Quartz sand micro-proppant flow conductivity simulation method and system

Abstract

The invention discloses a quartz sand micro-proppant flow conductivity simulation method and system, and relates to the technical field of geological exploration and development, and the quartz sand micro-proppant flow conductivity simulation method comprises the following steps: obtaining a quartz sand micro-proppant sample, and pretreating the quartz sand micro-proppant sample; performing three-dimensional imaging on the quartz sand micro-proppant through three-dimensional imaging equipment to obtain three-dimensional image data of the quartz sand micro-proppant; processing and analyzing the three-dimensional image data, and extracting structural parameters of the quartz sand micro-proppant; constructing a pore network model of the micro-proppant by utilizing a machine learning algorithm based on the structure parameters; and simulating the flow of fluid in the pore network of the micro-proppant based on the constructed pore network model, and calculating the seepage characteristics and the flow conductivity of the micro-proppant. According to the method, the relationship between the characteristics of the quartz sand micro-proppant and the diversion performance can be identified, automatic learning and optimization of the model are realized, research and development of the micro-proppant are accelerated, the oil and gas recovery efficiency is improved, and the development cost is reduced.

Description

Technical Field

[0001] The invention relates to the technical field of geological exploration and development, and more particularly to a method and system for simulating the flow conductivity of quartz sand micro-proppant. Background Art

[0002] In the exploitation of unconventional oil and gas resources (such as shale gas, tight oil, etc.), hydraulic fracturing technology is widely used. Micro-proppants are filled in reservoir cracks to maintain diversion channels, aiming to prop open the cracks and maintain their diversion capacity for a long time to ensure smooth oil and gas production. The quality of its performance directly affects the recovery rate of oil and gas and the economic benefits of exploitation. Therefore, it is very important to choose the right proppant.

[0003] Traditional research methods for microproppants rely on optical microscopes and scanning electron microscopes (SEM) to obtain two-dimensional image information to analyze the characteristics of microproppant particles. However, in actual reservoir environments, microproppants exist in a three-dimensional stacked form, and two-dimensional imaging cannot fully capture the three-dimensional structural details of the particles. For example, key geometric features such as the actual aspect ratio of the particles and complex curved surface shapes are severely simplified or even distorted in a two-dimensional perspective, making the subsequent shape factors calculated based on these data deviate from reality, making it difficult to accurately reflect the impact of particles on pore morphology and fluid flow in three-dimensional space, ultimately leading to deviations in the estimation of the overall conductivity of microproppants.

[0004] As an important proppant, quartz sand is widely used in the oil and gas extraction industry because its main component is silicon dioxide (SiO2) and it exhibits excellent physical and chemical properties. Its conductivity directly affects the recovery rate of oil and gas wells. However, due to the complexity and uncertainty of microproppants, traditional calculation models are difficult to accurately predict their conductivity and cannot systematically grasp the intrinsic relationship between the pore structure and conductivity of microproppants, which limits the advancement of efficient structural design and performance optimization of microproppants.

[0005] Therefore, how to provide a quartz sand microproppant conductivity simulation method and system, identify the relationship between the characteristics of the microproppant and the conductivity performance, realize automatic learning and optimization of the model, accelerate the research and development of microproppant, improve oil and gas recovery efficiency, and reduce development costs is a problem that technical personnel in this field urgently need to solve. Summary of the invention

[0006] In view of this, the present invention provides a method and system for simulating the conductivity of quartz sand microproppants, which can identify the relationship between the characteristics of microproppants and their conductivity, realize automatic learning and optimization of the model, accelerate the research and development of microproppants, improve oil and gas recovery efficiency, and reduce development costs.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] A method for simulating the conductivity of quartz sand micro-proppant, comprising:

[0009] Obtaining a quartz sand micro-proppant sample, and pre-treating the quartz sand micro-proppant sample;

[0010] Performing three-dimensional imaging on the quartz sand micro-proppant by using a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant;

[0011] Processing and analyzing the three-dimensional image data to extract structural parameters of the quartz sand micro-proppant;

[0012] Based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm;

[0013] Based on the constructed pore network model, the flow of fluid in the pore network of microproppant is simulated, and the seepage characteristics and conductivity of microproppant are calculated.

[0014] Preferably, processing and analyzing the three-dimensional image data includes: performing denoising, threshold segmentation and edge detection processing on the three-dimensional image data in sequence.

[0015] Preferably, the process of performing threshold segmentation processing on the three-dimensional image data includes:

[0016] Dividing the three-dimensional image data into a plurality of non-overlapping sub-regions;

[0017] For each sub-region, a local adaptive threshold is calculated according to the grayscale statistics of the pixels in the sub-region;

[0018] The connected domains in the threshold segmentation image are obtained, and the areas of the connected domains are counted to obtain the final target area.

[0019] Preferably, the edge detection process includes: performing morphological opening and closing operations on the final target area;

[0020] The edge tracking algorithm is used to determine the position and direction of the edge of the final target area to obtain the edge detection result.

[0021] Preferably, based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm, including: taking the structural parameters as feature vectors and collecting corresponding label data to construct a sample data set for supervised learning;

[0022] Using a random forest algorithm to perform Bootstrap random sampling on the sample data set to generate multiple data subsets;

[0023] According to the importance parameter value of each sample feature, the extraction probability of each sample feature is determined;

[0024] According to the extraction probability of each sample feature, sample features are randomly extracted to generate multiple feature subsets;

[0025] Randomly select a data subset and a feature subset, and use the CART feature splitting method based on the selected feature subset to split the root node into leaf nodes and build a decision tree;

[0026] Repeat the above steps to construct multiple decision trees, and then combine the multiple decision trees into a random forest to build a pore network model based on the random forest.

[0027] Preferably, simulating the flow of fluid in the microproppant pore network based on the constructed pore network model includes:

[0028] The finite element method is used to solve the flow equations of the fluid in the pore network model;

[0029] According to the set fluid properties and boundary conditions, the equation of fluid seepage in porous media is obtained;

[0030] The iterative method is used to solve the discretized equations to obtain the pressure distribution and flow velocity distribution at each node in the pore network model.

[0031] Preferably, calculating the flow characteristics and conductivity of the microproppant includes:

[0032] Based on the pressure and velocity distribution results obtained from the simulation, the permeability is calculated using Darcy's law;

[0033] According to the simulated permeability and the geometric parameters of the set pore network model, the formula for conductivity is:

[0034]

[0035] Where h is the thickness of the pore network, ΔL is the distance the fluid flows,

[0036] Preferably, a quartz sand micro-proppant conductivity simulation system comprises:

[0037] A sample acquisition module, used for acquiring a quartz sand micro-proppant sample and pre-treating the quartz sand micro-proppant sample;

[0038] A three-dimensional imaging module, used to perform three-dimensional imaging on the quartz sand micro-proppant through a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant;

[0039] A parameter extraction module, used to process and analyze the three-dimensional image data to extract the structural parameters of the quartz sand micro-proppant;

[0040] A model building module, for building a pore network model of the microproppant using a machine learning algorithm based on the structural parameters;

[0041] The calculation module is used to simulate the flow of fluid in the micro-proppant pore network based on the constructed pore network model and calculate the seepage characteristics and conductivity of the micro-proppant.

[0042] It can be known from the above technical scheme that compared with the prior art, the present invention discloses a method and system for simulating the flow conductivity of quartz sand microproppants, including: obtaining a quartz sand microproppant sample and pre-processing the quartz sand microproppant sample; performing three-dimensional imaging on the quartz sand microproppant through a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand microproppant; processing and analyzing the three-dimensional image data to extract the structural parameters of the quartz sand microproppant; based on the structural parameters, constructing a pore network model of the microproppant using a machine learning algorithm; simulating the flow of fluid in the pore network of the microproppant based on the constructed pore network model, and calculating the seepage characteristics and flow conductivity of the microproppant. The present invention can identify the relationship between the characteristics of the quartz sand microproppant and the flow conductivity, realize automatic learning and optimization of the model, accelerate the research and development of microproppants, improve oil and gas recovery efficiency, and reduce development costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0044] Figure 1 A schematic flow chart of a method for simulating the conductivity of quartz sand micro-proppant provided by the present invention.

[0045] Figure 2 A schematic diagram of the structure of a quartz sand micro-proppant conductivity simulation system provided by the present invention. DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0047] The embodiment of the present invention discloses a method and system for simulating the conductivity of quartz sand micro-proppants, which ensures the accuracy and reliability of the conductivity of micro-proppants from three-dimensional imaging to parameter extraction, model construction and CFD simulation. By comprehensively using image analysis, machine learning and CFD technology, the design is continuously optimized in the research and application of micro-proppants to improve their application efficiency in oil and gas recovery.

[0048] Quartz sand is quartz particles made by crushing quartz stone. Quartz stone is a non-metallic mineral, a hard, wear-resistant, chemically stable silicate mineral. The color of quartz sand is milky white or colorless and translucent, with a Mohs hardness of 7.

[0049] Quartz sand is an important industrial mineral raw material and a non-chemical hazardous product. It is widely used in industries such as glass, casting, ceramics and fireproof materials, smelting ferrosilicon, metallurgical flux, metallurgy, construction, chemicals, plastics, rubber, abrasives and filter materials.

[0050] The embodiment of the present invention discloses a method for simulating the conductivity of quartz sand micro-proppant, such as Figure 1 As shown, including:

[0051] Obtaining a quartz sand micro-proppant sample and pre-treating the quartz sand micro-proppant sample; the pre-treating of the quartz sand micro-proppant sample includes washing, drying, etc., to remove impurities and ensure the stability of the sample during the imaging process;

[0052] Performing three-dimensional imaging on the quartz sand micro-proppant by using a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant;

[0053] Processing and analyzing the three-dimensional image data to extract structural parameters of the quartz sand micro-proppant;

[0054] Based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm;

[0055] Based on the constructed pore network model, the flow of fluid in the pore network of microproppant is simulated, and the seepage characteristics and conductivity of microproppant are calculated.

[0056] Specifically, processing and analyzing the three-dimensional image data includes: performing denoising, threshold segmentation and edge detection processing on the three-dimensional image data in sequence.

[0057] The three-dimensional image data is subjected to denoising and contrast enhancement processing by using Gaussian filtering and automatic gamma correction, including:

[0058] The weight matrix is ​​taken from Gaussian distribution, and the space target observation image is convolved with the matrix. Each pixel in the output image takes the weighted average of the surrounding pixels to complete Gaussian filtering. The "center point" of the weight matrix is ​​taken as the origin, and the initial weights of other points in the neighborhood are calculated according to the Gaussian distribution. The two-dimensional Gaussian distribution is:

[0059]

[0060] Among them, x, y, z are the three-dimensional spatial coordinates, and σ is the standard deviation of the Gaussian distribution; it determines the width of the Gaussian kernel and controls the degree of filtering. By convolving this Gaussian kernel with each voxel in the three-dimensional image data, the Gaussian noise in the image can be effectively removed.

[0061] After obtaining the Gaussian weight matrix G, the three-dimensional image data I is convolved with G. The specific formula is:

[0062]

[0063] in, represents the convolution operation, I G Output image for convolution;

[0064] For the denoised image, automatic gamma correction is used to achieve contrast enhancement, improve the contrast of the image, and make the details in the image more clearly visible.

[0065] First calculate the average gray value μ of the image:

[0066]

[0067] Where N is the total number of voxels in the image, p i is the gray value of the i-th voxel. Then the gamma value γ is calculated based on the average gray value:

[0068]

[0069] Finally, the gray value p of each voxel in the image i To perform gamma correction:

[0070]

[0071] The gamma value is automatically adjusted according to the grayscale distribution characteristics of the image itself to enhance the image contrast and highlight the structural characteristics of the quartz sand micro-proppant.

[0072] Specifically, the process of performing threshold segmentation processing on the three-dimensional image data includes:

[0073] The three-dimensional image data is divided into a number of non-overlapping sub-areas; the shape of the sub-area can be a cube or a cuboid, and the size of the division is determined according to the resolution of the image and the size of the target; the side length of the sub-area in the embodiment of the present invention is between 10 and 50 voxels.

[0074] For each sub-region, a local adaptive threshold is calculated according to the grayscale statistics of the pixels in the sub-region;

[0075] The local adaptive threshold is based on a threshold calculation method of a local mean and a standard deviation, where threshold=local mean+k*local standard deviation, and k is a coefficient determined based on experience or experiments.

[0076] For each sub-region, the mean and standard deviation of the pixel grayscale values ​​in the sub-region are first calculated. Then, the local adaptive threshold is calculated based on the mean and standard deviation. In different sub-regions, the appropriate threshold can be determined according to its own grayscale changes, so as to better segment the quartz sand micro-proppant from the background.

[0077] The connected domains in the threshold segmentation image are obtained, and the areas of the connected domains are counted to obtain the final target area.

[0078] The connected domains of the image after threshold segmentation are marked, and each connected domain in the image is marked, and then the area of ​​each connected domain, that is, the number of voxels in the connected domain, is counted. An area threshold is set. When the area of ​​a connected domain is less than the area threshold, the pixels in the connected domain are set as background pixels (for example, the gray value is set to 0) to remove some small noise or interference connected domains. After that, a closed operation is used to dilate and erode the image using a three-dimensional structural element (such as a cubic structural element) to connect the remaining target connected domains into a larger connected domain. The connected domain with the largest area is selected as the final target area, which is the main distribution area of ​​quartz sand micro-proppant in the image.

[0079] Specifically, the edge detection process includes: performing morphological opening and closing operations on the final target area;

[0080] When the area of ​​the final target region is larger than the set area threshold for particle segmentation, the circumscribed rectangle of the connected domain is obtained based on the edge point set of the target in the connected domain, the image block of the region of interest is extracted, and the threshold segmentation is performed. The morphological opening and closing operations are used to process the connections and holes between particles.

[0081] For the opening operation, set the appropriate size of the rectangular structure element. The radius size m of the height is the value obtained by multiplying the width W of the region of interest by a certain coefficient and rounding up. The radius size n of the width is the value obtained by multiplying the height H of the region of interest by a certain coefficient and rounding up. The size of the rectangular structure element is (2m+1)*(2n+1). The opening operation smoothes the contours of the object, disconnects the narrow necks, and eliminates thin protrusions, weakening the connection between particles.

[0082] The result after the opening operation is closed. The size of the rectangular structure element is determined in a similar way to the opening operation. The holes in the target particles and the breaks in the contour lines are filled by the closing operation to obtain independent regions between the particles. The boundaries of these regions are the edges of the target particles. In this process, the size of the structure element is adaptively changed by adjusting the coefficients of the height and width of the rectangular structure elements of the opening and closing operations to meet the edge detection requirements of target particles of different sizes and shapes.

[0083] The edge tracking algorithm is used to determine the position and direction of the edge of the final target area to obtain the edge detection result.

[0084] For example, the edge tracking algorithm based on chain code. Starting from an edge point in the image, search for adjacent edge points in a certain direction order (for example, clockwise or counterclockwise), and record their coordinates. During the search process, the edge point is determined according to the change in the grayscale value of the pixel. When the pixel transitions from the inside of the target area to the background pixel, the edge point is considered to be found. By continuously tracking the edge points, a chain code sequence of edge points is formed, thereby determining the position and direction of the edge of the final target area, and obtaining accurate edge detection results. The shape, size and other structural parameters of the quartz sand microproppant are extracted through the edge detection results, providing basic data for further construction of the pore network model and analysis of the drainage capacity.

[0085] During the entire processing process, various parameters need to be continuously adjusted and optimized according to the actual processing results to adapt to the processing requirements of quartz sand micro-proppant three-dimensional image data of different batches and different characteristics, ensuring accurate and reliable processing results.

[0086] Specifically, based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm, including: taking the structural parameters as feature vectors, and collecting corresponding label data to construct a sample data set (such as physical properties, flow characteristics, etc.) for supervised learning;

[0087] The sample data set covers microproppants with different structural parameters;

[0088] Using a random forest algorithm to perform Bootstrap random sampling on the sample data set to generate multiple data subsets;

[0089] According to the importance parameter value of each sample feature, the extraction probability of each sample feature is determined;

[0090] According to the extraction probability of each sample feature, sample features are randomly extracted to generate multiple feature subsets;

[0091] Randomly select a data subset and a feature subset, and select the CART feature splitting method based on the selected feature subset, with the goal of minimizing the node impurity after splitting (such as the mean square error for regression problems and the Gini index for classification problems), split the root node into leaf nodes, and construct a decision tree;

[0092] Repeat the above steps to construct multiple decision trees, and then combine the multiple decision trees into a random forest to build a pore network model based on the random forest.

[0093] During prediction, for a new input sample (i.e., the structural parameter feature vector of quartz sand microproppant), it is input into each decision tree in the random forest. Each decision tree will give a prediction result (such as the predicted value of permeability or conductivity). These prediction results are then integrated to achieve the prediction and analysis of the pore network model of the microproppant.

[0094] In practical applications, it is necessary to tune the hyperparameters of the random forest model (such as the number of decision trees, the sampling ratio of data subsets, the number of feature subsets extracted, etc.), and find the optimal hyperparameter combination through cross-validation and other methods to improve the performance and accuracy of the model. At the same time, the constructed model is evaluated and verified, and compared with actual experimental data or known theoretical results to further optimize the model so that it can better reflect the pore network characteristics and related physical processes of quartz sand microproppants.

[0095] Specifically, based on the constructed pore network model, the flow of fluid in the micro-proppant pore network is simulated, including:

[0096] The finite element method is used to solve the flow equations of the fluid in the pore network model;

[0097] The numerical simulation method described in the embodiment of the present invention adopts the finite element method, which divides the model into a finite number of units, establishes the equilibrium equations on the unit nodes through the variational principle, and then assembles them into the overall equilibrium equation for solution. For pore network structures with complex shapes, the finite element method has better adaptability.

[0098] According to the set fluid properties (the fluid properties include fluid viscosity) and boundary conditions, an equation for fluid seepage in a porous medium is obtained;

[0099] The iterative method is used to solve the discretized equations to obtain the pressure distribution and flow velocity distribution at each node in the pore network model.

[0100] The iterative method is used to solve the linear equations after finite element discretization until the convergence condition is met (such as the difference between two adjacent iterative results is less than the set threshold).

[0101] Specifically, the seepage characteristics and conductivity of the microproppant are calculated, including:

[0102] Based on the pressure and velocity distribution results obtained from the simulation, the permeability is calculated using Darcy's law;

[0103] Permeability is a key seepage characteristic parameter, which can be back-calculated from the simulation results by Darcy's law, namely

[0104]

[0105] Where v is the seepage velocity, K is the permeability, is the pressure gradient, μ is the fluid viscosity;

[0106] According to the simulated permeability and the geometric parameters of the set pore network model, the formula for conductivity is:

[0107]

[0108] Where h represents the thickness of the pore network and ΔL is the distance the fluid flows;

[0109] This will evaluate the conductivity of quartz sand microproppant under specific fluid conditions and provide a theoretical basis and performance prediction for its application in fields such as petroleum engineering.

[0110] During the entire implementation process, it is necessary to continuously verify and optimize each link of model construction and simulation calculation, and compare and analyze with actual experimental data to improve the accuracy and reliability of the model, so as to better reflect the actual seepage and conductivity characteristics of quartz sand microproppants.

[0111] The embodiment of the present invention discloses a quartz sand micro-proppant conductivity simulation system, such as Figure 2 As shown, it includes: a sample acquisition module, used to obtain a quartz sand micro-proppant sample and pre-treat the quartz sand micro-proppant sample;

[0112] A three-dimensional imaging module, used to perform three-dimensional imaging on the quartz sand micro-proppant through a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant;

[0113] A parameter extraction module, used to process and analyze the three-dimensional image data to extract the structural parameters of the quartz sand micro-proppant;

[0114] A model building module, for building a pore network model of the microproppant using a machine learning algorithm based on the structural parameters;

[0115] The calculation module is used to simulate the flow of fluid in the micro-proppant pore network based on the constructed pore network model and calculate the seepage characteristics and conductivity of the micro-proppant.

[0116] In a specific embodiment of the present invention, a method for simulating the conductivity of quartz sand microproppant comprises:

[0117] 1. Sample preparation:

[0118] Obtain a batch of quartz sand micro-proppant samples, place them in a beaker, add an appropriate amount of deionized water, stir and wash for 10-15 minutes, then filter with filter paper, put the filtered samples into an oven at 70°C to dry, and after drying, use 20-mesh and 40-mesh standard sieves to sieve, and collect quartz sand micro-proppant samples between 20-40 mesh for later use.

[0119] 2. Three-dimensional imaging:

[0120] The sample was imaged using an X-ray micro-focus CT scanning device. The sample was placed on the sample stage of the device, and the X-ray tube voltage was set to 80 kV, the current was 120 μA, the scanning step was 0.05 mm, and the scanning angle range was 0-360 degrees. A full-range scan was performed to obtain the three-dimensional image data of the quartz sand micro-proppant, and the data was stored in DICOM format.

[0121] X-ray micro-focus CT scanning equipment can penetrate the sample and scan and image from different angles, providing high-resolution (such as micron level) three-dimensional image data, clearly presenting the internal structure of quartz sand micro-proppant, including particle morphology and pore distribution.

[0122] 3. Image data processing and analysis:

[0123] The median filtering algorithm was used to reduce the noise of the three-dimensional image data, and the image enhancement algorithm was used to adjust the contrast and brightness. Then, the threshold segmentation algorithm was used to set the threshold to 120 (grayscale value) according to the distribution of image pixel values ​​to separate the micro-proppant particles from the pore part to obtain a binary image.

[0124] The particle size distribution of the microproppant particles is calculated by image analysis software, and the number and proportion of particles in different particle size ranges (such as 20-25 mesh, 25-30 mesh, 30-35 mesh, and 35-40 mesh) are counted; the porosity is measured, and the proportion of the pore volume to the total volume is calculated; the pore connectivity is analyzed, and the embodiment of the present invention determines that the number of main connecting paths is 5, and the average diameter is 0.2 mm; the pore shape parameters are obtained, and the average equivalent spherical diameter of the embodiment of the present invention is 0.15 mm, and the average aspect ratio is 1.2.

[0125] 4. Constructing the pore network model:

[0126] The pore network model was constructed using a machine learning algorithm. The particle size distribution, porosity, pore connectivity and pore shape parameters calculated above were used as input data. Through training and learning of a large amount of sample data, the network weights were adjusted to construct a model that can accurately reflect the pore network structure of the microproppant.

[0127] 5. Simulate fluid flow and performance calculation:

[0128] Import the pore network model into the CFD software, set the physical properties of the fluid (such as viscosity, density, etc.) and boundary conditions, perform fluid flow simulation, calculate the seepage characteristics and conductivity of the micro-proppant layer, and analyze the flow path and flow velocity distribution.

[0129] The method of the embodiment of the present invention can efficiently and accurately simulate the conductivity of quartz sand microproppants, provide a scientific basis for oil and gas production, and significantly improve the application efficiency and economic benefits of microproppants by combining three-dimensional imaging, data processing and machine learning technology.

[0130] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0131] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be 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 will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for simulating the conductivity of quartz sand microproppant, characterized in that: include: Obtaining a quartz sand micro-proppant sample, and pre-treating the quartz sand micro-proppant sample; Performing three-dimensional imaging on the quartz sand micro-proppant by using a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant; Processing and analyzing the three-dimensional image data to extract structural parameters of the quartz sand micro-proppant; Based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm; Based on the constructed pore network model, the flow of fluid in the pore network of microproppant is simulated, and the seepage characteristics and conductivity of microproppant are calculated.

2. A method for simulating the conductivity of quartz sand microproppant according to claim 1, characterized in that: Processing and analyzing the three-dimensional image data includes: performing denoising, threshold segmentation and edge detection processing on the three-dimensional image data in sequence.

3. A method for simulating the conductivity of quartz sand microproppant according to claim 2, characterized in that: The process of threshold segmentation of three-dimensional image data includes: Dividing the three-dimensional image data into a plurality of non-overlapping sub-regions; For each sub-region, a local adaptive threshold is calculated according to the grayscale statistics of the pixels in the sub-region; The connected domains in the threshold segmentation image are obtained, and the areas of the connected domains are counted to obtain the final target area.

4. A method for simulating the conductivity of quartz sand microproppant according to claim 3, characterized in that: The edge detection process includes: performing morphological opening and closing operations on the final target area; The edge tracking algorithm is used to determine the position and direction of the edge of the final target area to obtain the edge detection result.

5. The method for simulating the conductivity of quartz sand microproppant according to claim 1, characterized in that: Based on the structural parameters, a pore network model of the microproppant is constructed using a machine learning algorithm, including: taking the structural parameters as feature vectors and collecting corresponding label data to construct a sample data set for supervised learning; Using a random forest algorithm to perform Bootstrap random sampling on the sample data set to generate multiple data subsets; According to the importance parameter value of each sample feature, the extraction probability of each sample feature is determined; According to the extraction probability of each sample feature, sample features are randomly extracted to generate multiple feature subsets; Randomly select a data subset and a feature subset, and use the CART feature splitting method based on the selected feature subset to split the root node into leaf nodes and build a decision tree; Repeat the above steps to construct multiple decision trees, and then combine the multiple decision trees into a random forest to build a pore network model based on the random forest.

6. The method for simulating the conductivity of quartz sand microproppant according to claim 1, characterized in that: Based on the constructed pore network model, the flow of fluid in the micro-proppant pore network is simulated, including: The finite element method is used to solve the flow equations of the fluid in the pore network model; According to the set fluid properties and boundary conditions, the equation of fluid seepage in porous media is obtained; The iterative method is used to solve the discretized equations to obtain the pressure distribution and flow velocity distribution at each node in the pore network model.

7. A method for simulating the conductivity of quartz sand microproppant according to claim 6, characterized in that: Calculate the flow characteristics and conductivity of microproppants, including: Based on the pressure and velocity distribution results obtained from the simulation, the permeability is calculated using Darcy's law; According to the simulated permeability and the geometric parameters of the set pore network model, the formula for conductivity is: Here, h represents the thickness of the pore network and ΔL is the distance the fluid flows.

8. A quartz sand micro-proppant conductivity simulation system, characterized in that: include: A sample acquisition module, used for acquiring a quartz sand micro-proppant sample and pre-treating the quartz sand micro-proppant sample; A three-dimensional imaging module, used for performing three-dimensional imaging on the quartz sand micro-proppant through a three-dimensional imaging device to obtain three-dimensional image data of the quartz sand micro-proppant; A parameter extraction module, used to process and analyze the three-dimensional image data to extract the structural parameters of the quartz sand micro-proppant; A model building module, for building a pore network model of the microproppant using a machine learning algorithm based on the structural parameters; The calculation module is used to simulate the flow of fluid in the micro-proppant pore network based on the constructed pore network model and calculate the seepage characteristics and conductivity of the micro-proppant.

Images