Perforation parameter optimization method and device for aftereffect body perforation technology

Abstract

The invention discloses a perforation parameter optimization method and device for an aftereffect body perforation technology. The method comprises the steps that an aftereffect body perforation penetration finite element model is established; determining a reservoir physical property parameter combination by adopting an orthogonal table; taking the reservoir physical property parameter combination as an input parameter, and outputting based on a pre-trained aftereffect body perforation process parameter prediction model to obtain a perforation process parameter range; and optimizing perforation parameters from the perforation process parameter range based on a global sensitivity quantitative analysis method.

Description

Technical Field

[0001] This application relates to the technical field of perforating completion operations for oil and gas wells, and particularly to a method and device for optimizing perforation parameters of after-effect body perforation technology. Background Art

[0002] Perforation is the main method in well completion. In oil and gas wells with perforating completion, factors such as the geology, fluids, and well types of the oil-bearing formation, the selection of perforation technology, the selection of perforating charges, the perforation damage mechanism, and the optimization design of perforation parameters (perforation depth, perforation density, hole diameter, phase) during the perforation process will directly affect the productivity of the oil and gas well. Summary of the Invention

[0003] This application provides a method and device for optimizing perforation parameters of after-effect body perforation technology. The method can simply obtain the process parameter range through the after-effect body perforation process parameter prediction model, and determine the optimal perforation parameter combination by combining the global sensitivity quantitative analysis method and productivity.

[0004] In a first aspect, this application provides a method for optimizing perforation parameters of after-effect body perforation technology, including: establishing an after-effect body perforation penetration finite element model; determining a combination of reservoir physical property parameters using an orthogonal table; using the combination of reservoir physical property parameters as input parameters, and obtaining the perforation process parameter range based on the output of a pre-trained after-effect body perforation process parameter prediction model; and optimizing the perforation parameters from the process parameter range based on the global sensitivity quantitative analysis method.

[0005] In a second aspect, an embodiment of the present invention further provides a device for optimizing perforation parameters of after-effect body perforation technology. The device includes: a memory and a processor; the memory is used to store a program for optimizing perforation parameters of after-effect body perforation technology, and the processor is used to read and execute the program for optimizing perforation parameters of after-effect body perforation technology, and execute the method described in any one of the above embodiments.

[0006] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a data processing program is stored, and the data processing program is executed by a processor to perform the method for optimizing perforation parameters of after-effect body perforation technology described in any one of the above embodiments.

[0007] Compared with related technologies, the present application provides a method and device for optimizing perforation parameters of after-effect body perforation technology, including: establishing a finite element model of an after-effect body perforating charge; determining reservoir physical property parameter combinations using an orthogonal table; using the reservoir physical property parameter combinations as input parameters, and obtaining a process parameter range based on a pre-trained prediction model of after-effect body perforation process parameters; and optimizing perforation parameters from the process parameter range based on a global sensitivity quantitative analysis method. The present application can accurately obtain the process parameter range through the prediction model of after-effect body perforation process parameters, and determine the final perforation parameter combination by combining the global sensitivity quantitative analysis method with productivity.

[0008] Other features and advantages of the present application will be described in the subsequent specification, and, in part, will be obvious from the specification, or will be understood by implementing the present application. Other advantages of the present application can be realized and obtained through the solutions described in the specification and the drawings. Description of the Drawings

[0009] The drawings are used to provide an understanding of the technical solutions of the present application, and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the technical solutions of the present application, and do not constitute a limitation to the technical solutions of the present application.

[0010] Figure 1 It is a flowchart of a method for optimizing perforation parameters of after-effect body perforation technology according to an embodiment of the present application;

[0011] Figure 2 It is a schematic diagram of a device for optimizing perforation parameters of after-effect body perforation technology according to an embodiment of the present application;

[0012] Figure 3 It is a schematic diagram of an after-effect body perforation penetration finite element model in some exemplary embodiments;

[0013] Figure 4 It is a schematic diagram of the penetration jet simulation of a perforating charge on a formation in some exemplary embodiments;

[0014] Figure 5 It is a flowchart of a method for optimizing perforation parameters based on global sensitivity quantitative analysis in some exemplary embodiments. Detailed Embodiments

[0015] This application describes multiple embodiments, but the description is exemplary rather than restrictive, and it will be obvious to those of ordinary skill in the art that there can be more embodiments and implementation solutions within the scope covered by the embodiments described in this application. Although many possible combinations of features are shown in the drawings and discussed in the detailed description, many other combinations of the disclosed features are also possible. Unless specifically restricted, any feature or element of any embodiment can be used in combination with any other feature or element in any other embodiment, or can replace any other feature or element in any other embodiment.

[0016] This application includes and contemplates combinations with features and elements known to those of ordinary skill in the art. The embodiments, features, and elements disclosed in this application can also be combined with any conventional features or elements to form unique inventive solutions defined by the claims. Any feature or element of any embodiment can also be combined with features or elements from other inventive solutions to form another unique inventive solution defined by the claims. Therefore, it should be understood that any feature shown and / or discussed in this application can be implemented alone or in any suitable combination. Therefore, the embodiments are not subject to other limitations except those made in accordance with the appended claims and their equivalents. In addition, various modifications and changes can be made within the scope of the appended claims.

[0017] In addition, when describing representative embodiments, the specification may have presented the method and / or process as a specific sequence of steps. However, to the extent that the method or process does not depend on the specific order of the steps described herein, the method or process should not be limited to the specific order of steps described. As will be understood by those of ordinary skill in the art, other step sequences are possible. Therefore, the specific order of steps set forth in the specification should not be construed as a limitation on the claims. In addition, the claims directed to the method and / or process should not be limited to performing their steps in the order written, and those skilled in the art can easily understand that these orders can vary and still remain within the spirit and scope of the embodiments of this application.

[0018] The basic principle of traditional perforation methods is to perforate oil and gas wells by squeezing the rock to create an oil and gas flow channel. Problems existing in traditional perforation completion:

[0019] (1) The perforation compaction zone contaminates the formation: When the perforating charge generates a high-speed jet acting on the formation, fragmented rock debris and residues of the perforating charge magma debris are extremely likely to block the formation, forming a perforation compaction zone. According to statistics, the squeezing effect of the jet will cause serious compaction pollution in the pore channel. The average thickness of this pollution layer is 1.2 - 1.3 cm, the porosity decreases by 13 - 22%, and the permeability decreases by 72 - 78%, seriously affecting the productivity of oil and gas wells and subsequent oil and gas production.

[0020] (2) Serious energy loss in the explosion of conventional perforating charges: After the perforating charge is detonated, the high-speed metal jet generated first penetrates the perforating gun, casing, and cement sheath, and only the remaining energy can participate in the process of creating fractures in the formation to form an oil and gas seepage channel. As a result, the energy effectively utilized for opening the formation pores and extending the fractures is greatly reduced.

[0021] Compared with the traditional perforating method, the after-effect body perforating technology uses the theory of deflagration-to-detonation transition to install an after-effect body made of chemical material particles at the mouth of the conventional perforating charge. After the perforating charge is detonated, the after-effect body is carried into the pore channel and instantly excited. With the help of the gravitational force of the eddy current field generated by the explosion, the chemical material particles are directionally aggregated and dragged into the perforating pore channel in a cloud state. After the chemical particles are excited, a large amount of chemical energy and heat energy are generated due to the principle of dust explosion, directly acting on the formation in the pore. The local hot deflagration rapidly extends to the detonation explosion of the entire pore channel, and secondary diversion fractures are formed. On the premise of ensuring the penetration performance, the pore diameter can be enlarged, the seepage area of the pore channel can be increased, the compaction pollution around the pore channel can be removed, and the purpose of increasing oil production and injection in oil and water wells can be achieved.

[0022] In the existing stage of the after-effect body perforating technology, most of the experimental research and theoretical analysis are carried out based on numerical penetration simulation. Although numerical simulation can realize the influence of various factors such as hole depth, hole density, and hole diameter, it brings a certain degree of accuracy and authenticity to the experiment. However, its calculation process is very cumbersome and time-consuming, and the requirements for computer resources are also limited in the engineering application range.

[0023] In view of the problems existing in the numerical simulation of the after-effect body perforating technology, the inventor proposed a new method for optimizing the perforating parameters of the after-effect body perforating process. By establishing a finite element model of after-effect body perforation penetration, using the orthogonal method to set multiple groups of different parameter combinations, using the jet penetration model for simulation, and establishing a prediction model of hole depth and hole diameter based on the CNN neural network. On the range of hole depth and hole diameter obtained by the neural network model, using the concept of model variance, based on the fact that the variance of the output caused by the individual or coupled action of the input perforating parameters (hole depth, hole diameter, hole density, phase angle) is different, the sensitivity of the input parameters (perforating parameters) to the output quantity (productivity ratio) is obtained, and the influence degree of multiple perforating parameters on the productivity ratio is quantitatively evaluated and analyzed, so as to optimize the parameter combination based on the influence on the productivity ratio.

[0024] An embodiment of the present invention provides a method for optimizing the perforating parameters of the after-effect body perforating process, as Figure 1 shown, the method includes steps S100 - S130:

[0025] S100: Establish a finite element model of after-effect body perforation penetration;

[0026] S110: Use the orthogonal table to establish a combination of reservoir physical property parameters,

[0027] S120: Use the reservoir physical property parameter combination as the input parameter, and obtain the range of perforation process parameters based on the pre-trained aftereffect body perforation process parameter prediction model;

[0028] S130: Determine the final perforation parameters from the range of perforation process parameters based on the global sensitivity quantitative analysis method.

[0029] In this embodiment, the perforation parameters include hole depth, hole diameter, hole density, and phase angle.

[0030] In an exemplary embodiment, an aftereffect body perforation penetration finite element model is established. The specific implementation process includes:

[0031] S1: The aftereffect body perforation penetration finite element model includes: a perforating charge model, a casing model, a cement sheath model, and a formation model; As Figure 3 shown, the aftereffect body perforation penetration finite element model is composed of a perforating charge shell, a liner, a shaped charge, an air domain, a gun slice, a water layer, a casing slice, a cement sheath, and a rock target. When building the formation model, the formation target model should be established according to the actual well condition formation model, such as: cement target, sandstone target, etc. To ensure that the simulated hole depth and hole diameter are closer to the actual downhole penetration depth and hole diameter.

[0032] S2: Mesh generation and boundary condition setting.

[0033] In order to obtain good simulation results while reducing the calculation time, it is necessary to control the element size and total number. Tetrahedral meshes are mainly used in mesh generation, and the mesh size is 0.5 mm. The boundary conditions are set according to the characteristics of the finite element model, including symmetric boundaries and non-reflective boundary conditions. The symmetric boundary condition is that when the model has symmetry, in order to reduce the model and the calculation amount, a 1 / 2 symmetric model can be used for perforation penetration simulation.

[0034] In an exemplary embodiment, the orthogonal experimental design method is a scientific method for researching and processing multi-factor experiments. For single-factor or two-factor experiments, since the number of factors is small, the design, implementation, and analysis of the experiments are relatively simple. However, in actual work, it is often necessary to simultaneously examine three or more experimental factors. Multi-factor experimental theories include: factor rotation method, random experimental method, comprehensive comparison method, uniform design method, full-factor experimental method, and orthogonal experimental design method. Conducting multi-factor analysis is a very complicated task. If the computational workload is not effectively and reasonably arranged, not only will the number of calculations increase, but the calculated data often cannot provide sufficient information, so that the purpose of calculation cannot be achieved in multi-factor problems. Therefore, after the calculation model is established, it is required to establish a quantitative relationship between factors with higher accuracy with fewer calculation times. This requires selecting calculation points according to the purpose of calculation and data analysis, so that the maximum information can be obtained from the least amount of calculation data. This is the problem to be studied in orthogonal combination design.

[0035] In this embodiment, the implementation process of combining reservoir physical property parameters using an orthogonal table is as follows:

[0036] S3: Use an orthogonal table to combine reservoir physical property parameters.

[0037] The orthogonal table is generally expressed as:

[0038] L n (m k )

[0039] Where L represents the orthogonal table; k = 5 represents the number of index factors (Factor) or the number of columns of the orthogonal table; m = 4 represents the number of different levels (Level) or the number of treatment methods under each factor; n represents the number of experiments or the number of rows of the orthogonal table, and n = k×(m - 1)+1.

[0040] In general multi-factor experiments, they are all fully implemented. For example, when conducting a multi-factor experiment with 5 factors and 4 levels, if fully implemented, it requires 4 5 =1024 combinations. However, if an L 16 (4 5 ) orthogonal table is used to arrange the experiment, 16 treatment combinations are enough. Here, L represents the orthogonal table, "16" represents the total number of experiments is 16 times, "4" represents that each factor has 4 levels, and "5" represents that 5 factors are considered.

[0041] S4: Select appropriate reservoir physical property parameters, experimental factors, and levels according to the well conditions, and obtain multiple groups of reservoir parameter combinations based on the orthogonal table.

[0042] In an exemplary embodiment, penetration simulations are performed under different combinations of reservoir parameters. Here, a jet penetration model is used for simulation based on the combination of reservoir physical property parameters obtained using the orthogonal array, and the hole depth and hole diameter under each combination of reservoir physical property parameters are obtained.

[0043] S5: Based on the combination of reservoir physical property parameters obtained from the orthogonal array, use the jet penetration model to obtain the hole depth and hole diameter under each combination of reservoir physical property parameters.

[0044] As Figure 4 shown, use the jet penetration model for simulation and record the simulation results, so as to obtain the hole depth and hole diameter under different combinations of reservoir physical property parameters in the afterbody perforation process.

[0045] S6: Outlier processing

[0046] Outliers refer to data that are unreasonable. In some modeling scenarios, ignoring outliers will lead to serious deviations in the conclusions drawn. The value range can be determined, and the values outside the value range are treated as outliers and deleted.

[0047] In an exemplary embodiment, the process of using a convolutional neural network to construct a prediction model for afterbody perforation process parameters is as follows:

[0048] S7: Determine the sample data set for constructing the prediction model of afterbody perforation process parameters using a convolutional neural network

[0049] Use the combination of reservoir parameters and the simulated process parameters as sample data to train the prediction model of afterbody perforation process parameters.

[0050] Obtain a characteristic data set of perforation parameters based on the determined influencing factors of hole depth (hole diameter), and obtain a label data set of the hole depth (hole diameter) corresponding to the combination of reservoir physical property parameters; use the characteristic data set and the label data set to establish a training data set for the model.

[0051] Obtain a characteristic data set of reservoir physical property parameters X based on the determined influencing factors of hole depth and hole diameter,

[0052] where the reservoir physical property parameter X is expressed as:

[0053] X = {X 孔隙度 , X 抗压强度 , X 负压 , X 剪切模量 ,...}

[0054] Obtain a label data set of the hole depth (hole diameter) corresponding to the obtained reservoir physical property parameters.

[0055] S8: Divide the data set for training the model

[0056] For example: Obtain 1000 sets of X and Y data; divide the feature dataset and its corresponding label dataset into two parts: 70% of the data is used for model training, and 30% of the data is used for verifying the training effect of the model; it can also be divided according to specific circumstances.

[0057] S9: Establish a prediction model for after-effect perforation process parameters based on CNN (Convolutional Neural Network);

[0058] Among them, the CNN convolutional neural network is as follows:

[0059] (1) One input layer, where the input reservoir physical property factor data (i.e., X) is converted into data in the format of n×1×9, n represents the number of samples in the training set and the validation set, and 9 is the feature dimension;

[0060] (2) One output layer, where the output layer has only one neuron, representing the perforation depth (aperture) output;

[0061] (3) Three convolutional layers and three pooling layers, where the convolutional layers and pooling layers are alternately arranged. The convolutional kernel of each convolutional layer is 3*3 in size, and the pooling uses the maximum pooling method; and two fully connected layers are finally set, and the number of neurons in each layer is set.

[0062] S10: If the prediction model for after-effect perforation process parameters meets the predetermined conditions, then determine the current prediction model for after-effect perforation process parameters as the trained prediction model for after-effect perforation process parameters.

[0063] Determine whether it meets the predetermined conditions by judging the accuracy of the model, that is, the mean absolute percentage error (MAPE); among them, the mean absolute percentage error (MAPE):

[0064]

[0065] In the formula: y i , y i , y i are the measured values, y' i , y' i , y' n are the predicted values, y' i is the measured average value, and n is the number of samples in the training model. The measured value here is the simulated value of the perforation depth or aperture obtained by simulating through the jet penetration model, and the predicted value is the predicted value of the perforation depth or aperture predicted by the prediction model for after-effect perforation process parameters.

[0066] The Mean Absolute Percentage Error (MAPE) can intuitively reflect the error. The closer the value is to 0, the more accurate the model. A threshold can be set. When it is less than the threshold and it is determined that the model meets the predetermined conditions, the current aftereffect perforation process parameter prediction model is determined to be a trained aftereffect perforation process parameter prediction model.

[0067] S11: Using the trained aftereffect perforation process parameter prediction model, with the reservoir physical property parameter combination as the input parameter, predict the process parameter range; in this step, the parameter ranges of the hole depth and hole diameter obtained by prediction using the aftereffect perforation process parameter prediction model are obtained.

[0068] In an exemplary embodiment, based on the EFAST global sensitivity quantitative analysis method, the perforation parameters are optimized in combination with the obtained process parameter range, and the implementation process is as Figure 5 shown.

[0069] Under the premise of ensuring the penetration performance, the aftereffect perforation technology can expand the hole diameter, increase the seepage area of the hole channel, relieve the compaction pollution around the hole channel, and achieve the purpose of increasing production and injection of oil and water wells. The influence of the aftereffect perforation parameters on productivity mainly considers four aspects: hole depth, hole diameter, hole density, and phase angle. Based on the hole depth and hole diameter ranges obtained from penetration simulation and neural network training, and a certain range of hole density and phase angle is set. In this embodiment, the ranges of hole density and phase angle are generally set according to the actual working conditions.

[0070] S12: Establish a productivity model

[0071] The productivity model is: Y = f(X);

[0072] The input perforation parameters are X(x1, x2,...x n ); where x1, x2,...x n are respectively the perforation parameters related to productivity, including: hole depth, hole diameter, hole density, and phase angle, etc. Based on the value ranges of the perforation parameters, a multi-dimensional parameter value space is formed.

[0073] S13: Establish a conversion function R;

[0074] The conversion function R converts the model Y = f(X), that is, Y = f(x1, x2,...x n ) into Y = f(s). Here, the conversion function R i is related to the probability distribution function of the perforation parameter x i :

[0075] x i (s) = R i (sin ω i s), i = 1, 2,..., n (2)

[0076] where: s is a scalar, and s ∈ (-∞, +∞), {ω i} is the frequency point related to the perforation parameter, and i = 1, 2, …, n.

[0077] S14: Perform Fourier transform on f(s):

[0078]

[0079] where:

[0080]

[0081] where: s is the mapping of the oscillation frequency of the perforation parameter X, which is a variable, and the value range of s can be [0, 2π]; sample s at equal intervals within the interval [0, 2Π], convert the sampled values into the values of each parameter through the conversion function, and input them into the model.

[0082] S15: Determine the frequency spectrum curve of the Fourier series;

[0083] Then the frequency spectrum curve of the Fourier series is defined as:

[0084]

[0085] where: p is the Fourier transform parameter, p ∈ Z = {-∞, …, -1, 0, 1, +∞}, A m and B n are Fourier coefficients.

[0086] S16: The variance V caused by each perforation parameter in the production capacity model result i can be determined by the change range of the input perforation parameter, and the variance V i is expressed as:

[0087]

[0088] where: g ∈ (1, +∞), which is the number of perforation parameters selected according to the actual working conditions, and then the range of i is determined, and g is an integer.

[0089] S17: The total variance V of the model

[0090] is obtained from the Fourier coefficients and the frequency ω corresponding to the perforation parameter x i i i i i and the total variance V of the model output can be obtained through equations (5) and (6).

[0091] The total variance of the production capacity model is:

[0092] V = ∑V i + ∑Vij +∑V ijk +∑V ijkm …+∑V 12…n

[0093] =(V1 + V2 + V3 + V4)+(V 12 +V 13 +V 14 +V 23 +V 24 +V 34 )+(V 123 +V 124 +V 234 )+…+V 12…n

[0094] In the formula: V 12…n is the variance under the action of n parameters;

[0095] V ij (V 12 , V 13 , …, i < j) is the variance under the interaction of parameter x i and parameter x j ;

[0096] V ij = V[E(Y|x i , x j |)] - V i - V j , i ≠ j

[0097] V ijk (V 123 + V 124 + V 234 , i < j < k) is the variance under the joint action of parameters x i , x j , x k ;

[0098] V ijk = V[E(Y|x i , x j , x k |)] - V i - V j - V k , i ≠ j ≠ k

[0099] ∑V ijmk (V 1234 , i < j < k < m) is the result of the joint action of parameters x i , x j , x k , x m ;

[0100] V ijkg = V[E(Y|x i ,x j ,x k ,x m |)] - V i -V j -V k -V m ,i ≠ j ≠ k ≠ m.

[0101] S18: Determine the total sensitivity index of the perforation parameters.

[0102] The first-order sensitivity index S i of the perforation parameter x i can be defined as follows:

[0103]

[0104] This sensitivity index reflects the proportion of the perforation parameter in the total output variance of the productivity model. Similarly, the second-order and third-order sensitivity indices S i of the parameter x ij 、S ijm can be defined as:

[0105]

[0106] The total sensitivity index of the perforation parameter is equal to the sum of the sensitivity indices of each order obtained from the productivity model. It can be expressed as:

[0107] S Ti = S i + S ij + S ijm + … + S 12…i…k (13)

[0108] The total sensitivity index intuitively reflects the sum of the contribution rate of the perforation parameter and the contribution of the interaction with other parameters indirectly to the total variance of the productivity model output.

[0109] Through the EFAST global sensitivity quantitative analysis method, select the perforation parameter combination with the largest total sensitivity index as the optimal aftereffect body perforation parameter combination. Study the results obtained from LS-DYNA numerical simulation and neural network training, and obtain the primary and secondary relationship of the influence of each perforation parameter on productivity and the influence trend of each perforation parameter on the productivity ratio with the change of the parameter value range through variance, providing a reliable basis for the optimization of the aftereffect body perforation parameters.

[0110] In the second aspect, the embodiment of the present invention also provides a perforation parameter optimization device for the aftereffect body perforation technology, such as Figure 2As shown in the figure, the device includes: a memory 200 and a processor 210; the memory is used to store a program for optimizing perforation parameters of the aftereffect body perforation technology, and the processor is used to read and execute the program for optimizing perforation parameters of the aftereffect body perforation technology, and execute the method described in any one of the above embodiments.

[0111] In a third aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a data processing program is stored, and the data processing program is executed by a processor to perform the method for optimizing perforation parameters of the aftereffect body perforation technology described in any one of the above embodiments.

[0112] The embodiments of the present application have the following technical effects:

[0113] (1) The orthogonal method is used to combine reservoir physical property parameters, and the finite element numerical simulation method is used to simulate the penetration of the aftereffect body perforation technology under different reservoir parameter combinations;

[0114] (2) Using a convolutional neural network to train and construct a prediction model for aftereffect body perforation process parameters can save calculation time;

[0115] (3) The EFAST global sensitivity analysis method is used to analyze and calculate the result variance of the productivity model. Clearly obtain the contribution weights of each perforation parameter to the total variance, quantitatively analyze the influence of perforation parameters on productivity, and complete the optimization of perforation parameter combinations.

[0116] Example 1

[0117] For the actual working condition parameters of a certain oilfield group in the Bohai Sea, the thickness of the casing slice is set to 10.36 mm, and the thickness of the cement sheath is 20.64 mm. A dynamic perforation numerical simulation scheme for the aftereffect body perforation technology is formulated, and the DP46HMX45 perforating charge is selected, with 5 reservoir performance parameters and 5 variable values for each parameter. The optimization of aftereffect body perforation parameters for this oilfield is completed.

[0118] The first step: Establish a finite element model for aftereffect body perforation penetration

[0119] Step 1. Based on the actual working condition parameters of a certain oilfield group in the Bohai Sea, establish a finite element model for aftereffect body perforation penetration. The finite element model for aftereffect body perforation penetration consists of a perforating cartridge case, a liner, a shaped charge, an air domain, a gun slice, a water layer, a casing slice, a cement sheath, and a rock target. When building the formation model, the formation target model should be established according to the formation model of the actual well condition, such as: cement target, sandstone target, etc., to ensure that the simulated hole depth and hole diameter are closer to the actual downhole penetration depth and hole diameter.

[0120] Step 2: Mesh generation and boundary condition setting. To obtain good simulation results while reducing the calculation time, it is necessary to control the element size and total number. There are 1,538,218 nodes and 1,451,279 elements in the finite element model of the aftereffect perforating charge system. Tetrahedral meshes are mainly used in the mesh generation, and the mesh size is 0.5 mm. According to the characteristics of the finite element model, symmetric boundaries and non-reflective boundary conditions are set. When the model has symmetry, a 1 / 2 symmetric model can be used for perforation penetration simulation to reduce the model and calculation amount.

[0121] Step 2: Use the orthogonal experiment method to combine reservoir physical property parameters

[0122] Through literature research, the actual data reference range of a certain oilfield group in the Bohai Sea, and the above numerical simulation model, among the controllable parameters, 5 influencing parameters are selected for analysis, including: porosity, shear modulus, compressive strength, confining pressure, and negative pressure value. The value ranges of the 5 parameters for the orthogonal experiment are as follows:

[0123] Porosity (%) : 5, 10, 15, 20, 25;

[0124] Confining pressure (MPa) : 0, 10, 20, 30, 40;

[0125] Negative pressure (MPa) : 0, 2, 5, 7, 10;

[0126] Compressive strength (MPa) : 5, 10, 20, 30, 60;

[0127] Shear modulus (GPa) : 0.5, 1.0, 1.5, 3.0, 6.0.

[0128] According to the parameter variables, test factors and number of levels, select and make an L 25 (5 5 ) orthogonal experimental design table (as shown in Table 1).

[0129] Table 1

[0130]

[0131]

[0132] Step 3: Conduct penetration simulations under different reservoir parameter combinations:

[0133] (1) Based on the reservoir physical property parameter combinations in the above orthogonal table, use the finite element numerical simulation method to conduct penetration simulations to obtain the hole depth and hole diameter under different reservoir physical property parameter combinations. The obtained hole depth range is 451.110 mm - 603.777 mm, and the hole diameter range is 12.586 mm - 13.629 mm.

[0134] (2) Outlier handling: Delete outliers

[0135] Step 4: Use a convolutional neural network to construct a prediction model for after-effect perforation process parameters:

[0136] (1) Obtain a feature dataset of perforation parameters based on the influencing factors of the determined hole depth (hole diameter), and obtain a label dataset of the hole depth (hole diameter) corresponding to the reservoir physical property parameter combination; use the feature dataset and the label dataset to establish a hole depth (hole diameter) prediction model based on a convolutional neural network; obtain a feature dataset of reservoir physical property parameters X based on the influencing factors of the determined hole depth and hole diameter, where the reservoir physical property parameter X is expressed as:

[0137] X = {X 孔隙度 , X 抗压强度 , X 负压 , X 剪切模量 ,...}

[0138] (2) The hole depth (hole diameter) Y corresponding to the obtained reservoir physical property parameters forms a label dataset, and 1000 groups of X and Y data are obtained. The feature dataset and its corresponding label dataset are divided into three parts: 70% of the data is used to train the model, and 30% of the data is used to verify the training effect of the model.

[0139] (3) The established CNN (convolutional neural network) includes:

[0140] A: 1 input layer, where the input reservoir physical property factor data (i.e., X) is converted into data in the format of n×1×9, n represents the number of samples in the training set and the validation set, and 9 is the feature dimension;

[0141] B: 1 output layer, where the output layer has only 1 neuron, representing the output hole depth (hole diameter);

[0142] C: 3 convolutional layers and 3 pooling layers, where the convolutional layers and the pooling layers are alternately arranged. The convolutional kernel of each convolutional layer has a size of 3*3, and the pooling uses the maximum pooling method; and two fully connected layers are finally set, and the number of neurons in each layer is set.

[0143] (4) The evaluation of the hole depth and hole diameter prediction model can be represented by the mean absolute percentage error (MAPE). MAPE can intuitively reflect the error. The closer the value is to 0, the more accurate the model is.

[0144]

[0145] In the formula: y i , y i , y i are the measured values, y'i , y' i , y' n is the predicted value, and y' i is the measured average value, and n is the number of samples of data in the training model.

[0146] After calculation, the mean absolute percentage errors (MAPE) of the hole depth and hole diameter prediction models are 2.63% and 1.58% respectively, which is in line.

[0147] Step 5: A perforation parameter optimization method based on the global sensitivity quantitative analysis of EFAST.

[0148] (1) Under the premise of ensuring the penetration performance, the after-effect body perforation technology can expand the hole diameter, increase the seepage area of the hole channel, relieve the compaction pollution around the hole channel, and achieve the purpose of increasing production and injection of oil and water wells. The influence of after-effect body perforation parameters on productivity mainly considers four aspects: hole depth, hole diameter, hole density, and phase angle. Based on the hole depth and hole diameter ranges obtained from penetration simulation and neural network training, and a certain range of hole density and phase angle is set.

[0149] (2) The hole depth range is 451.110 mm - 603.777 mm, and the hole diameter range is 12.586 mm - 13.629 mm. The hole density is set to 12, 18, 24, 30 holes / m, and the phase angle is set to 60°, 90°, 120°, 180°.

[0150] (3) Given a productivity model according to the actual well conditions, and based on the value range of perforation parameters, a multi-dimensional parameter value space is formed. A conversion function R is provided to convert the productivity model into Y = f(s). And perform Fourier transform on f(s). And perform variance processing. Through variance, the primary and secondary relationships of the influence of each perforation parameter on productivity and the influence trend of each perforation parameter on the productivity ratio with the change of the parameter value range are obtained, providing a reliable basis for the optimization of perforation parameters.

[0151] In the above example, through the perforation parameter optimization method using the after-effect body perforation technology, through the following specific technical features, the following technical effects are achieved:

[0152] (1) Using the orthogonal experiment method, multiple groups of different parameter combinations are formed to simulate the penetration effects under different parameters (porosity, compressive strength, negative pressure, etc.), and trained through the CNN convolutional neural network to obtain the hole depth range and hole diameter range under different parameter combinations; solving the problem that most current penetration tests are only limited to experimental research and theoretical analysis based on numerical penetration simulation, and the calculation process will be very cumbersome and relatively time-consuming.

[0153] (2) On the basis of the formed hole depth and aperture range, use the global sensitivity analysis method to quantitatively evaluate the influence of perforation parameters and the coupling of perforation parameters on productivity, identify the factors sensitive to productivity during the aftereffect body perforation process, and at the same time consider the uncertainty of some main control factors to optimize the perforation parameters of the aftereffect body perforation process. Solve the problem that currently it is impossible to specifically analyze the influence magnitude of each factor on productivity. The global sensitivity analysis adopts the concept of model variance. Based on the fact that the variance of the output is different due to the individual or coupled effects of the input perforation parameters (hole depth, aperture, perforation density, phase angle), the sensitivity degree of the input parameters (perforation parameters) to the output quantity (productivity ratio) is obtained, and the influence degree of multiple perforation parameters on the productivity ratio is quantitatively evaluated and analyzed. Understand the influence of reservoir physical properties parameters (porosity, negative pressure, compressive strength, confining pressure shear modulus, etc.) on hole depth, aperture and productivity in sandstone formations, bringing a certain degree of accuracy and authenticity to the experiment.

[0154] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, and their appropriate combinations. In the hardware implementation, the division between the functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or one function or step may be executed by several physical components in cooperation. Some or all components may be implemented as software executed by a processor, such as a digital signal processor or a microprocessor, or may be implemented as hardware, or may be implemented as an integrated circuit, such as an application-specific integrated circuit. Such software can be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cassette, tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those of ordinary skill in the art that a communication medium typically contains computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium.

Claims

1. A method for optimizing perforation parameters of aftereffect body perforation technology, characterized in that, Including: Establish a finite element model for aftereffect perforation penetration; Use an orthogonal array to determine the combination of reservoir physical property parameters; Taking the combination of the reservoir physical property parameters as input parameters, obtain the range of perforation process parameters based on the pre-trained prediction model of aftereffect perforation process parameters; Determine the perforation parameters from the range of perforation process parameters based on the global sensitivity quantitative analysis method.

2. The method for optimizing perforation parameters of after-effect body perforation technology according to claim 1, characterized in that The training process of the prediction model of aftereffect perforation process parameters is as follows: For multiple combinations of reservoir parameters, use the jet penetration model to conduct penetration simulations respectively to obtain the simulated range of process parameters; Use the combination of the reservoir parameters and the simulated process parameters as sample data to train the prediction model of aftereffect perforation process parameters; If the prediction model of aftereffect perforation process parameters meets the predetermined conditions, determine the current prediction model of aftereffect perforation process parameters as the trained prediction model of aftereffect perforation process parameters.

3. The method for optimizing perforation parameters of after-effect body perforation technology according to claim 2, characterized in that, The predetermined conditions are: the mean absolute percentage error MAPE meets a predetermined value; Wherein, the mean absolute percentage error MAPE is: In the above formula, y i , y i , y i are measured values, y' i , y' i , y' n are predicted values, y' i is the measured average value, and n is the number of samples.

4. The method for optimizing perforation parameters of the aftereffect body perforation technology according to claim 3, wherein, The combination of reservoir parameters X in the sample data is: X = {X 孔隙度 , X 抗压强度 , X 负压 , X 剪切模量 ,...} The simulated perforation process parameters Y in the sample data are: hole depth data or hole diameter data.

5. The method for optimizing perforation parameters of the aftereffect perforation technology according to claim 3, characterized in that The prediction model of aftereffect perforation process parameters is a model based on the CNN convolutional neural network, and the model includes: 1 input layer, 1 output layer, 3 convolutional layers and 3 pooling layers.

6. The method for optimizing perforation parameters of the aftereffect perforation technology according to claim 1, characterized in that The orthogonal array is: L n (m k ) Wherein, L represents the orthogonal array, k represents the number of reservoir parameters, m represents the number of levels under each factor, and n is the number of experiments, n = k×(m - 1) + 1.

7. The perforation parameter optimization method for aftereffect body perforation technology according to claim 1, wherein The determination of the perforation parameters from the range of perforation process parameters based on the global sensitivity quantitative analysis method includes: Establish a productivity model and use a transformation function to determine the Fourier coefficients of the productivity model; For each perforation parameter, determine the total variance of the perforation parameter according to the Fourier coefficients and using the productivity model; Determine the sensitivity index corresponding to each perforation parameter according to the total variance of the perforation parameter; Determine the final perforation parameters according to the sensitivity index.

8. The method for optimizing perforation parameters of aftereffect body perforation technology according to claim 7, characterized in that The establishment of the productivity model and the use of the transformation function to determine the Fourier coefficients of the productivity model include: Establish a productivity model and use a transformation function to obtain the productivity model in the frequency domain; Perform Fourier transform according to the productivity model in the frequency domain to obtain the Fourier coefficients of the productivity model; The productivity model in the frequency domain is: The Fourier coefficients A m and B n :

9. The perforation parameter optimization method of the aftereffect body perforation technology according to claim 7, characterized in that The determination of the total variance of each perforation parameter according to the Fourier coefficients and using the productivity model includes: Determine the spectral curve of the Fourier series according to the Fourier coefficients; Determine the variance of the productivity model according to the spectral curve of the Fourier series and the numerical range of each perforation parameter; Determine the total variance of the productivity model according to the variance of each perforation parameter.

10. The perforation parameter optimization method for after-effect body perforation technology according to claim 7, characterized in that, The variance of the productivity model is: In the above formula, V i is the variance of the perforation parameter i, g is the number of perforation parameters, i is the perforation parameter, and p is the Fourier transform parameter.

11. A perforation parameter optimization device for after-effect body perforation technology, characterized in that, The device includes: a memory and a processor; the memory is used to store a program for optimizing perforation parameters of the aftereffect body perforation technology, and the processor is used to read and execute the program for optimizing perforation parameters of the aftereffect body perforation technology, and execute the method according to any one of claims 1-10.

12. A computer-readable storage medium, on which a data processing program is stored, and the data processing program is executed by a processor to perform the method for optimizing perforation parameters of the aftereffect body perforation technology according to any one of claims 1-10.

Images