A method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit

Abstract

This invention provides a method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, comprising the following steps: S1, conducting a proppant-fiber transport experiment, dividing the experimental image into local distributions according to a preset method, performing threshold segmentation on each local block, and obtaining the overall mask matrix and coordinates of dark pixels in the experimental image; S2, expanding outwards from each dark pixel as the initial center to form a cluster unit index matrix, thereby achieving cluster unit segmentation; S3, filling the internal holes of the cluster units, identifying hole regions completely surrounded by a single cluster unit by scanning the zero-value elements in the cluster unit index matrix, and filling the holes with the index of the outer clusters; S4, statistically analyzing all cluster units, converting the pixel area into the actual area, and outputting the size distribution parameters of the cluster units. This invention enables the automatic identification of amorphous proppant-fiber cluster units within a slit and the output of cluster unit characteristic parameters.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field development engineering, and in particular to a method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit. Background Technology

[0002] China possesses enormous unconventional oil and gas resources with unique endowments, primarily found in shale oil and gas, tight oil and gas, and coal-fired gas, widely distributed in basins such as Sichuan, Ordos, Junggar, and Songliao. Hydraulic fracturing technology is a core technique for the efficient development of unconventional oil and gas reservoirs. Due to the extremely high complexity requirements of fractures in unconventional oil and gas reservoirs, large-scale use of low-viscosity fracturing fluids is often employed during fracturing to reduce operational friction, connect natural fractures, and promote the formation of complex fractures. Core sampling results from field fracturing tests indicate that low-viscosity fracturing fluids lead to rapid proppant settling and short delivery distances, making it difficult to reach the distal ends of fractures and cover the entire fracture height. This phenomenon results in a large number of dynamic fractures not being effectively supported, failing to form "highways" with sufficient flow capacity. Optimizing proppant delivery strategies to ensure effective support throughout the entire fracture length and height is crucial for improving the utilization rate of artificial fractures and enhancing reservoir stimulation effects.

[0003] To address the challenge of limited longitudinal and transverse support range in fractures, fibers are incorporated into the fracturing process. On one hand, the fibers increase the apparent viscosity of the fracturing fluid, effectively enhancing the proppant's suspension and transport capabilities. On the other hand, the collision and contact between proppant particles and fibers form a network structure, altering the particle settling pattern and slowing its settling velocity, thus enhancing the proppant's transport capacity within the fracturing fluid. During this process, proppant and fibers have a certain probability of combining to form amorphous cluster units, which then create discontinuous support within the fracture, increasing the fracture's height and length. The fracture support effect is related to the formation and dispersion process of the proppant-fiber cluster units, as well as the size distribution of these cluster units. The cluster unit size is controlled by intrinsic material properties and construction parameters such as fiber length, fiber concentration, proppant particle size, proppant concentration, fracturing fluid viscosity, and delivery rate. However, existing research methods mainly rely on proppant-fiber hybrid transport experiments to qualitatively observe the discontinuous distribution morphology of solid particles after sedimentation and accumulation within the fracture. They lack methods for identifying proppant-fiber cluster units and characterizing their size distribution, making it difficult to quantitatively analyze the influence of intrinsic material properties and construction parameters on the formation and dispersion of proppant-fiber clusters. This hinders the provision of a scientific theory for proppant-fiber hybrid fracturing design and parameter optimization. Therefore, characterizing the size of amorphous units within proppant-fiber clusters is crucial for studying fiber-assisted proppant transport and fracture support. Summary of the Invention

[0004] To overcome the problems in the prior art, this invention provides a method, system, electronic device, and storage medium for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, thereby achieving automatic identification of amorphous proppant-fiber cluster units within the slit and outputting the characteristic parameters of the cluster units. To achieve the above objective, this invention provides the following solution:

[0005] In a first aspect, embodiments of the present invention also provide a method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, comprising the following steps:

[0006] S1. Conduct proppant-fiber transport experiment, obtain experimental images, divide the experimental images into local distributions according to a preset method, perform threshold segmentation on each local block, and obtain the overall mask matrix and dark pixel coordinates of the experimental images.

[0007] S2. Using each dark pixel as the initial center, expand outwards according to the search radius r and brightness difference constraints to form a cluster unit index matrix, thereby realizing cluster unit segmentation;

[0008] S3. Fill the holes inside the cluster unit. By scanning the zero-value elements in the cluster unit index matrix, identify the hole area completely surrounded by a single cluster unit, and fill the hole with the index of the outer cluster to ensure the integrity of the cluster unit area.

[0009] S4. Perform statistics on all cluster units, convert the pixel area to the actual area, and output the size distribution parameters of the cluster units.

[0010] A further technical solution is that, in step S1, the experimental image is divided into local distributions according to a preset method, including: dividing the experimental image into m×n blocks equally, where m is the number of horizontal blocks and n is the number of vertical blocks.

[0011] A further technical solution is that in step S1, threshold segmentation is performed on each local block, including: based on the V channel value of the HSV color space representing pixel brightness, the threshold is calculated using the minimum value of the V channel in the local block and the scaling factor k, the dark areas with V channels less than the threshold are extracted to obtain the local mask matrix of each block, and the local mask matrices of all blocks are spliced ​​together according to their spatial positions to obtain the overall mask matrix.

[0012] A further technical solution is that, in step S3, filling the holes inside the cluster unit includes: according to the cluster unit index matrix, if the top, bottom, left, and right sides of a hole are completely surrounded by the same number, then the hole is filled with that number; otherwise, the hole is not filled. The hole is an eight-connected region with an index of 0.

[0013] A further technical solution is that the size distribution parameters in step S4 include: total number of cluster units, average area, average equivalent diameter, maximum equivalent diameter, and cluster unit area ratio.

[0014] Secondly, embodiments of the present invention also provide a system for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, comprising the following modules:

[0015] The experimental image acquisition module is used to conduct proppant-fiber delivery experiments, acquire experimental images, divide the experimental images into local distributions according to a preset method, perform threshold segmentation on each local block, and obtain the overall mask matrix and dark pixel coordinates of the experimental image.

[0016] The cluster unit segmentation module is used to expand outwards from each dark pixel as the initial center according to the search radius r and brightness difference constraints to form a cluster unit index matrix, thereby realizing cluster unit segmentation;

[0017] The hole filling module is used to fill the holes inside the cluster unit. By scanning the zero-value elements in the cluster unit index matrix, it identifies the hole area completely surrounded by a single cluster unit and fills the hole with the index of the outer cluster to ensure the integrity of the cluster unit area.

[0018] The statistical output modulus is used to perform statistics on all cluster units, convert the pixel area into the actual area, and output the size distribution parameters of the cluster units.

[0019] Thirdly, embodiments of the present invention also provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the cluster unit identification and size distribution characterization method as described in any of the embodiments of the present invention.

[0020] Fourthly, embodiments of the present invention also provide a storage medium for storing computer-executable instructions, which, when executed by a computer processor, are used to perform the cluster unit identification and size distribution characterization method as described in any of the embodiments of the present invention.

[0021] Compared with existing technologies, the advantages of this invention are: the image processing method provided by this invention can automatically identify and segment cluster units in experimental images, and automatically determine characteristic parameters such as cluster unit area, equivalent diameter, and average diameter through statistical analysis and feature extraction of cluster units. The annotation method clearly displays the boundary and size distribution of each cluster unit, facilitating observation of the cluster's shape characteristics and spatial distribution.

[0022] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of the method flow provided by the present invention;

[0025] Figure 2 This is a diagram showing the experimental results of proppant-fiber mixed transport in one embodiment of the present invention;

[0026] Figure 3 This is a dark pixel distribution map after block threshold segmentation in one embodiment of the present invention;

[0027] Figure 4 This is a cluster unit annotation diagram in one embodiment of the present invention;

[0028] Figure 5 This is a histogram showing the size distribution of cluster units in one embodiment of the present invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The purpose of this invention is to provide a method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit. To make the above-mentioned objectives, features and advantages of this invention more apparent and understandable, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0031] like Figure 1 As shown, this invention provides a method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, comprising the following steps:

[0032] S1. Conduct proppant-fiber transport experiment, obtain experimental images, divide the experimental images into local distributions according to a preset method, perform threshold segmentation on each local block, and obtain the overall mask matrix and dark pixel coordinates of the experimental images.

[0033] S2. Using each dark pixel as the initial center, expand outwards according to the search radius r and brightness difference constraints to form a cluster unit index matrix, thereby achieving cluster unit segmentation.

[0034] S3. Fill the holes inside the cluster unit. By scanning the zero-value elements in the cluster unit index matrix, identify the hole area completely surrounded by a single cluster unit, and fill the hole with the index of the outer cluster to ensure the integrity of the cluster unit area.

[0035] S4. Perform statistics on all cluster units, convert the pixel area to the actual area, and output the size distribution parameters of the cluster units.

[0036] In step S1, as Figure 2 As shown, after acquiring the experimental images, a computer is used to acquire and read the experimental images, and the original experimental images are loaded through the input image path.

[0037] Specifically, technicians can use MATLAB's imread function to read experimental images from a specified path, and the user can input the actual height H of the image for conversion between pixels and actual size.

[0038] Subsequently, the input image is divided into local blocks with a specified number of rows and columns using a computer. Thresholding segmentation is then performed on each block to obtain the overall mask matrix and the coordinates of dark pixels. Specifically, such as... Figure 3 As shown, the image can be equally divided into m×n blocks, where m is the number of horizontal blocks and n is the number of vertical blocks.

[0039] When performing threshold segmentation on blocks, the threshold is calculated based on the V channel value in the HSV color space, which represents pixel brightness. The minimum value of the V channel within the block and the scaling factor k are used to extract dark regions where the V channel value is less than the threshold, resulting in the local mask matrix for each block. The overall mask matrix is ​​obtained by concatenating the local mask matrices of all blocks according to their spatial positions. The expression is as follows:

[0040] For block B i Let j be the i-th block, and let the V channel value of the pixels in this block be V(x,y). Its minimum value is:

[0041]

[0042] The threshold for this block is:

[0043]

[0044] In the formula, k is the scaling factor, and the local mask matrix is ​​defined as:

[0045]

[0046] The overall mask matrix is ​​obtained by concatenating all the local mask matrices of the blocks according to their spatial positions.

[0047]

[0048] In the formula, Ω is the set of pixel coordinates of the entire image, and the pixel points that satisfy M(x,y)=1 are defined as dark pixels.

[0049] In step S2, the search area expands outwards from the dark pixel. The candidate pixels must simultaneously satisfy the search range condition and the brightness difference condition, as expressed below:

[0050] Let the coordinates of a dark pixel be (x0, y0), and its V channel value be V(x0, y0). The set of pixels within the search radius r is:

[0051]

[0052] For any pixel (x,y)∈Ω r (x0, y0), the brightness difference is:

[0053] △V(x,y)=V(x,y)-V(x0,y0) (6)

[0054] The user specifies the brightness difference threshold as diff, and the brightness difference condition must be met as follows:

[0055] △V(x,y)≤diff (7)

[0056] The total number of dark pixels is N p The coordinates of the i-th dark pixel are (x... i ,y i Given that the coordinates of any pixel are (x, y), and the cluster unit index matrix M... c The expression for (x, y) is:

[0057]

[0058] If dark pixels numbered n, n+1, ..., n+m have overlapping cluster units due to their close proximity, resulting in the cluster unit index of the pixel at coordinate (x, y) belonging to dark pixels numbered n, n+1, ..., n+m simultaneously, then let M... c (x,y)=n,M c (x n ,y n) = M c (x n+1 ,y n+1 ) = ... = M c (x n+m ,y n+m ) = n.

[0059] The cluster cell index matrix is ​​rearranged so that its index values ​​increase sequentially from 1, for example:

[0060]

[0061] All sets of pixels with the same index form a cluster unit, and the set of pixels in cluster unit i is:

[0062] C i ={(x,y)∈Ω∣M c (x,y)=i} (10)

[0063] In step S3, the cluster unit holes are filled. Based on the cluster unit index matrix, if the top, bottom, left, and right sides of a hole (an 8-connected region with index 0) are completely surrounded by the same number, then the hole is filled with that number; otherwise, the hole is not filled. For example:

[0064]

[0065] In step S4, the area of ​​all cluster units is statistically analyzed, and the pixel area S of cluster unit i is calculated. i Cluster unit index matrix M c The expression for converting pixel area to actual area based on the number of times the digit i appears is:

[0066]

[0067] The equivalent diameter is calculated using the equivalent circle, and the expression is:

[0068]

[0069] In the formula: A i Let i be the actual area of ​​cluster unit i, in mm. 2 S i D is the pixel area of ​​cluster unit i; H is the true height of the original input image, in mm; M is the number of pixels in the height direction of the input image, in mm. eq,i Let be the equivalent diameter of cluster unit i, in mm.

[0070] Let the total number of cluster units be N. c The following size distribution parameters can be obtained:

[0071] Average area (unit: mm) 2 ):

[0072]

[0073] Average equivalent diameter (unit: mm):

[0074]

[0075] Maximum equivalent diameter (unit: mm):

[0076] D eq,max =max(D eq,i ), i = 1, ..., N c (17)

[0077] Cluster unit area ratio:

[0078]

[0079] In the formula: A tot The true area of ​​the original input image, in mm. 2 .

[0080] To facilitate the visualization of statistical results, different random colors can be assigned to different cluster units on the original input image to ensure that adjacent cluster units have different colors, thereby enhancing the boundary recognition of cluster units and the overall visualization effect.

[0081] In addition, other cluster unit size distribution parameters can also be visualized, for example, all cluster units can be divided into several intervals according to their equivalent diameter [b k ,b k+1 ), b k With b k+1 Let the upper and lower bounds of the k-th grouping interval be defined, and calculate the percentage of cluster units within each interval (number of clusters within the interval / total number of clusters). Then, plot a histogram of cluster unit size distribution, as shown below. Figure 5 As shown.

[0082] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0083] Propionate delivery experiments based on fracturing parameters of a shale reservoir in the target work area were conducted, and the original experimental images were obtained as follows: Figure 2 As shown, the actual height of the image is H = 140 mm.

[0084] Step S1 is used to perform block thresholding of the image, specifying the number of horizontal blocks m = 8, the number of vertical blocks n = 7, and the minimum value of the V channel in block (1,1). With a scaling factor k = 0.45, the V channel threshold for this block is T. (1,1)= (0.45+1)×0.2824=0.41, the local mask matrix of this block is:

[0085]

[0086] The overall mask matrix M(x,y) is obtained by concatenating all the local mask matrices according to their spatial positions. Pixels that satisfy M(x,y) = 1 are dark pixels. In the example image, there are a total of 89,666 dark pixels, and the distribution of all dark pixels is as follows: Figure 3 As shown.

[0087] The cluster units are divided using step S2, expanding outwards from each dark pixel as the center, with a specified search radius r = 30 and a brightness difference threshold of diff = 0.09. The cluster unit index matrix is ​​as follows:

[0088]

[0089] Overlapping cluster units were merged, and the cluster unit index matrix was rearranged so that the index values ​​increased sequentially from 1. All sets of pixels with the same index constituted a cluster unit, and a total of 298 cluster units were identified.

[0090] Step S3 is used to fill the cluster unit holes. According to the cluster unit index matrix, if the top, bottom, left and right sides of a certain hole (an 8-connected region with index 0) are completely surrounded by the same number, then the hole is filled with that number; otherwise, the hole is not filled.

[0091] Step S4 is used to statistically analyze all cluster units. Given that the input image height H = 140 mm, the number of pixels in the image height direction M = 1134, and the pixel area S of one cluster unit... i =665, its actual area is:

[0092]

[0093] The equivalent diameter is:

[0094]

[0095] Following the method described above, the size distribution information of all cluster units can be obtained as shown in the table below:

[0096]

[0097] Furthermore, the statistical results can be visualized and histograms plotted. Different random colors are assigned to different cluster units on the original input image, ensuring that adjacent cluster units have different colors. The processed result is as follows: Figure 4 As shown.

[0098] Divide all cluster units into intervals [0,1), [1,2), ..., [16,17) according to their equivalent diameter, and calculate the percentage of cluster units in each interval (number of clusters in the interval / total number of clusters). Plot a histogram of cluster unit size distribution as follows: Figure 5 As shown.

[0099] The method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit provided by this invention can quantitatively characterize the size and proportion distribution of proppant-fiber clusters formed under specific construction parameters. This is beneficial for in-depth analysis of the influence of intrinsic material properties and construction parameters on the proppant-fiber cluster support effect, and thus provides a theoretical method for the optimized design of proppant-fiber hybrid fracturing.

[0100] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, characterized in that, Includes the following steps: S1. Conduct proppant-fiber transport experiment, obtain experimental images, divide the experimental images into local distributions according to a preset method, perform threshold segmentation on each local block, and obtain the overall mask matrix and dark pixel coordinates of the experimental images. In step S1, threshold segmentation is performed on each local block, including: based on the V channel value of the HSV color space representing pixel brightness, the threshold is calculated using the minimum value of the V channel in the local block and the scaling factor k, dark areas with V channels less than the threshold are extracted to obtain the local mask matrix of each block, and the local mask matrices of all blocks are spliced ​​together according to their spatial positions to obtain the overall mask matrix. For block B i , j Let the i-th, j-th block have a V channel value of V(x, y) for the pixels within that block, and its minimum value be: ； The threshold for this block is: ； In the formula, k is the scaling factor, and the local mask matrix is ​​defined as: ； The overall mask matrix is ​​obtained by concatenating all the local mask matrices of the blocks according to their spatial positions. ； In the formula, Ω is the set of all pixel coordinates of the image, and the pixels that satisfy M(x, y)=1 are defined as dark pixels. S2. Using each dark pixel as the initial center, expand outwards according to the search radius r and brightness difference constraints to form a cluster unit index matrix, thereby achieving cluster unit segmentation. Let the coordinates of a dark pixel be (x0, y0), and its V channel value be V(x0, y0). The set of pixels within the search radius r is: ； For any pixel The brightness difference is: ； The user specifies the brightness difference threshold as diff, and the brightness difference condition must be met as follows: ； The total number of dark pixels is N p The coordinates of the i-th dark pixel are (x... i , y i Given that the coordinates of any pixel are (x, y), and the cluster unit index matrix M... c The expression for (x, y) is: ； S3. Fill the holes inside the cluster unit. By scanning the zero-value elements in the cluster unit index matrix, identify the hole area completely surrounded by a single cluster unit, and fill the hole with the index of the outer cluster. The step S3 of filling the holes inside the cluster unit includes: according to the cluster unit index matrix, if the top, bottom, left and right sides of a hole are completely surrounded by the same number, then the hole is filled with that number; otherwise, the hole is not filled. The hole is an eight-connected region with an index of 0. S4. Perform statistics on all cluster units, convert the pixel area to the actual area, and output the size distribution parameters of the cluster units; The expression for the actual area is: ； In the formula: A i Let i be the actual area of ​​cluster unit i, in mm. 2 S i H is the pixel area of ​​cluster unit i; H is the true height of the original input image, in mm; M is the number of pixels in the height direction of the input image.

2. The method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit according to claim 1, wherein step S1 involves dividing the experimental images into local distributions according to a preset method, including: The experimental image was divided into m×n blocks, where m is the number of horizontal blocks and n is the number of vertical blocks.

3. The method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit according to claim 1, wherein the size distribution parameters in step S4 include: Total number of cluster units, average area, average equivalent diameter, maximum equivalent diameter, and percentage of cluster unit area.

4. A system for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit, implementing the method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit as described in any one of claims 1-3, characterized in that, Includes the following modules: The experimental image acquisition module is used to perform proppant-fiber delivery experiments, acquire experimental images, divide the experimental images into local distributions according to a preset method, perform threshold segmentation on each local block, and obtain the overall mask matrix and dark pixel coordinates of the experimental image. The cluster unit segmentation module is used to expand outwards from each dark pixel as the initial center according to the search radius r and brightness difference constraints to form a cluster unit index matrix, thereby realizing cluster unit segmentation. The hole filling module is used to fill the holes inside the cluster unit. By scanning the zero-value elements in the cluster unit index matrix, it identifies the hole area completely surrounded by a single cluster unit and fills the hole with the index of the outer cluster. The statistical output modulus is used to perform statistics on all cluster units, convert the pixel area into the actual area, and output the size distribution parameters of the cluster units.

5. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit as described in any one of claims 1-3.

6. A storage medium storing computer-executable instructions, which, when executed by a computer processor, are used to perform the method for identifying and characterizing the size distribution of amorphous proppant-fiber cluster units within a slit as described in any one of claims 1-3.

Images