to the filtered list of segment points. For points with a depth greater than or equal to the specified value, add them to the list unconditionally.

[0067] 4-5. Obtain the actual segmentation point locations:

[0068] Using the filtered segment point index, obtain the corresponding actual segment point position from x_data (depth or time data);

[0069] 4-6. Output:

[0070] Output the filtered segment point positions.

[0071] Preferably, step 5 specifically includes the following steps:

[0072] 5-1. Data Segmentation:

[0073] Based on the segmentation points obtained in step 4, the np.split function is used to divide x_data (depth or time data) and y_data (sonic velocity data) into multiple subsequences (segments_x and segments_y).

[0074] 5-2. Define the fitting function:

[0075] Choose an appropriate fitting function to describe the relationship between sound wave velocity and depth or time;

[0076] 5-3. Piecewise Fitting:

[0077] Use the curve_fit function to fit each segment and obtain the fitting parameters;

[0078] 5-4. Output and Visualization:

[0079] After fitting each segment, output the fitting parameters;

[0080] Visualization techniques are used to intuitively display the original data and fitting results.

[0081] Preferably, step 6 specifically includes the following steps:

[0082] 6-1. Initialization: Initialize an array corrected_y of the same size as the original y data to store the corrected acoustic velocity values;

[0083] 6-2. Traverse segments: For each segment (defined by breakpoints), determine the range of x values ​​corresponding to that segment and find the corresponding y values;

[0084] 6-3. Apply fitting parameters: Use the fitting parameters of the current segment (obtained from fitted_params) to calculate the corrected y value through the fitting function;

[0085] 6-4. Store the correction value: Store the calculated corrected y value in the corresponding position in the corrected_y array;

[0086] 6-5. Handling Boundary Cases: Ensure proper handling at segment boundaries and the start / end points of data to avoid data loss or duplication;

[0087] 6-6: Verify the correction effect: Compare and analyze the corrected acoustic velocity data with the VSP velocity data to evaluate the accuracy and effectiveness of the correction method.

[0088] Preferably, the quality check in step 7 includes at least one of data integrity check, fit quality assessment, parameter rationality check, and boundary condition processing.

[0089] Preferably, step 7 specifically includes the following steps:

[0090] 7-1. Data Integrity Check:

[0091] Check the integrity of the raw logging sonic velocity data (y_data) and depth or time data (x_data) to ensure there are no missing or outliers; verify the validity of the breakpoints.

[0092] 7-2. Fit quality assessment:

[0093] For each segment, the quality of the fitting result can be evaluated by checking the fitting residuals (i.e., the difference between the original data and the fitted curve).

[0094] 7-3. Parameter rationality check:

[0095] Check the reasonableness of the fitted parameters (fitted_params);

[0096] If unreasonable parameter values ​​are found, further examination of the data and fitting process is needed to determine the cause of the problem.

[0097] 7-4. Boundary Condition Handling:

[0098] Ensure proper handling at segment boundaries and data start / end points to avoid data loss or duplication; ensure the continuity of correction results.

[0099] Preferably, the output in step 8 includes corrected data, fitting parameters, quality control report, and visualization data.

[0100] Preferably, step 8 specifically includes the following steps:

[0101] 8-1. Output of corrected data:

[0102] Output the corrected logging sonic velocity data (corrected_y) in a suitable format;

[0103] The output file should include the index or depth / time value of the original data for correspondence and comparison with the original data;

[0104] 8-2. Fitting parameter output:

[0105] Output the fitted parameters (fitted_params) for each segment for subsequent analysis and validation;

[0106] 8-3. Quality Control Report:

[0107] Generate a quality control report summarizing the results of data integrity checks, fit quality assessments, parameter rationality checks, and boundary condition handling.

[0108] 8-4. Visualization Output:

[0109] You can choose to use charts, images, and other visualization methods to display the data and fitted curves before and after correction, so that users can understand the correction effect more intuitively.

[0110] Preferably, the visualization output may include at least one of the following: a scatter plot of the original data, a fitted curve plot, and a residual plot.

[0111] Furthermore, this invention provides the application of the above-described methods in oil and gas exploration and development, geophysical data processing and analysis, and mineral resource exploration in the Shunbei area of ​​the Tarim Basin.

[0112] Compared with the prior art, the present invention has the following beneficial effects:

[0113] This invention proposes a method for correcting logging acoustic velocity to VSP velocity trend based on automatic segmentation, which shows significant advantages in the field of oil and gas exploration.

[0114] First, by introducing automatic segmentation and correction algorithms, this invention enables rapid processing and analysis of well logging acoustic velocity data, greatly improving correction efficiency and reducing the manpower and time required by traditional manual correction methods, thereby enhancing the overall efficiency of exploration work.

[0115] Secondly, the automatic segmentation algorithm can objectively and accurately identify changes in formation or geological features, avoiding interference from human factors and thus ensuring the accuracy and reliability of the correction results. The correction algorithm fully considers changes in formation or geological features, enabling precise correction of well logging sonic velocity data, providing more reliable geological data support for oil and gas exploration and development.

[0116] Furthermore, this invention not only improves calibration efficiency and accuracy but also reduces exploration costs. By reducing the need for manpower and time, it lowers resource consumption during the exploration process. Simultaneously, accurate calibration results reduce exploration risks, increase exploration success rates, further reduce VSP project deployment, and achieve effective cost control.

[0117] Furthermore, the corrected well logging sonic velocity data, along with key parameters and quality assessment information from the correction process, provide important reference value for subsequent geological interpretation and oil and gas prediction. Comprehensive analysis of this information helps to more accurately understand subsurface geological structures and improve the targeting and effectiveness of exploration work.

[0118] In summary, the automatic segmentation-based logging sonic velocity correction method of this invention not only improves exploration efficiency and accuracy and reduces costs, but also enhances data interpretability, providing strong support for oil and gas exploration and development. Its strong adaptability and flexibility make it suitable for various types of logging sonic velocity data and geological conditions, bringing great convenience and benefits to exploration work. Attached Figure Description

[0119] Figure 1 This is a flowchart illustrating the specific implementation of Example 1.

[0120] Figure 2 This is a graph comparing the corrected value with the actual value. Detailed Implementation

[0121] To make the technical means, creative features, achieved objectives, and effects of this invention readily understandable, the invention is further illustrated below with specific embodiments. However, these embodiments are merely preferred embodiments and not all embodiments. Other embodiments obtained by those skilled in the art based on the embodiments described herein without creative effort are all within the scope of protection of this invention. It is worth noting that the raw materials used in this invention are all common commercially available products, and their sources are not specifically limited. The technical and scientific terms used in the embodiments have the meanings commonly understood by those skilled in the art to which this invention pertains.

[0122] Example 1

[0123] Step 1: Data Preparation

[0124] For the target conversion well, data from wells and formations with similar geological structures were collected to ensure coverage of the target well and its adjacent wells, including depth, sonic logging velocity, VSP logging velocity, and formation stratification information. The data collected in this study covers the Shunbei No. 4 fault zone (41X, 42X, 43X, 44X, 45X, 46X, and 47X) in the Shunbei Oil and Gas Field. The original sonic and VSP logging data are shown in Tables 1 and 2.

[0125] Table 1. Original sonic logging data of a well in Shunbei No. 4 belt

[0126]

[0127]

[0128] Table 2. Original VSP Time-Depth Relationship and Velocity Table of a Well in Shunbei No. 4 Belt

[0129]

[0130]

[0131]

[0132] Step 2: Data Preprocessing

[0133] Data is cleaned to remove outliers, missing values, or erroneous data to ensure data quality.

[0134] Perform necessary depth or time calibration on the data to ensure comparisons are made within the same formation depth or time range. Based on well location information, match and align the logging data with the VSP data to ensure a corresponding depth. Generally, sonic logging has a higher sampling density frequency, with ten sampling points per meter, while VSP uses one sampling point every ten meters or twenty meters, as shown in Table 3.

[0135] Table 3. Comparison of acoustic logging velocity and VSP velocity with time and depth

[0136]

[0137]

[0138] Step 3: Analyze VSP speed trend

[0139] Analyze the changes in velocity parameters with depth to identify the main velocity-changing strata or trends. These strata or trends may correspond to geological structures, lithological changes, etc. Compare VSP logging strata, surface depth, base level depth, surface bimodal time, and stratum velocity to check for any anomalies.

[0140] 3-1. Data Preparation:

[0141] Collect and organize VSP logging data, including depth information, time information, and corresponding velocity data.

[0142] Ensure the accuracy and integrity of the data, and preprocess missing or abnormal data.

[0143] 3-2. Analysis of velocity variation with depth:

[0144] Using depth as the independent variable and velocity as the dependent variable, plot a curve showing the change in velocity with depth.

[0145] Carefully observe the graph to identify the layers or intervals where significant velocity changes occur. These changes may manifest as sudden increases or decreases in velocity, or significant shifts in the velocity gradient.

[0146] 3-3. Identification of key velocity change layers or trends:

[0147] Analyze the geological significance of velocity variations, including their stratigraphic levels or trends, in conjunction with background geological data. These variations may correspond to geological features such as stratigraphic interfaces, faults, and lithological changes.

[0148] These key velocity variation layers or intervals are marked to provide a basis for subsequent geological interpretation.

[0149] 3-4. Outlier Check:

[0150] By comparing information such as VSP logging layers, surface depth, reference level depth, and surface two-way travel time, check for outliers in the velocity data. Outliers may be caused by measurement errors, data processing errors, or geological reasons.

[0151] Outliers are marked and recorded for further analysis and processing.

[0152] 3-5. Speed ​​parameter calculation:

[0153] Calculate velocity parameters such as average velocity, layer velocity, and root mean square velocity. These parameters can provide quantitative information about the distribution and variation of formation velocities.

[0154] Formula for calculating average speed:

[0155]

[0156] in,

[0157] V a The average velocity represents the average propagation speed of sound waves within a specific well section.

[0158] Hd i The depth of the observation point (measured from the ground surface);

[0159] T i / 2 Vertical one-way travel time (from the ground surface).

[0160] Preferably, in steps 3-5, the formula for calculating the layer velocity is:

[0161]

[0162] in,

[0163] V ni For layer velocity;

[0164] Hd i The depth of the observation point (measured from the ground surface);

[0165] T i / 2 Vertical one-way time (from the ground surface);

[0166] Preferably, in steps 3-5, the root mean square velocity calculation formula is as follows:

[0167]

[0168] in,

[0169] It is the root mean square velocity;

[0170] T i It is the time value of the i-th measurement point;

[0171] V i It is the velocity value at the i-th measurement point;

[0172] n is the total number of measurement points.

[0173] 3-6. Outlier Handling:

[0174] Outlier checks are performed on the calculated average velocity, layer velocity, and root mean square velocity. If outliers are found, further analysis and processing are required, such as outlier removal and parameter recalculation.

[0175] 3-7. VSP Solving Speed ​​Selection:

[0176] After outlier processing, the average velocity was selected as the velocity for the next step of VSP calculation. The average velocity can well reflect the overall distribution and variation trend of formation velocities, and is suitable for subsequent seismic data processing and interpretation.

[0177] Step 4: Automatic Segmentation

[0178] The data is segmented using an automatic segmentation algorithm. A threshold is set based on the rate of change of VSP velocity trends. The default threshold automatically divides the logging sonic data into different segments according to the velocity data variation characteristics. These segments should have similar velocity characteristics.

[0179] 4-1. Data Preparation:

[0180] Import necessary libraries, such as NumPy, for numerical computation.

[0181] Load the logging sonic velocity data (y_data) and the corresponding depth or time data (x_data).

[0182] 4-2. Threshold setting:

[0183] Set an initial threshold to identify significant changes in the sound velocity data. This threshold can be set according to actual needs through user input (e.g., from a web form request.form['threshold']) or a preset value.

[0184] If the user does not provide a threshold, the default value (such as 30) will be used.

[0185] 4-3. Automatic segmentation algorithm:

[0186] Define a function find_breakpoints that takes sound velocity data (y), threshold (threshold), and minimum length (min_length) as input.

[0187] Inside the function, the difference (diff) between adjacent velocity values ​​is calculated first.

[0188] Find the indices where the absolute value of the difference is greater than the threshold; these indices correspond to possible segmentation points.

[0189] Iterate through all possible segmentation points, checking if there is already a segmentation point before the current segmentation point, and whether the number of data points between the two points is greater than the minimum length (min_length). If the conditions are met, add the current segmentation point to the segmentation point list.

[0190] Ensure the last segment point is set correctly. If the distance between the last segment point and the end of the data is greater than the minimum length minus one, then use the end of the data as the segment point.

[0191] 4-4. Segmentation point filtering:

[0192] Initialize a new list (filtered_segments) to store the filtered segment points. The first point is always retained.

[0193] Iterate through the list of segment points (starting from the second point), checking if the difference between the current point and the previous point is greater than or equal to a preset value (e.g., 5). If the condition is met and the current point's depth is less than 4000 (or other specified value), add the current point to the filtered list of segment points. For points with a depth greater than or equal to 4000, add them to the list unconditionally.

[0194] 4-5. Obtain the actual segmentation point locations:

[0195] Using the filtered segment point index, retrieve the corresponding actual segment point position from x_data (depth or time data).

[0196] 4-6. Output:

[0197] The output shows the locations of the filtered segment points, which will be used in the subsequent process of correcting the logging acoustic velocity to the VSP velocity trend.

[0198] Step 5: Piecewise Fitting

[0199] After completing the automatic segmentation, we will next perform fitting on each segment to correct the logging sonic velocity and predict or adjust it to the VSP velocity trend. The following is the specific implementation details of the segment fitting section:

[0200] 5-1. Data Segmentation:

[0201] Based on the segmentation points obtained from the automatic segmentation algorithm above, the np.split function is used to divide x_data (depth or time data) and y_data (sonic velocity data) into multiple subsequences (segments_x and segments_y).

[0202] 5-2. Define the fitting function:

[0203] Choose an appropriate fitting function to describe the relationship between sound wave velocity and depth or time. In this example, we define a simple linear function `fit_func(x,a,b)`, where `a` is the slope and `b` is the intercept. However, depending on the specific application and the characteristics of the data, more complex functions (such as polynomials, exponential functions, etc.) can be chosen.

[0204] 5-3. Piecewise Fitting:

[0205] Iterate through seg_x and seg_y for each subsequence (segment).

[0206] Use the curve_fit function (from the SciPy library) to fit each segment and obtain the fitting parameters (such as the slope and intercept of the linear function).

[0207] If the length of a segment is less than or equal to 2 (i.e., there are not enough data points to fit), skip that segment and output the corresponding prompt message.

[0208] The fitting parameters for each segment are stored in the fitted_params list for easy subsequent analysis and application.

[0209] 5-4. Output and Visualization:

[0210] After fitting each segment, output the fitting parameters (such as slope and intercept) for subsequent analysis and validation.

[0211] Visualization techniques (such as scatter plots and fitted curves) are used to visually represent the original data and the fitting results, enabling a better understanding of the data characteristics and the fitting effect.

[0212] Step 6: Correct the acoustic logging velocity

[0213] Based on the fitting parameters for each segment, the trend of sonic logging velocity to VSP velocity can be predicted or adjusted. This can be achieved by applying the fitting function to the entire (or a specific) range of logging data, thereby obtaining corrected sonic velocity data. The specific implementation steps are as follows:

[0214] 6-1. Initialization: First, initialize an array corrected_y with the same size as the original y data to store the corrected sound velocity values.

[0215] 6-2. Traverse segments: For each segment (defined by breakpoints), determine the range of x values ​​corresponding to that segment and find the corresponding y values.

[0216] 6-3. Apply fitting parameters: Use the fitting parameters of the current segment (obtained from fitted_params) to calculate the corrected y value through the fitting function (e.g., a linear function).

[0217] 6-4. Store the correction value: Store the calculated corrected y value in the corresponding position in the corrected_y array.

[0218] 6-5. Handling Boundary Cases: Ensure proper handling at segment boundaries and the start / end points of data to avoid data loss or duplication.

[0219] 6-6. Verify the correction effect: Compare and analyze the corrected acoustic velocity data with the VSP velocity data to evaluate the accuracy and effectiveness of the correction method (e.g., Figure 2 ).

[0220] Step 7: Quality Control

[0221] In the process of correcting well logging sonic velocity to VSP velocity trends, quality control is a crucial step to ensure the accuracy and reliability of the results. The following are our specific implementation details regarding quality control:

[0222] 7-1. Data Integrity Check:

[0223] Before starting the calibration, check the integrity of the original logging sonic velocity data (y_data) and depth or time data (x_data) to ensure there are no missing or outliers.

[0224] Verify the validity of breakpoints to ensure they correctly divide the data into multiple reasonable segments.

[0225] 7-2. Fit quality assessment:

[0226] For each segment, evaluate the quality of the fit. The quality of the fit can be assessed by examining the fit residuals (i.e., the difference between the original data and the fitted curve).

[0227] If the fitting quality of a certain segment is poor (e.g., the residuals are too large), you can consider readjusting the segment points or choosing a more suitable fitting function, as shown in Table 4:

[0228] Table 4. Output segmentation points, threshold function, and fitting parameters

[0229]

[0230]

[0231]

[0232]

[0233] 7-3. Parameter rationality check:

[0234] Check the reasonableness of the fitted parameters (fitted_params). For linear fitting, check whether the slope and intercept are within a reasonable range.

[0235] If unreasonable parameter values ​​are found (e.g., the slope is too large or too small), further examination of the data and fitting process is needed to determine the cause of the problem.

[0236] 7-4. Boundary Condition Handling:

[0237] Ensure proper handling at segment boundaries and the start / end points of data to avoid data loss or duplication.

[0238] At the boundaries, specific interpolation or smoothing methods may be required to ensure the continuity of the correction results.

[0239] Step 8: Output Results

[0240] After completing the calibration process, we need to output the results in an appropriate manner for subsequent analysis and application. The following are our specific implementation details regarding result output:

[0241] 8-1. Output of corrected data:

[0242] Output the corrected logging sonic velocity data (corrected_y) in a suitable format, such as a CSV file or database.

[0243] The output file should include the index or depth / time value of the original data for correspondence and comparison with the original data.

[0244] 8-2. Fitting parameter output:

[0245] Output the fitted parameters (fitted_params) for each segment for subsequent analysis and validation.

[0246] The fitting parameters can be output along with the corrected data, or in a separate file.

[0247] 8-3. Quality Control Report:

[0248] Generate a quality control report summarizing the results of data integrity checks, fit quality assessments, parameter rationality checks, and boundary condition handling.

[0249] The report should include the evaluation results and potential problems for each segment so that users can understand the quality and reliability of the calibration process.

[0250] 8-4. Visualization Output:

[0251] You can choose to use charts, images, and other visualization methods to display the data and fitted curves before and after correction, so that users can understand the correction effect more intuitively.

[0252] Visualization outputs can include scatter plots of the original data, fitted curve plots, residual plots, etc.

[0253] This invention proposes an innovative method for correcting logging acoustic velocity. The core of this method is to correct the logging acoustic velocity data to the VSP velocity trend, thereby improving the accuracy and efficiency of well control imaging and thus enhancing the efficiency and accuracy of oil and gas exploration.

[0254] 1. Automatic segmentation and correction:

[0255] This invention introduces an automatic segmentation technology that can automatically segment well logging sonic velocity data based on VSP velocity trends and formation characteristics using an algorithm. This step avoids the tediousness and subjectivity of manual segmentation in traditional methods, improving the accuracy and efficiency of calibration.

[0256] By automatically segmenting, this invention can more accurately identify the influence of different formations on the logging acoustic velocity, thereby using a more suitable mathematical function for fitting within each segment to obtain a more accurate correction factor or correction curve.

[0257] 2. Application of VSP speed trend:

[0258] This invention utilizes VSP data to identify key velocity variation layers or trends and constructs the rate of change of VSP velocities. This step provides an important reference for subsequent well logging sonic velocity correction.

[0259] By introducing the VSP velocity trend, this invention can more comprehensively consider the variation law of formation velocity, making the corrected logging sonic velocity more consistent with the actual geological conditions.

[0260] 3. Mathematical function fitting and correction:

[0261] Within each segment, this invention uses mathematical functions to fit the logging acoustic velocity to obtain a correction factor or correction curve. This step further improves the accuracy and reliability of the correction.

[0262] By fitting mathematical functions, this invention can more accurately describe the relationship between logging acoustic velocity and VSP velocity, thereby obtaining more accurate correction results.

[0263] This invention not only improves exploration efficiency and accuracy and reduces costs, but also enhances data interpretability. By correcting the acoustic velocity of well logging, the data becomes more consistent with actual geological conditions. Furthermore, the method of this invention has broad applicability and can be applied to different types of oil and gas exploration projects, providing strong technical support for the effective development of oil and gas resources.

[0264] In summary, this invention proposes a method for correcting logging sonic velocity to VSP velocity trends based on automatic segmentation. Through steps such as automatic segmentation, application of VSP velocity trends, and mathematical function fitting and correction, this invention can more accurately correct logging sonic velocity, improve exploration efficiency and accuracy, and provide strong technical support for oil and gas exploration and development.

[0265] Finally, it should be emphasized that the above content is only used to illustrate the technical solution of the present invention, and is not intended to limit the scope of protection of the present invention. Those skilled in the art can make simple modifications or equivalent substitutions to the technical solution of the present invention according to actual needs, and such modifications or substitutions do not depart from the essence and scope of the technical solution of the present invention.

Claims

1. A method for correcting logging acoustic velocity to VSP velocity trend based on automatic segmentation, characterized in that, Includes the following steps: Step 1: Data Preparation: Collect well logging sonic velocity data, VSP data, and geological stratification data; Step 2: Data Preprocessing: Preprocess the logging sonic velocity data and VSP data, and then perform calibration; Step 3: Analyze VSP velocity trends: Analyze the VSP data obtained in Step 2, identify the main velocity change layers or velocity trends, and construct the VSP velocity change rate; Step 4: Automatic segmentation: Based on the VSP velocity trend and formation characteristics, the logging sonic velocity data is automatically segmented using an automatic algorithm; Step 5: Segmented Fitting: Within each segment, use mathematical functions to fit the logging sonic velocity to obtain a correction factor or correction curve; Step 6: Correct the logging sonic velocity: Correct the original logging sonic velocity according to the correction factor or correction curve obtained from the fitting. Step 7: Quality Control: Perform a quality check on the corrected logging sonic velocity to ensure that no unreasonable velocity changes or abnormal values ​​were introduced during the correction process; Step 8: Output Results: Output the corrected data.

2. The method according to claim 1, characterized in that, In step 1, the logging acoustic velocity data includes the depth of the target well and adjacent wells, and the acoustic logging velocity; the VSP data includes the depth of the target well and adjacent wells, and the VSP logging velocity; the geological stratification data includes well depth and stratigraphic information.

3. The method according to claim 1, characterized in that, In step 2, the preprocessing includes at least one of cleaning, filtering, denoising, and removing outliers, missing values, or erroneous data.

4. The method according to claim 1, characterized in that, In step 2, the calibration includes depth and / or time calibration.

5. The method according to claim 1, characterized in that, Step 3 specifically includes the following steps: 3-1. Data Preparation: Ensure the accuracy and completeness of the data, and preprocess missing or abnormal data; 3-2. Velocity variation analysis with depth: Identify the layers or intervals where the velocity changes significantly with depth; 3-3. Identification of major velocity change strata or trends: Combine geological background data to analyze the geological significance of velocity change strata or trends; 3-4. Outlier Check: Check for outliers in the speed data, and mark and record outliers. 3-5. Calculation of velocity parameters: Calculate the average velocity, layer velocity, and root mean square velocity; 3-6. Handling of outlier speed parameters: Check the obtained speed parameters for outliers, remove outliers and / or recalculate them; 3-7. VSP Solution Speed ​​Selection: After outlier handling, the average speed is selected as the VSP solution speed for the next step.

6. The method according to claim 5, characterized in that, In section 3-5, the formula for calculating the average velocity is: Among them, V a The average velocity represents the average propagation speed of sound waves within a specific well section. Hd i The depth of the observation point is measured from the ground surface; T i / 2 Vertical one-way time, calculated from the ground surface.

7. The method according to claim 5, characterized in that, In section 3-5, the formula for calculating the layer velocity is: in, V ni For layer velocity; Hd i The depth of the observation point is measured from the ground surface; T i / 2 Vertical one-way time, calculated from the ground surface.

8. The method according to claim 5, characterized in that, In section 3-5, the formula for calculating the root mean square velocity is: in, It is the root mean square velocity; T i It is the time value of the i-th measurement point; V i It is the velocity value at the i-th measurement point; n is the total number of measurement points.

9. The method according to claim 1, characterized in that, In step 4, the automatic algorithm includes at least one of cluster analysis, dynamic programming, and sliding window.

10. The method according to claim 1, characterized in that, Step 4 specifically involves the following steps: 4-1. Data Preparation: Import the database and load the logging acoustic velocity data and the corresponding depth or time data; 4-2. Threshold setting: Set an initial threshold to identify significant changes in the sound wave velocity data; 4-3. Automatic segmentation algorithm: Define a function find_breakpoints that takes sound velocity data, a threshold, and a minimum length as input; Inside the function, the difference between adjacent velocity values ​​is calculated; Indices whose absolute difference is greater than a threshold correspond to possible segmentation points; Iterate through all possible segmentation points. If there is already a segmentation point before the current segmentation point or if the number of data points between two points is greater than the minimum length, then add the current segmentation point to the segmentation point list. Ensure that the last segment point is set correctly; If the distance between the last segment point and the end of the data is greater than the minimum length minus one, then the end of the data is taken as the segment point; 4-4. Segmentation point filtering: Initialize a new list to store the filtered segmentation points; The first point is always retained; Iterate through the list of segment points and check whether the difference between the current point and the previous point is greater than or equal to a preset value. If the condition is met and the current point's depth is less than the specified value, then add the current point to the filtered segment point list; for points with a depth greater than or equal to the specified value, add them to the list unconditionally. 4-5. Obtain the actual segmentation point locations: Use the filtered segmentation point index to obtain the corresponding actual segmentation point position from x_data; 4-6. Output: Output the filtered segment point positions.

11. The method according to claim 1, characterized in that, Step 5 specifically involves the following steps: 5-1. Data Segmentation: Based on the segmentation points obtained in step 4, the np.split function is used to divide x_data and y_data into multiple subsequences; 5-2. Define the fitting function: Choose an appropriate fitting function to describe the relationship between sound wave velocity and depth or time; 5-3. Piecewise Fitting: Use the curve_fit function to fit each segment and obtain the fitting parameters; 5-4. Output and Visualization: After fitting each segment, output the fitting parameters; Visualization techniques are used to intuitively display the original data and fitting results.

12. The method according to claim 1, characterized in that, Step 6 is as follows: 6-1. Initialization: Initialize an array corrected_y of the same size as the original y data to store the corrected acoustic velocity values; 6-2. Traverse the segments: For each segment, determine the range of x values ​​corresponding to that segment and find the corresponding y values; 6-3. Apply fitting parameters: Use the fitting parameters of the current segment to calculate the corrected y value through the fitting function; 6-4. Store the correction value: Store the calculated corrected y value in the corresponding position in the corrected_y array; 6-5. Handling Boundary Cases: Ensure proper handling at segment boundaries and the start / end points of data to avoid data loss or duplication; 6-6: Verify the correction effect: Compare and analyze the corrected acoustic velocity data with the VSP velocity data to evaluate the accuracy and effectiveness of the correction method.

13. The method according to claim 1, characterized in that, The quality check described in step 7 includes at least one of the following: data integrity check, fit quality assessment, parameter rationality check, and boundary condition handling.

14. The method according to claim 1, characterized in that, Step 7 is specifically as follows: 7-1. Data Integrity Check: Check the integrity of the raw logging sonic velocity and depth or time data to ensure there are no missing or outlier values; verify the validity of the segment points. 7-2. Fit quality assessment: For each segment, the quality of the fitting results can be evaluated by checking the fitting residuals; 7-3. Parameter rationality check: Check the reasonableness of the fitted parameters; If unreasonable parameter values ​​are found, further examination of the data and fitting process is needed to determine the cause of the problem. 7-4. Boundary Condition Handling: Ensure proper handling at segment boundaries and data start / end points to avoid data loss or duplication; ensure the continuity of correction results.

15. The method according to claim 1, characterized in that, The output described in step 8 includes corrected data, fitting parameters, quality control report, and visualization data.

16. The method according to claim 1, characterized in that, Step 8 is specifically as follows: 8-1. Output of corrected data: Output the corrected logging acoustic velocity data in a suitable format; The output file should include the index or depth / time value of the original data for correspondence and comparison with the original data; 8-2. Fitting parameter output: Output the fitting parameters for each segment for subsequent analysis and validation; 8-3. Quality Control Report: Generate a quality control report summarizing the results of data integrity checks, fit quality assessments, parameter rationality checks, and boundary condition handling. 8-4. Visualization Output: You can choose to use charts, images, and other visualization methods to display the data and fitted curves before and after correction, so that users can understand the correction effect more intuitively.

17. The method according to claim 16, characterized in that, The visualization output may include at least one of the following: a scatter plot of the original data, a fitted curve plot, and a residual plot.

18. The application of the method according to any one of claims 1-17 in oil and gas exploration and development, geophysical data processing and analysis, and mineral resource exploration in the Shunbei area of ​​the Tarim Basin.

Images