Three-dimensional seam network reconstruction method and device, storage medium and electronic equipment

Abstract

The invention provides a three-dimensional seam network reconstruction method and device, a storage medium and electronic equipment. For the microseismic point cloud of each fracturing section of the target well, randomly extracting a microseismic event point, and intercepting a first neighborhood point set; performing azimuth information identification of the crack based on the first neighborhood point set, and performing preliminary geometrical shape identification of the current crack when the current crack meets the requirement to obtain a second neighborhood point set corresponding to the preliminary geometrical shape boundary of the current crack; clustering the microseismic event points in the second neighborhood point set, identifying and removing noisy points to obtain a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain a final geometric shape of the current crack; and combining the fractures of the fractured sections of the target well based on the azimuth information and the final geometrical shape to form a reconstructed three-dimensional fracture network of the target well. According to the method, the interference of noisy points in a micro-seismic event on the fracture network can be effectively overcome, and a high-precision fracture three-dimensional complex fracture network reconstruction result is obtained.

Description

[0001] Technology Neighborhood

[0002] The invention relates to the technical field of oil and gas reservoir development, and in particular to a three-dimensional fracture network reconstruction method, device, storage medium and electronic equipment. Background Art

[0003] Shale gas reservoirs are characterized by low permeability and low porosity, and hydraulic fracturing is required for production enhancement. Due to the heterogeneity of shale reservoirs, the distribution of natural fractures and stress fields is very complex, and the distribution of complex fracture networks after hydraulic fracturing is difficult to predict and characterize, resulting in a lack of reliable decision-making basis for the effectiveness and construction optimization of shale gas hydraulic fracturing.

[0004] According to different ideas of constructing fracture networks, the current three-dimensional fracture network construction methods are mainly divided into fracture simulation methods and fracture network reconstruction methods.

[0005] The pre-fracturing fracture network simulation method simplifies the basic relationship within the fracture network into the connectivity relationship between fractures. After obtaining the distribution of natural fractures, the expansion path of hydraulic fractures in the natural fracture network is simulated based on the hydraulic-natural fracture expansion criterion and the corresponding numerical calculation method. After correction of actual construction data, the three-dimensional complex fracture network of fracturing is obtained. Currently, the commonly used numerical methods include finite element method, extended finite element method, displacement discontinuity method, discrete element method and phase field simulation method developed in recent years. The accuracy of the fracture expansion simulation method depends on the accurate description of the global natural fractures and rock mechanical properties, which is difficult to obtain in actual construction, resulting in the limited accuracy and reliability of the complex fracture network obtained by the fracture simulation method.

[0006] The fracture network reconstruction method simplifies the basic relationship within the fracture network into the topological relationship between event points and fracture surfaces. After obtaining the microseismic monitoring event points in the multi-stage fracturing process, a high-robustness regression method is used to directly identify the three-dimensional complex fracture network from the event points through a reasonable fracture network-event point distribution relationship assumption. The advantage of the fracture network reconstruction method is that the fracture network is directly constructed from the fracturing monitoring results, and the obtained three-dimensional fracture network is basically consistent with the monitoring results. However, due to the presence of a considerable number of invalid noise points that cannot be removed in the microseismic event points, the reconstructed fracture network and the event points have a poor fit. Therefore, this neighborhood urgently needs to design a three-dimensional fracture network reconstruction method with high noise resistance to effectively overcome the interference of invalid noise points in microseismic event points, so as to obtain high-precision fracturing complex fracture network reconstruction results. Summary of the invention

[0007] In order to solve the above problems, the present invention provides a three-dimensional seam network reconstruction method, device, storage medium and electronic device.

[0008] In a first aspect, an embodiment of the present invention provides a three-dimensional seam network reconstruction method, comprising:

[0009] For the microseismic point cloud of each fracturing section of the target well, a microseismic event point is randomly selected, and the first neighborhood point set of the randomly selected microseismic event point is intercepted;

[0010] Identify the orientation information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the orientation information, a preset fit index, and a preset distance index;

[0011] When the current crack meets the requirements, the preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack;

[0012] Clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein, obtaining a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture, wherein the noise points are microseismic event points in the second neighborhood point set whose density values ​​in the area are less than a preset core point density threshold and which do not fall within the neighborhood radius of any core point;

[0013] The fractures of each fracturing stage of the target well are combined based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

[0014] In some implementations, intercepting a first neighborhood point set of a randomly selected microseismic event point includes:

[0015] Calculate the distance between the remaining microseismic event points in the microseismic event point cloud of the current fracturing section and the randomly selected microseismic event point;

[0016] Based on the distance, a first neighborhood point set of the randomly selected microseismic event point is intercepted.

[0017] In some implementations, the identifying the orientation information of the crack based on the first neighborhood point set, and determining whether the current crack meets the requirements according to the orientation information, a preset fit index, and a preset distance index, includes:

[0018] Fitting the intercept formula of the crack plane based on the first neighborhood point set to obtain the orientation information of the crack;

[0019] According to the azimuth information, the number of microseismic event points in the first neighborhood point set that meet a preset distance index is counted to obtain a fit degree of the first neighborhood point set;

[0020] If the fit meets the preset fit index, it is determined that the current crack meets the requirement.

[0021] In some implementations, the orientation information includes the direction, inclination and dip of the fracture, and the calculation formula is as follows:

[0022]

[0023]

[0024]

[0025] Among them, λ1 represents the direction of the crack, λ2 represents the inclination of the crack, λ3 represents the tendency of the crack, and a, b, and c represent the intercepts of the crack on the x, y, and z axes, respectively.

[0026] In some implementations, when the current crack meets the requirements, a preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack, including:

[0027] When the current fracture meets the requirements, extract the microseismic event points in the first neighborhood point set that meet the preset fit index to obtain the fracture neighborhood point set;

[0028] Projecting the crack neighborhood point set onto the plane where the current crack is located to obtain a projection point set of the current crack;

[0029] A disk with a set radius is used to roll at the microseismic event points at the boundary of the projection point set, and two adjacent microseismic event points at the boundary in contact with the disk are sequentially connected as a boundary segment to obtain the preliminary geometric shape boundary of the current crack and the corresponding second neighborhood point set.

[0030] In some implementations, clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein to obtain a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain a final geometric shape of the current fracture includes:

[0031] According to the set neighborhood radius and core point density threshold, the microseismic event points in the second neighborhood point set are clustered to obtain a valid point set and a noise point set. The valid point set includes a core point set and a boundary point set. The core point is a microseismic event point in the second neighborhood point set whose regional density value is not less than the preset core point density threshold, and the boundary point is a microseismic event point whose regional density value is less than the neighborhood radius and falls within the neighborhood radius of other core points.

[0032] In some implementations, after obtaining the final geometric shape of the current crack, the method further includes:

[0033] Removing the fracture neighborhood point set from the microseismic point cloud of the current fracturing section to obtain a remaining point set;

[0034] The remaining point set is used as a new first neighborhood point set, and the orientation information of the crack is identified based on the first neighborhood point set. According to the orientation information, a preset fit index and a preset distance index, it is determined whether the current crack meets the requirements, and the orientation information of the next crack is identified to obtain the final geometric shape of the next crack until the number of microseismic event points in the remaining point set is less than the set value.

[0035] In a second aspect, an embodiment of the present invention provides a three-dimensional seam network reconstruction device, comprising:

[0036] A random extraction module is used to randomly extract a microseismic event point for each fracturing section of the target well, and intercept the first neighborhood point set of the randomly extracted microseismic event point;

[0037] an orientation recognition module, configured to recognize the orientation information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the orientation information, a preset fit index and a preset distance index;

[0038] A shape recognition module is used to perform preliminary geometric shape recognition of the current crack when the current crack meets the requirements, and obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack;

[0039] A denoising processing module is used to cluster the microseismic event points in the second neighborhood point set, identify and remove the noise points therein, obtain a third neighborhood point set, and connect the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture, wherein the noise points are microseismic event points in the second neighborhood point set whose density value in the area is less than a preset core point density threshold and do not fall within the neighborhood radius of any core point;

[0040] A fracture network reconstruction module is used to combine the fractures of each fracturing section of the target well based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

[0041] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by at least one processor, the method described in the first aspect is implemented.

[0042] In a fourth aspect, an embodiment of the present invention provides an electronic device, comprising a memory and at least one processor, wherein the memory stores a computer program, and when the computer program is executed by the at least one processor, the method described in the first aspect is implemented.

[0043] Beneficial effects:

[0044] In view of the problem that the current fracture network reconstruction method has insufficient noise resistance, the present invention establishes a three-dimensional complex fracture network reconstruction method with high noise resistance, which can effectively overcome the interference of noise points in microseismic events on the fracture network, realize the effective identification of the three-dimensional fracture orientation and shape of the point cloud of complex structure microseismic events, and obtain high-precision fracturing three-dimensional complex fracture network reconstruction results, which provides a more reliable decision-making basis for fracturing construction optimization, for example, it can provide a reliable evaluation basis for the fine diagnosis and real-time regulation of hydraulic fracturing construction, and can also provide key support for the integrated technology of hydraulic fracturing geology and engineering in unconventional reservoirs. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and should not be regarded as limiting the scope.

[0046] Figure 1 is a flow chart of a three-dimensional seam network reconstruction method provided by an embodiment of the present invention;

[0047] Figure 2 It is a schematic diagram of a three-dimensional seam network reconstruction process provided by an embodiment of the present invention;

[0048] Figure 3 : is a schematic diagram of the distribution of microseismic monitoring event points of fracturing in Well A provided by an embodiment of the present invention;

[0049] Figure 4 is a schematic diagram of the reconstruction result of the three-dimensional complex fracture network of Well A provided by an embodiment of the present invention;

[0050] Figure 5 It is a comparison of the fit between the microseismic event point of Well A and the three-dimensional complex fracture network provided by an embodiment of the present invention;

[0051] Figure 6 It is a block diagram of a three-dimensional seam network reconstruction device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The components of the embodiments of the present invention generally described and represented in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present invention.

[0053] Embodiment 1

[0054] This embodiment provides a three-dimensional seam network reconstruction method. Figure 1 As shown, including:

[0055] Step S101: for the microseismic point cloud of each fracturing section of the target well, a microseismic event point is randomly selected, and the first neighborhood point set of the randomly selected microseismic event point is intercepted.

[0056] In some implementations, intercepting a first neighborhood point set of a randomly selected microseismic event point includes:

[0057] Step S101a, calculating the distance between the remaining microseismic event points in the microseismic event point cloud of the current fracturing section and the randomly selected microseismic event point.

[0058] In practical applications, the set of microseismic point clouds of the target single well can be identified as M, then

[0059] M={M I ,M o}={M1,M2,M3,...,M N}

[0060] Among them, M I represents a subset of valid points, M o represents a subset of noise points, M N It represents the subset of microseismic event points of the Nth fracturing stage, that is, the microseismic event point cloud of the Nth fracturing stage, where N represents the number of fracturing stages of the target well.

[0061] The distance in this embodiment may be but is not limited to the Euclidean distance. A microseismic event point p is extracted from the microseismic event point subset M1 of the current fracturing stage by a simple random sampling method, and the Euclidean distances from the remaining microseismic event points in the microseismic event point subset to p are calculated.

[0062] Step S101b: based on the distance, intercept the first neighborhood point set of the randomly selected microseismic event point.

[0063] After sorting the Euclidean distances, a preset number of microseismic event points closest to the microseismic event point are obtained to obtain a random neighborhood point set P as the first neighborhood point set, thus completing the sampling of the random neighborhood point set. The preset number can be any integer between 5 and 10.

[0064] Step S102: Identify the orientation information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the orientation information, a preset fit index and a preset distance index.

[0065] In this embodiment, the first neighborhood point set obtained in step S101 can be substituted into the highly noise-resistant RANSAC (Random Sample Consensus) algorithm to identify the orientation information of a single crack to obtain preliminary orientation information of a single crack. The orientation information may include the direction, inclination and dip of the crack.

[0066] Identifying the orientation information of the crack based on the first neighborhood point set, and determining whether the current crack meets the requirements according to the orientation information, a preset fit index and a preset distance index, may further include:

[0067] Step S102a: fitting the intercept formula of the crack plane based on the first neighborhood point set to obtain the orientation information of the crack.

[0068] The fracture surface can be expressed by the intercept formula as follows:

[0069]

[0070] Among them, a, b, and c represent the intercepts of the crack on the x, y, and z coordinate axes, respectively, and θ represents the azimuth vector of the crack plane.

[0071] The intercepts of the cracks on the x, y, and z coordinate axes are obtained by fitting the microseismic event points m(x, y, z) in the first neighborhood point set, and then the azimuth vector of the cracks is obtained.

[0072] Step S102b: count the number of microseismic event points in the first neighborhood point set that meet a preset distance index according to the azimuth information to obtain the fit degree of the first neighborhood point set.

[0073] The orientation vector is verified by the preset fit index and preset distance index of the RANSAC algorithm:

[0074] The Euclidean distance from any event point in the first neighborhood point set to the crack plane can be calculated by the intercept:

[0075]

[0076] Among them, f(m i ; θ) represents the microseismic event point m i Euclidean distance to the fracture plane, microseismic event point m i represents any microseismic event point after removing the first neighborhood point set in M1, that is, m i ∈M1-P.

[0077] The fit degree SN of the first neighborhood point set can be obtained by counting the number of microseismic event points that meet the preset distance index:

[0078]

[0079] Among them, c(m i ; θ) represents the judgment function, the preset distance index is the preset distance threshold τ, if the Euclidean distance is less than τ, it is a microseismic event point that meets the preset distance index, marked as 1, otherwise, it is marked as 0, thus, the number SN of microseismic event points that meet the preset distance index can be counted.

[0080] Step S102c: If the fit meets the preset fit index, it is determined that the current crack meets the requirement.

[0081] SN is compared with a preset fit index sn. If SN≥sn, the currently obtained crack is a crack that meets the fit standard. Otherwise, the process returns to step S101 and re-extracts a neighborhood point set until a crack that meets the fit standard is found.

[0082] In some implementations, the strike, dip, and inclination of the fracture are calculated as follows:

[0083]

[0084]

[0085]

[0086] Among them, λ1 represents the direction of the crack, λ2 represents the inclination of the crack, λ3 represents the tendency of the crack, sgn represents the sign function, the range of the direction λ1 is [0,π], and the range of the inclination λ2 is The range of the inclination λ3 is [0,2π]. It can be seen that after obtaining the azimuth vector, we can Calculate the position information.

[0087] Step S103: When the current crack meets the requirements, the preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack.

[0088] In this embodiment, the Alpha Shape shape recognition algorithm may be used to perform preliminary geometric shape recognition of the current crack to achieve three-dimensional crack shape recognition.

[0089] When the current crack meets the requirements, the preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack, which may further include:

[0090] Step S103a: when the current fracture meets the requirements, extract the microseismic event points in the first neighborhood point set that meet the preset fit index to obtain the fracture neighborhood point set P1.

[0091] Step S103b: Project the crack neighborhood point set onto the plane where the current crack is located to obtain the projection point set V2 of the current crack.

[0092] Specifically, the shrinkage factor of the Alpha Shape shape recognition algorithm is set, and the microseismic event points in the fracture neighborhood point set are vertically projected onto the plane where the fracture identified in step S102 is located, so as to obtain a projection point set corresponding to a single fracture.

[0093] Step S103c, use a disk with a set radius to roll at the microseismic event points at the boundary of the projection point set, and sequentially connect two adjacent microseismic event points at the boundary in contact with the disk into a boundary section to obtain the preliminary geometric shape boundary of the current crack and the corresponding second neighborhood point set P2.

[0094] Specifically, the fracture projection point set is substituted into the Alpha Shape shape recognition algorithm to identify the vertices (microseismic event points) at the fracture boundary, and the vertices are connected in sequence to obtain the geometric shape of the three-dimensional fracture.

[0095] For the projection point set V2, the Alpha Shape algorithm needs to define a disk with a radius of a. When the disk rolls at the boundary of the projection point set V2, it will touch two adjacent boundary points in V2 at the same time (microseismic event points at the contacting boundaries). The boundary point and the disk intersect, and the two boundary points are connected by a straight line and identified as a section of the boundary of the crack.

[0096] The value range of the radius a of the above disk is [a min ,a max ], the size of a is determined by the shrinkage factor k of the polygonal boundary of the projection point set V2, and the radius of the disk can be expressed as:

[0097] a=(1-k)a max +ka min

[0098] Among them: a max Indicates the maximum value of the disk radius corresponding to V2, a min Indicates the minimum value of the disk radius corresponding to V2.

[0099] In one example, setting the shrinkage factor k to 0.5, we can get

[0100] Substituting the projection point set V2 and the disk radius a into the Alpha Shape algorithm to identify the three-dimensional fracture boundary (set shape), we can obtain the second neighborhood point set V1 corresponding to the geometric shape boundary points of the polygonal fracture. The microseismic event points in V1 are connected in sequence to obtain the initial geometric shape of the three-dimensional fracture.

[0101] Step S104, cluster the microseismic event points in the second neighborhood point set, identify and remove the noise points therein, obtain the third neighborhood point set, connect the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture, and the noise points are microseismic event points in the second neighborhood point set whose density values ​​in the area are less than the preset core point density threshold and do not fall within the neighborhood radius of any core point.

[0102] In practical applications, since there is a certain proportion of noise points in the microseismic point cloud that cannot be removed, these noise points appear as isolated and discrete cluster distributions. Since this type of microseismic event points exist at the boundary of the microseismic event point set, the cracks obtained in step S103 may be over-extended, and there are a large number of blank areas of microseismic event points inside the cracks, which reduces the fit between the cracks and the event points, resulting in large calculation errors.

[0103] This embodiment can use the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm to perform cluster analysis and shrinkage processing to further improve the fit between cracks and microseismic event points. Specifically, by integrating the spatial orientation information and geometric shape of the cracks, a complete three-dimensional single crack model can be obtained. The DBSCAN density-based clustering algorithm of non-parametric discriminant analysis is used to denoise the cracks, which can effectively remove the blank areas of microseismic event points in the cracks. By setting the core point density threshold to optimize the boundaries of the cracks, the closeness of the cracks and microseismic event points is further improved, and a three-dimensional single crack model with high fit is obtained.

[0104] In some implementations, clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein to obtain a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain a final geometric shape of the current fracture includes:

[0105] Step S104a, clustering the microseismic event points in the second neighborhood point set according to the set neighborhood radius and core point density threshold to obtain a valid point set and a noise point set, wherein the valid point set includes a core point set and a boundary point set, wherein the core point is a microseismic event point in the second neighborhood point set whose regional density value is not less than the preset core point density threshold, and the boundary point is a microseismic event point whose regional density value is less than the neighborhood radius and falls within the neighborhood radius of other core points.

[0106] In an application example, for the cluster analysis of the second neighborhood point set P2, first, two key parameters of the DBSCAN algorithm are set: the neighborhood radius Eps and the core point density threshold MinPts. According to the density distribution of microseismic event points, the neighborhood radius Eps can be set to 100 and the core point density threshold MinPts can be set to 2.

[0107] The second neighborhood point set P2 in step S103 can be expressed as P2 = {p1, p2, ..., p H}, take any point p∈P2 in the second neighborhood point set, x is any other point in the second neighborhood point set that is different from p, then the Eps neighborhood is the circular area with a neighborhood radius of Eps of point p, which can be expressed as:

[0108] D Eps (p)={x∈P2:d(x,p)≤Eps}

[0109] Wherein, d(x, p) represents the distance between x and p.

[0110] For any point p, the point density of the area where it is located can be defined as the number of microseismic event points falling within its Eps neighborhood, the size of which is determined by the neighborhood radius Eps and is expressed as r(p):

[0111] r(p)=|D Eps (p)|

[0112] For the second neighborhood point set P2, given the neighborhood radius Eps and the core point density threshold MinPts, the core point set P in the second neighborhood point set P2 c It can be expressed as:

[0113] P c ={x∈P2:ρ(x)≥MinPts}

[0114] The boundary point is defined as: the point set density in the area where the microseismic event point is located is less than Eps, and falls within the Eps neighborhood of other core points. The boundary point set P in the second neighborhood point set P2 b It can be expressed as:

[0115]

[0116] Where y represents P c Any point in .

[0117] Noise points are defined as points where the density threshold of the microseismic event point is less than MinPts and does not fall within the Eps neighborhood of any core point. The noise point set P in the second neighborhood point set P2 is n It can be expressed as:

[0118]

[0119] After DBSCAN processing, the second neighborhood point set is divided into three types according to the two parameters of neighborhood radius Eps and core point density threshold MinPts: core point set P c , boundary point set P band the noise set P n , where the core point set P c and the boundary point set P b Can be merged into a valid point set P e , the second neighborhood point set P2 can be divided into valid points P by DBSCAN algorithm. e and noise P n The union of:

[0120] P2=P c +P b +P n =P e +P n

[0121] The noise points in P2 are sparsely distributed around the core points and are identified by the DBSCAN algorithm because of their discrete and low-density features. n Afterwards, the remaining valid point set P in the second neighborhood point set e Substituting into step S103 again, the Alpha Shape shape recognition algorithm is used to re-recognize the fracture boundary points, and the point set corresponding to the vertices of the polygonal fracture geometric boundary after denoising can be obtained. The microseismic event points therein are sequentially connected to obtain the final three-dimensional single fracture geometry.

[0122] After obtaining the orientation information and final geometric shape of the current crack, the method of this embodiment further includes:

[0123] Remove the fracture neighborhood point set from the microseismic point cloud of the current fracturing section to obtain a remaining point set;

[0124] The remaining point set is used as the new first neighborhood point set, and the orientation information of the crack is identified based on the first neighborhood point set. According to the orientation information, the preset fit index and the preset distance index, it is determined whether the current crack meets the requirements, and the orientation information of the next crack is identified to obtain the final geometric shape of the next crack until the number of microseismic event points in the remaining point set is less than the set value.

[0125] To prevent repeated identification, the fracture neighborhood point set P1 corresponding to the current fracture is removed from the microseismic point cloud subset M1 of the current fracturing section to obtain the remaining point set M1 r , where M1 r It can be expressed as:

[0126] M1 r =M1-P1

[0127] M1 rContinue to substitute step S102 to step S104 to identify the orientation information and collective shape of the next fracture, and obtain more single fractures that meet the distance threshold and fit threshold through iteration, until the number of microseismic event points in the remaining point set in M1 is less than the minimum value of microseismic event points required for the specified reconstruction of fractures. Combine all single fractures identified by M1 of the current fracturing stage to form a three-dimensional complex fracture network of a single fracturing stage.

[0128] Step S105: combining the fractures of each fracturing stage of the target well based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

[0129] After the complex fracture network of a single fracturing section is identified, the microseismic point cloud subsets of other fracturing sections in the microseismic point cloud set M of the target single well are sequentially substituted into steps S101 to S105 until the complex fracture networks of all fracturing sections are identified, and the complex fracture networks of single fracturing sections are combined to obtain the complete three-dimensional complex fracture network of single well fracturing.

[0130] The method of this embodiment effectively overcomes the reconstruction error caused by noise points in microseismic events. Figure 2 As shown in the figure, by adopting the RANSAC pattern recognition algorithm, the three-dimensional fracture orientation information with high fit can be obtained. By using the Alpha Shape pattern recognition algorithm and the DBSCAN denoising algorithm, the three-dimensional fracture geometry with high fit can be obtained. By combining and iterating single fractures, the final three-dimensional complex fracture network can be obtained. The fracture network inversion and interpretation of the microseismic monitoring results of fracturing construction are realized, and a reliable analysis basis is provided for post-fracturing construction optimization and fracturing effect evaluation.

[0131] Embodiment 2

[0132] Take shale oil well A in the Biyang Sag area of ​​the Nanxiang Basin in Henan Province as an example. The vertical depth of target point A is 2820.21m, the vertical depth of target point B is 3182.70m, the horizontal section length is 2000m, and well A is divided into 33 sections, with 169 perforation clusters and an average cluster spacing of 8.0-12.0m. The target layer rock of the well has medium elastic modulus and high Poisson's ratio characteristics, and the mechanical brittleness index is low, but natural fractures are developed in local areas along the well. In order to form a complex fracture network through fracturing, well A adopts the "multi-segment multi-cluster + dense cutting" fracturing method to increase the number of fractures, while reducing the cluster spacing to strengthen the interference between fractures, increase the complexity of fractures, and increase the oil and gas leakage area. In order to obtain the geometric parameters and extension orientation of the three-dimensional fractures in Well A during the hydraulic fracturing process, the reservoir fracture signals during the fracturing process were collected by a three-component detector array arranged in advance. After positioning interpretation and inversion, the three-dimensional microseismic event points were obtained. A total of 3305 effective event points were obtained in 33 fracturing sections during the whole fracturing process of Well A. The event points are distributed along the well as shown in the following figure. Figure 3 shown.

[0133] The RANSAC pattern recognition algorithm is used to identify the orientation of the cracks, and the Alpha Shape algorithm is used to identify the three-dimensional crack shape, and then ellipse processing is performed after identification. After obtaining the three-dimensional crack shape and orientation, the DBSCAN algorithm is used to denoise the crack network to improve the fit between the cracks and event points. The parameter settings of each algorithm are shown in Table 1.

[0134] Table 1 Parameter settings of the complex fracture network identification algorithm of Well A

[0135]

[0136] like Figure 4 As shown in the figure, the reconstruction results of the three-dimensional complex fracture network of Well A are obtained based on the microseismic monitoring data. It can be seen from the figure that after the fracturing of Well A, a complex fracture network is basically formed. The complex fracture network is mainly composed of main fractures connected to the wellbore. There are natural fractures at the boundaries and inside of the fracture network that are oblique to the fracture network. These natural fractures lead to the response of large-magnitude event points and increase the complexity of the fracture network.

[0137] Figure 5 A comparison diagram of the reconstructed fracture network and microseismic event points is further given. It can be seen from the figure that the spatial distribution of the reconstructed fracture network is highly consistent with the spatial distribution of the original microseismic event points. The reconstructed three-dimensional complex fracture network and the microseismic monitoring event points have a high degree of fit. The model can effectively overcome the influence of noise points in the microseismic event point cloud and obtain effective complex fracture network reconstruction results.

[0138] Embodiment 3

[0139] This embodiment provides a three-dimensional seam network reconstruction device, such as Figure 6 As shown, including:

[0140] A random extraction module 201 is used to randomly extract a microseismic event point for each fracturing section of the target well, and intercept a first neighborhood point set of the randomly extracted microseismic event point;

[0141] A position identification module 202 is used to identify the position information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the position information, a preset fit index and a preset distance index;

[0142] The shape recognition module 203 is used to perform preliminary geometric shape recognition of the current crack when the current crack meets the requirements, and obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack;

[0143] The denoising processing module 204 is used to cluster the microseismic event points in the second neighborhood point set, identify and remove the noise points therein, obtain a third neighborhood point set, and connect the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture. The noise points are microseismic event points in the second neighborhood point set whose density value in the area is less than a preset core point density threshold and do not fall within the neighborhood radius of any core point.

[0144] The fracture network reconstruction module 205 is used to combine the fractures of each fracturing stage of the target well based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

[0145] In some implementations, intercepting a first neighborhood point set of a randomly selected microseismic event point includes:

[0146] The distance between the remaining microseismic event points in the microseismic event point cloud of the current fracturing section and the randomly selected microseismic event point is calculated. Based on the distance, the first neighborhood point set of the randomly selected microseismic event point is intercepted.

[0147] In some implementations, identifying the orientation information of the crack based on the first neighborhood point set, and determining whether the current crack meets the requirements according to the orientation information, a preset fit index, and a preset distance index may further include:

[0148] The intercept formula of the fracture plane is fitted based on the first neighborhood point set to obtain the orientation information of the fracture. According to the orientation information, the number of microseismic event points in the first neighborhood point set that meet the preset distance index is counted to obtain the fit of the first neighborhood point set. If the fit meets the preset fit index, it is determined that the current fracture meets the requirements.

[0149] In some implementations, the strike, dip, and inclination of the fracture are calculated as follows:

[0150]

[0151]

[0152]

[0153] Among them, λ1 represents the direction of the crack, λ2 represents the inclination of the crack, λ3 represents the tendency of the crack, sgn represents the sign function, the range of the direction λ1 is [0,π], and the range of the inclination λ2 is The range of the inclination λ3 is [0,2π]. It can be seen that after obtaining the azimuth vector, we can Calculate the position information.

[0154] In some implementations, when the current crack meets the requirements, performing preliminary geometric shape recognition of the current crack to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack may further include:

[0155] When the current crack meets the requirements, extract the microseismic event points in the first neighborhood point set that meet the preset fit index to obtain the crack neighborhood point set P1. Project the crack neighborhood point set to the plane where the current crack is located to obtain the projection point set V2 of the current crack. Use a disk with a set radius to roll at the microseismic event points at the boundary of the projection point set, and sequentially connect the two adjacent microseismic event points at the boundary in contact with the disk into a boundary section to obtain the preliminary geometric shape boundary of the current crack and the corresponding second neighborhood point set P2.

[0156] In some implementations, clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein to obtain a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain a final geometric shape of the current fracture includes:

[0157] According to the set neighborhood radius and core point density threshold, the microseismic event points in the second neighborhood point set are clustered to obtain a valid point set and a noise point set. The valid point set includes a core point set and a boundary point set. The core point is a microseismic event point in the second neighborhood point set whose regional density value is not less than the preset core point density threshold. The boundary point is a microseismic event point whose regional density value is less than the neighborhood radius and falls within the neighborhood radius of other core points.

[0158] After obtaining the orientation information and final geometric shape of the current crack, the orientation recognition module 202, shape recognition module 203 and denoising processing module 204 of this embodiment are further used to:

[0159] Remove the fracture neighborhood point set from the microseismic point cloud of the current fracturing section to obtain a remaining point set;

[0160] The remaining point set is used as the new first neighborhood point set, and the orientation information of the crack is identified based on the first neighborhood point set. According to the orientation information, the preset fit index and the preset distance index, it is determined whether the current crack meets the requirements, and the orientation information of the next crack is identified to obtain the final geometric shape of the next crack until the number of microseismic event points in the remaining point set is less than the set value.

[0161] The specific implementation of each module can be found in the first embodiment, which will not be described in detail in this embodiment. It should be understood that the device of this embodiment has at least all the beneficial effects of the first embodiment.

[0162] Embodiment 4

[0163] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by at least one processor, the method of the first embodiment is implemented.

[0164] The computer-readable storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0165] The processor can be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller unit (MCU), a microprocessor or other electronic components. In the several embodiments provided in the embodiments of the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are merely illustrative.

[0166] Embodiment 5

[0167] This embodiment provides an electronic device, including a memory and at least one processor. The memory stores a computer program, and when the computer program is executed by the at least one processor, the method of the first embodiment is implemented.

[0168] The processor can be an application specific integrated circuit (ASIC), a digital signal processor (DSP), a digital signal processing device (DSPD), a programmable logic device (PLD), a field programmable gate array (FPGA), a controller, a microcontroller unit (MCU), a microprocessor or other electronic components. In the several embodiments provided in the embodiments of the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device and method embodiments described above are merely illustrative.

[0169] It should be noted that, in this article, the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. The terms "comprise", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a..." does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0170] Although the embodiments disclosed in the present invention are as above, the above contents are only embodiments adopted for facilitating the understanding of the present invention and are not intended to limit the present invention. Any technician in the technical field to which the present invention belongs can make any modifications and changes in the form and details of the implementation without departing from the spirit and scope disclosed in the present invention, but the patent protection scope of the present invention shall still be subject to the scope defined in the attached claims.

Claims

1. A three-dimensional seam network reconstruction method, characterized in that: include: For the microseismic point cloud of each fracturing section of the target well, a microseismic event point is randomly selected, and the first neighborhood point set of the randomly selected microseismic event point is intercepted; Identify the orientation information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the orientation information, a preset fit index, and a preset distance index; When the current crack meets the requirements, the preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack; Clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein, obtaining a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture, wherein the noise points are microseismic event points in the second neighborhood point set whose density values ​​in the area are less than a preset core point density threshold and which do not fall within the neighborhood radius of any core point; The fractures of each fracturing stage of the target well are combined based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

2. The three-dimensional seam network reconstruction method according to claim 1, characterized in that: The first neighborhood point set of the randomly extracted microseismic event point is intercepted, comprising: Calculate the distance between the remaining microseismic event points in the microseismic event point cloud of the current fracturing section and the randomly selected microseismic event point; Based on the distance, a first neighborhood point set of the randomly selected microseismic event point is intercepted.

3. The three-dimensional seam network reconstruction method according to claim 1, characterized in that: The identifying the orientation information of the crack based on the first neighborhood point set, and determining whether the current crack meets the requirements according to the orientation information, a preset fit index and a preset distance index, includes: Fitting the intercept formula of the crack plane based on the first neighborhood point set to obtain the orientation information of the crack; According to the azimuth information, the number of microseismic event points in the first neighborhood point set that meet a preset distance index is counted to obtain a fit degree of the first neighborhood point set; If the fit meets the preset fit index, it is determined that the current crack meets the requirement.

4. The three-dimensional seam network reconstruction method according to claim 3, characterized in that: The orientation information includes the direction, inclination and dip of the fracture, and the calculation formula is as follows: Among them, λ1 represents the direction of the crack, λ2 represents the inclination of the crack, λ3 represents the tendency of the crack, and a, b, and c represent the intercepts of the crack on the x, y, and z axes, respectively.

5. The three-dimensional seam network reconstruction method according to claim 1, characterized in that: When the current crack meets the requirements, the preliminary geometric shape of the current crack is identified to obtain a second neighborhood point set corresponding to the preliminary geometric shape boundary of the current crack, including: When the current fracture meets the requirements, extract the microseismic event points in the first neighborhood point set that meet the preset fit index to obtain the fracture neighborhood point set; Projecting the crack neighborhood point set onto the plane where the current crack is located to obtain a projection point set of the current crack; A disk with a set radius is used to roll at the microseismic event points at the boundary of the projection point set, and two adjacent microseismic event points at the boundary in contact with the disk are sequentially connected as a boundary segment to obtain the preliminary geometric shape boundary of the current crack and the corresponding second neighborhood point set.

6. The three-dimensional seam network reconstruction method according to claim 1, characterized in that: The step of clustering the microseismic event points in the second neighborhood point set, identifying and removing noise points therein to obtain a third neighborhood point set, and connecting the microseismic event points in the third neighborhood point set to obtain a final geometric shape of the current fracture includes: According to the set neighborhood radius and core point density threshold, the microseismic event points in the second neighborhood point set are clustered to obtain a valid point set and a noise point set. The valid point set includes a core point set and a boundary point set. The core point is a microseismic event point in the second neighborhood point set whose regional density value is not less than the preset core point density threshold, and the boundary point is a microseismic event point whose regional density value is less than the neighborhood radius and falls within the neighborhood radius of other core points.

7. The three-dimensional seam network reconstruction method according to claim 5, characterized in that: After obtaining the final geometry of the current crack, it also includes: Removing the fracture neighborhood point set from the microseismic point cloud of the current fracturing section to obtain a remaining point set; The remaining point set is used as a new first neighborhood point set, and the orientation information of the crack is identified based on the first neighborhood point set. According to the orientation information, a preset fit index and a preset distance index, it is determined whether the current crack meets the requirements, and the orientation information of the next crack is identified to obtain the final geometric shape of the next crack until the number of microseismic event points in the remaining point set is less than the set value.

8. A three-dimensional seam network reconstruction device, characterized in that: include: A random extraction module is used to randomly extract a microseismic event point for each fracturing section of the target well, and intercept the first neighborhood point set of the randomly extracted microseismic event point; an orientation recognition module, configured to recognize the orientation information of the crack based on the first neighborhood point set, and determine whether the current crack meets the requirements according to the orientation information, a preset fit index and a preset distance index; A shape recognition module is used to perform preliminary geometric shape recognition of the current crack when the current crack meets the requirements, and obtain a second neighborhood point set corresponding to the boundary of the preliminary geometric shape of the current crack; A denoising processing module is used to cluster the microseismic event points in the second neighborhood point set, identify and remove the noise points therein, obtain a third neighborhood point set, and connect the microseismic event points in the third neighborhood point set to obtain the final geometric shape of the current fracture, wherein the noise points are microseismic event points in the second neighborhood point set whose density value in the area is less than a preset core point density threshold and do not fall within the neighborhood radius of any core point; A fracture network reconstruction module is used to combine the fractures of each fracturing section of the target well based on the orientation information and the final geometric shape to form a reconstructed three-dimensional fracture network of the target well.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by at least one processor, the method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: The method comprises a memory and at least one processor, wherein a computer program is stored in the memory, and when the computer program is executed by the at least one processor, the method according to any one of claims 1 to 7 is implemented.

Images