Core fracture identification and modeling method based on CT scanning data

Abstract

This invention discloses a method for identifying and modeling rock core fractures based on CT scan data, belonging to the field of fracture identification technology. The fracture identification method comprises extracting fracture point information from CT scan data based on CT numerical values; performing ellipsoid fitting on the three-dimensional fracture point information; clustering the fitted characteristic data to obtain multiple fracture clusters; and identifying rock core fractures based on the multiple fracture clusters. This invention improves data processing efficiency and enhances fracture identification accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of crack identification in rock fracturing, and more particularly to a core crack identification and modeling method based on CT scanning data. Background Art

[0002] In fields such as oil and gas development and geotechnical engineering, core fracturing experiments are a key technique for studying rock seepage characteristics, mechanical properties, and fracture patterns. Traditional core fracturing analysis methods often rely on acoustic emission signals (microseismic data) to monitor the formation and propagation of fractures. However, this method has certain limitations, such as low resolution and an inability to visually demonstrate the specific morphology of fractures.

[0003] With the development of CT scanning technology, it has become possible to use CT scanning data to identify fractures in rock cores. CT scanning provides high-resolution three-dimensional images, allowing researchers to visually observe the fracture morphology and propagation process within the rock core. However, due to the large amount of data generated during dynamic CT scanning of rock cores, how to effectively process and analyze this data to accurately identify and predict fractures in the core has become a technical challenge.

[0004] In the existing technology, although there are some core fracture identification methods based on CT scanning data, these methods often have the following shortcomings:

[0005] The data processing efficiency is low, making it difficult to efficiently analyze large amounts of CT scan data.

[0006] The accuracy of crack identification is not high, and it is difficult to accurately distinguish complex crack combination relationships.

[0007] The lack of quantitative analysis of fracture parameters cannot provide sufficient information support for fracturing process optimization.

[0008] Therefore, how to overcome the above-mentioned shortcomings and propose a method for accurately identifying cracks is an urgent problem that those skilled in the art need to solve. Summary of the Invention

[0009] In view of this, the present invention provides a core fracture identification and modeling method based on CT scanning data; it aims to improve data processing efficiency, enhance fracture identification accuracy, and provide more accurate and comprehensive fracture parameter analysis for fields such as oil and gas development.

[0010] In order to achieve the above object, the present invention adopts the following technical solutions:

[0011] This application first discloses a core fracture identification method based on CT scanning data, the steps comprising:

[0012] Extract the crack point information in CT scan data based on CT numerical value;

[0013] Perform ellipsoid fitting on 3D crack point element information;

[0014] Clustering is performed based on the characteristic data obtained by fitting to obtain multiple crack clusters;

[0015] Core fractures were identified based on multiple fracture clusters.

[0016] According to an embodiment of the present invention, extracting crack point metadata from CT scan data based on CT numerical values ​​includes:

[0017] The rock skeleton data in the CT scan data is eliminated according to the CT value range corresponding to the crack to obtain the crack data;

[0018] Extract 2D CT scan slices from the crack data and identify pixel information of each crack;

[0019] Using two-dimensional CT scan slices as indexes, the crack pixel information is converted into three-dimensional spatial form to obtain the crack point element information.

[0020] According to an embodiment of the present invention, performing ellipsoid fitting on three-dimensional crack point element information includes:

[0021] S21, randomly select a crack point element information, set the initial ellipsoid parameters; the ellipsoid parameters include the main semi-major axis vector , minor semi-major axis vector and the shortest semi-major axis vector , for cracks, , Together they represent the direction of crack expansion; Indicates the opening of the crack;

[0022] S22. When the ellipsoid covers other point elements, the ellipsoid parameters are updated in combination with the positions of other point elements until the updated ellipsoid cannot cover the new point element, and the feature set of the current crack point element is obtained, which is expressed as (Val, X, Y, Z, , , );

[0023] S23. Traverse all crack point elements according to steps S21 and S22 to obtain a feature set of all crack point elements.

[0024] According to an embodiment of the present invention, when the ellipsoid fitted to a crack point element cannot cover other point elements, the current crack point element is marked and discarded.

[0025] According to an embodiment of the present invention, clustering is performed based on the characteristic data obtained by fitting to obtain multiple crack clusters, including:

[0026] S31, randomly selecting characteristic data of a crack point element, determining the deviation angle between the current crack point element and other point elements based on the main semi-major axis, and determining the neighborhood radius between the current crack point element and other point elements based on the shortest semi-major axis;

[0027] S32, based on the deviation angle and the neighborhood radius, using a distributed parallel merging method to identify crack point elements that meet the conditions and generate multiple crack clusters;

[0028] As an option, it also includes:

[0029] S33. Based on the spatial distance of the crack clusters and the deviation angle of the semi-major axis vector, the overlapping crack clusters are merged, and the merged noise points, that is, the point elements that do not meet the conditions, are further removed.

[0030] According to an embodiment of the present invention, the present invention further includes:

[0031] Cluster each crack cluster in turn to obtain multiple crack subsets.

[0032] The mean of the shortest semi-major axis of each crack point element in each crack subset is calculated, and the crack subset is graded according to the mean of the shortest semi-major axis.

[0033] On the other hand, the present application also provides a core fracture modeling method based on CT scanning data.

[0034] This method is used to extract the three-dimensional morphology of the fracture clusters generated using any of the above core fracture identification methods based on CT scan data, construct a three-dimensional fracture model, and assign fracture attribute parameters.

[0035] The fracture attribute parameters include fracture volume, curvature, maximum extension area, dip, aperture, and / or grade.

[0036] As can be seen from the above technical solutions, the present invention discloses a method for identifying and modeling core fractures based on CT scanning data. Compared with the prior art, the present invention has the following significant beneficial effects:

[0037] (1) The present invention can quantitatively analyze the fracture point metadata through ellipsoid fitting and clustering, thereby effectively identifying complex fracture combinations and distinguishing interlaced fractures;

[0038] (2) The present invention can efficiently and quickly construct the shape of core cracks and accurately extract the crack volume, curvature, maximum expansion area, inclination, aperture and other attribute parameters, thereby achieving an accurate description of the internal crack structure of the core and intuitively displaying the morphology and distribution of the cracks, providing researchers with a more accurate means of crack analysis.

[0039] This method not only improves the accuracy and efficiency of hydraulic fracturing fracture modeling in rock cores, but also has broad applicability, reduces experimental costs, improves research efficiency, and has a positive impact on actual production data. It can provide a scientific and economical solution for oil and gas field development, with significant technical advantages and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0041] Figure 1 This is a flow chart of the core fracture identification method based on CT scanning data of the present invention;

[0042] Figure 2 It is a 2D slice image of CT scan of the present invention;

[0043] Figure 3 This is a CT scan three-dimensional slice stack image of the present invention;

[0044] Figure 4 This is an example of the ellipsoid of the present invention. DETAILED DESCRIPTION

[0045] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] Figure 1 This is a flow chart of the core fracture identification method based on CT scan data of the present invention. The present invention can handle large amounts of CT scan data; and by spatially identifying the rock fracture crack information in the CT scan data and classifying the fracture levels, a more accurate description of the internal fracture structure of the core and an accurate prediction of the hydraulic fracturing effect can be achieved.

[0047] The specific execution steps of the crack identification method in this embodiment include:

[0048] Extract the crack point information in CT scan data based on CT numerical value;

[0049] Perform ellipsoid fitting on 3D crack point element information;

[0050] Cluster the obtained characteristic data through fitting to obtain multiple fracture clusters;

[0051] Identify the core fractures based on the multiple fracture clusters.

[0052] In one embodiment,

[0053] Extract the fracture point element information from the CT scan data based on the CT value, and the steps include:

[0054] To obtain the fracture spatial distribution information, first剔除 the rock skeleton data from the CT scan data according to the CT value range corresponding to the fractures to obtain the fracture data; in this embodiment, for a single CT scan data volume (M*M*L), M and L are the CT scan image grid parameters, usually 256<M<1024, 512<L<2048. According to the CT value difference between the rock skeleton and the fractures, statistically obtain the CT value range [v1, v2] corresponding to the fractures, so as to separate the fracture data volume from the rock skeleton data volume, facilitating the next step of fracture identification and modeling processing for the fracture data volume;

[0055] Extract the two-dimensional CT scan slices from the fracture data, and identify each fracture pixel information, including the CT value (val) and the spatial position (x, y);

[0056] Taking the two-dimensional CT scan slice as an index, convert the fracture pixel information into a three-dimensional spatial form to obtain the fracture point element information. After the fracture information is identified and extracted in this application, using the spatial position relationship between the continuous two-dimensional CT scan slices of the entire three-dimensional data volume, convert all two-dimensional fracture information (val, x, y) into a three-dimensional data volume (Val, X, Y, Z), where (val, x, y) is set according to the CT scan data slice information, and the two-dimensional slice information is as Figure 2 shown, Z represents the index of the two-dimensional CT slice, and the three-dimensional slice stacking is as Figure 3 shown.

[0057] In a preferred embodiment, due to the differences in scale and intensity of the fractures formed by the rupture (the final manifestation is the difference in the value of v in a single fracture pixel), to facilitate comparison and establish the combined relationship between different fracture surfaces, thin out the fracture data volume and extract the core areas of each fracture surface to reduce the calculation amount.

[0058] In one embodiment,

[0059] Considering that the fracture surfaces of different levels of fractures may be interleaved with each other, the present invention identifies and classifies the interleaved fracture surfaces by introducing a spatial cyclic ellipsoid fitting method. The ellipsoid fitting example is referred to Figure 4 , and the specific fitting steps include:

[0060] S21, randomly select a crack point information N i (Val i , X i , Y i , Z i ), according to the PKN crack modeling principle, the spatial ellipsoid fitting modeling is performed. The two-dimensional PKN model is a long and narrow ellipse. In this application, it is three-dimensionalized into a long and narrow ellipsoid. The ellipsoid has three directions of semi-major axis, namely the main semi-major axis vector , minor semi-major axis vector and the shortest semi-major axis vector For cracks, , Together they indicate the expansion and extension behavior of the crack body; Indicates the opening of the crack body, that is, the crack width.

[0061] In this embodiment, the initial ellipsoid parameters are first set; the radius of the adjacent regions is changed by cyclic increments, that is, the semi-major axis vector feature of the ellipsoid ( , , ), thereby obtaining the crack point meta-information feature dataset;

[0062] It should be noted that, in this embodiment, as long as the main semi-major axis is guaranteed Relative to the minor semi-major axis and the shortest semimajor axis It is large enough so that the main semi-major axis can be directly As the crack length parameter; optionally, the initial relationship between the three can be set as: ≧2 ≧4 ; And allow the situation where A / B is much greater than C during the fitting process;

[0063] S22. When the ellipsoid covers other point elements, the ellipsoid parameters are updated in combination with the positions of other point elements until the updated ellipsoid cannot cover the new point element, and the feature set of the current crack point element is obtained, which is expressed as (Val, X, Y, Z, , , );

[0064] In the present invention, the vector A is generated after the ellipse is initially fitted. i 、B i 、C i (ie the above feature set), and then traverse the new point element, if the new point element is consistent with vector A i 、B i 、C i With spatial similarity, update A i 、B i 、Ci , otherwise generate a new vector A j 、B j 、C j .

[0065] The present application covers other point elements by enveloping adjacent points together, thereby recalculating the ellipse model based on the absorbed point elements, and then updating the semi-major axis parameters until there are no more points that can satisfy the current ellipsoid set.

[0066] It can be understood that the above process of this application is to establish an ellipsoid model with a core point as the center of the ellipsoid, and the basic ratio of the three semi-major axes is 4-2-1. If there are other point elements in the established ellipsoid model, they will be absorbed and recalculated until there are no more point elements to be absorbed. It is worth noting that the neighborhood radius is defined based on the vector, so there are not only size constraints but also direction constraints.

[0067] S23. Traverse all crack point elements according to steps S21 and S22 to obtain a feature set of all crack point elements.

[0068] In a preferred embodiment, when the ellipsoid fitted to the crack point element cannot cover other point elements, the current crack point element is marked and discarded.

[0069] In one embodiment,

[0070] Clustering is performed based on the characteristic data obtained by fitting to obtain multiple crack clusters. That is, according to the characteristic information of the semi-axis (including length and direction), the crack point elements with similar characteristics are divided into a single crack cluster set Frac. All crack point elements are traversed and divided into several crack clusters to obtain Frac1…Frac n .

[0071] The partitioning steps include:

[0072] S31, randomly select a crack point element feature data, determine the deviation angle between the current crack point element and other point elements according to the main semi-major axis, and determine the neighborhood radius between the current crack point element and other point elements according to the shortest semi-major axis,

[0073] S32, based on the deviation angle and neighborhood radius, combined with the preset deviation angle threshold and neighborhood radius threshold, a distributed parallel merging method is used to identify crack point elements that meet the conditions, and multiple crack clusters Frac are generated;

[0074] Among them, different threshold settings lead to different classification results.

[0075] S33. Based on the spatial distance of the crack clusters and the deviation angle of the semi-major axis vector, the overlapping crack clusters are merged, and the merged noise points, that is, the point elements that do not meet the conditions, are further removed.

[0076] In one embodiment, the merging principle is:

[0077] The spatial distance error (calculated based on the X, Y, Z features of each point element in the set) is less than 1%;

[0078] The deviation angle of the semi-major axis vector is less than 1°.

[0079] This application adopts a distributed parallel merging approach to increase the generation speed of crack clusters;

[0080] At the same time, this clustering method is applied to the image dataset after CT scanning, and the functions of pixel processing and recognition are adaptively added. That is, since the pixel values ​​val of noise and abnormal points are completely different from the val of normal points, they can be classified by pixel values ​​(val, also called CT values). Furthermore, abnormal points and noise points can be identified based on pixel values;

[0081] In addition, this application adds the concept of vector, namely directionality, based on the SUBCLU clustering algorithm, so as to further distinguish and identify the direction of different small cracks (such as tortuosity, bifurcation, etc.) according to directionality.

[0082] In one embodiment,

[0083] The above steps can distinguish the crack information with obvious spatial positions and crack body morphology angles, but small tortuous cracks still need to be further divided and graded.

[0084] In order to deal with extremely small feature differences, this application introduces the LM K-means algorithm (an algorithm for calculating extremely small difference features of a data set) to (Val i , , ) are further clustered to obtain multiple subsets; because the A and B semi-axis vectors of different zigzag fractures are different, more precise crack identification is obtained.

[0085] In another embodiment, the average of the shortest semi-major axes of the fracture points in each fracture subset is further calculated, and the fracture subsets are graded according to the average of the shortest semi-major axes.

[0086] This process is used to optimize the fracture model, that is, in the process of establishing a fracture zone, the neighborhood radius is set (in fact, the core point The size of the crack point is used to identify and cluster other crack point elements into the crack zone set. Inevitably, this crack zone set will have different (Of course they are close), so averaging is needed as information about the set of crack clusters.

[0087] In this embodiment, the crack level is determined based on the size of the crack opening. One of the crack classification standards is:

[0088] The opening is less than 0.1mm, which is a tiny crack. It is usually difficult to observe with the naked eye and requires the help of a microscope or other professional equipment.

[0089] The opening is between 0.1mm and 1mm, which is a small crack that can be observed with the naked eye, but the extension length and spacing of the crack are small;

[0090] The cracks with an opening between 1mm and 10mm are medium cracks. The cracks are more obvious and may have a certain impact on the stability of the rock mass.

[0091] If the opening is greater than 10mm, it is a large crack. The crack is very obvious and has a significant impact on the stability of the rock mass. Special engineering treatment may be required.

[0092] It should be noted that this solution can be customized to classify crack levels for subsequent statistical analysis. At the same time, this step loops through all crack clusters to complete crack identification and crack classification for all crack clusters.

[0093] In one embodiment,

[0094] This application also provides a core fracture modeling method based on CT scan data, which is used to extract the three-dimensional morphology of a single fracture cluster generated in the above steps and assign fracture attribute parameters, and then traverse all fracture clusters to establish a three-dimensional fracture model of the entire rock sample.

[0095] Because a fracture cluster is a three-dimensional object with certain fracture characteristics (not an ellipsoid, but rather an arrangement of multiple ellipsoids), its volume can be calculated. The curvature, area, and inclination of the surface calculated by fitting the three-dimensional surface are used to refine the fracture information of the fracture cluster. Therefore, in this embodiment, fracture attribute parameters may include fracture volume, curvature, maximum expansion area, inclination, aperture, and / or grade.

[0096] The core hydraulic fracturing fracture modeling method of the present invention based on CT scanning data can generate three-dimensional imaging through non-destructive testing, provide quantitative analysis of fracture distribution and parameters, simulate and verify the fracturing process and reduce testing costs.

[0097] At the same time, the improved SUBCLU clustering algorithm in the present invention can effectively identify complex fracture combination relationships, distinguish interlaced fractures and conduct quantitative analysis. It has a wider applicability and also has a good effect on actual production data. It can efficiently and quickly construct the core fracture shape, thereby achieving an accurate description of the internal fracture structure of the core and the prediction of the hydraulic fracturing effect.

[0098] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0099] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A core fracture identification method based on CT scanning data, characterized in that: include: Extract the crack point information in CT scan data based on CT numerical value; Extract the two-dimensional CT scan slices from the crack data and identify the pixel information of each crack, including the CT value val and the spatial position (x, y); Using the spatial position relationship between consecutive 2D CT scan slices of the entire 3D data volume, all 2D crack information (val, x, y) is converted into a 3D data volume (Val, X, Y, Z), where Z represents the index of the 2D CT slice; Perform ellipsoid fitting on 3D crack point element information; including: S21, randomly select a crack point element information and set the initial ellipsoid parameters; the ellipsoid parameters include the major semi-major axis vector A, the minor semi-major axis vector B and the shortest semi-major axis vector C; S22. When the ellipsoid covers other point elements, update the ellipsoid parameters based on the positions of the other point elements until the updated ellipsoid cannot cover the new point element, thereby obtaining a feature set of the current crack point element; S23, traverse all crack point elements according to steps S21 and S22 to obtain a feature set of all crack point elements; Clustering is performed based on the fitted characteristic data to obtain multiple crack clusters, including: S31, randomly selecting characteristic data of a crack point element, determining the deviation angle between the current crack point element and other point elements based on the main semi-major axis, and determining the neighborhood radius between the current crack point element and other point elements based on the shortest semi-major axis; S32, based on the deviation angle and the neighborhood radius, using a distributed parallel merging method to identify crack point elements that meet the conditions and generate multiple crack clusters; Core fractures are identified based on multiple fracture clusters, including extracting the three-dimensional morphology of each fracture cluster, constructing a three-dimensional fracture model, and assigning fracture attribute parameters; fracture attribute parameters include fracture volume, curvature, maximum expansion area, dip, aperture, and / or grade.

2. The core fracture identification method according to claim 1, characterized in that: Extract the crack point metadata from CT scan data based on CT numerical values, including: The rock skeleton data in the CT scan data is eliminated according to the CT value range corresponding to the crack to obtain the crack data; Extract 2D CT scan slices from the crack data and identify pixel information of each crack; Using two-dimensional CT scan slices as indexes, the crack pixel information is converted into three-dimensional spatial form to obtain the crack point element information.

3. The core fracture identification method according to claim 1, characterized in that: When the ellipsoid fitting of the crack point element cannot cover other point elements, the current crack point element will be marked and discarded.

4. The core fracture identification method according to claim 1, characterized in that: Also includes: Cluster each crack cluster in turn to obtain multiple crack subsets; The mean of the shortest semi-major axis of each crack point element in each crack subset is calculated, and the crack subset is graded according to the mean of the shortest semi-major axis.

Images