Drilling parameter optimization method and device based on clustering analysis

Abstract

The invention provides a well drilling parameter optimization method and device based on clustering analysis, and relates to the technical field of well drilling construction data analys.The method comprises the steps that for a to-be-optimized target well, well logging data and well drilling data of adjacent wells of a target block where the to-be-optimized target well is located are obtained; according to the logging information of the target well to be optimized and the logging data of the adjacent well, dimension reduction processing is carried out, and a logging curve after dimension reduction processing is obtained; performing clustering analysis according to the logging curve, and determining similar wells of the same category as the target well to be optimized; according to the mechanical specific energy and the mechanical drilling speed of the same kind of wells, a drilling parameter range corresponding to each layer position is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range; and optimizing the drilling parameters of the target well to be optimized by using the drilling data of the optimal adjacent well to obtain optimized drilling parameters.

Description

Technical Field

[0001] The invention relates to the technical field of drilling construction data analysis, and in particular to a drilling parameter optimization method and device based on cluster analysis. Background Art

[0002] This section is intended to provide a background or context for embodiments of the present invention. No description herein is admitted to be prior art by virtue of its inclusion in this section.

[0003] During the drilling process, if the formation has poor drillability and the rock is highly abrasive, the drill bit will be worn and the downhole will be complicated frequently, which will seriously limit the overall drilling efficiency. Taking a certain structural block as an example, due to the development of natural fractures, strong rock abrasiveness, and faults encountered in multiple layers of shale gas wells in the block, there are many downhole complexities such as leakage, overflow, and drill jamming. The drilling process is frequently tripped and drilled, and a single well is tripped and drilled up to 26 times. The minimum footage of a single drill bit is only about 100m, which greatly limits the drilling efficiency. In response to this, drilling optimization schemes such as drilling tool combination optimization, drill bit optimization, and coring speed-up measures have been designed, and certain results have been achieved. However, the mechanical drilling speed of some sections is less than 5m / h, and further research is still needed. In the existing technology, most of the research content analyzes the relationship between mechanical specific energy and mechanical drilling speed by statistical drilling data, and does not combine the existing logging data for more refined analysis.

[0004] In summary, a technical solution is proposed to reasonably analyze the bottom layer to be drilled, optimize drilling parameters, and improve construction efficiency. Summary of the invention

[0005] In order to solve the problems existing in the prior art, the present invention proposes a drilling parameter optimization method and device based on cluster analysis. The present invention fully exploits the deep geological correlation between wells in the same layer, conducts more accurate drilling parameter optimization analysis, and performs multi-dimensional clustering of logging curves of vertical wells in the target block through unsupervised learning methods. On this basis, the mechanical drilling speed of each section of the same well is analyzed in combination with the mechanical specific energy theory, and the optimal drilling parameter range for each layer is further determined, and verified through typical wells, providing strong technical and data support for the drilling field.

[0006] In a first aspect of an embodiment of the present invention, a drilling parameter optimization method based on cluster analysis is proposed, comprising:

[0007] For the target well to be optimized, the logging data and drilling data of the adjacent wells in the target block are obtained;

[0008] Performing dimensionality reduction processing on the logging data of the target well to be optimized and the logging data of adjacent wells to obtain a logging curve after dimensionality reduction processing;

[0009] Performing cluster analysis based on the well logging curves to determine similar wells of the same category as the target well to be optimized;

[0010] According to the mechanical specific energy and mechanical drilling speed of similar wells, the drilling parameter range corresponding to each layer is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range;

[0011] The drilling data of the optimal adjacent well is used to optimize the drilling parameters of the target well to be optimized, so as to obtain optimized drilling parameters.

[0012] In a second aspect of an embodiment of the present invention, a drilling parameter optimization device based on cluster analysis is proposed, comprising:

[0013] A data acquisition module is used to acquire the logging data and drilling data of the adjacent wells in the target block where the target well to be optimized is located;

[0014] A dimension reduction processing module is used to perform dimension reduction processing on the well logging data of the target well to be optimized and the well logging data of the adjacent wells to obtain a well logging curve after dimension reduction processing;

[0015] A cluster analysis module, used for performing cluster analysis according to the well logging curves to determine similar wells of the same category as the target well to be optimized;

[0016] A data analysis module is used to screen out the drilling parameter range corresponding to each layer in the drilling data according to the mechanical specific energy and mechanical drilling speed of similar wells, and determine the optimal adjacent well corresponding to the target well to be optimized through the drilling parameter range;

[0017] The parameter optimization module is used to optimize the drilling parameters of the target well to be optimized by using the drilling data of the optimal adjacent well to obtain the optimized drilling parameters.

[0018] In a third aspect of an embodiment of the present invention, a computer device is proposed, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements a drilling parameter optimization method based on cluster analysis when executing the computer program.

[0019] In a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a drilling parameter optimization method based on cluster analysis is implemented.

[0020] In a fifth aspect of an embodiment of the present invention, a computer program product is proposed. The computer program product includes a computer program. When the computer program is executed by a processor, a drilling parameter optimization method based on cluster analysis is implemented.

[0021] The drilling parameter optimization method and device based on cluster analysis proposed in the present invention obtain the logging data and drilling data of the adjacent wells in the target block where the target well to be optimized is located; perform dimensionality reduction processing based on the logging data of the target well to be optimized and the logging data of the adjacent wells to obtain the logging curve after dimensionality reduction processing; perform cluster analysis based on the logging curve to determine the same type of wells as the target well to be optimized; based on the mechanical specific energy and mechanical drilling speed of the same type of wells, screen out the drilling parameter range corresponding to each layer in the drilling data, and determine the optimal adjacent well corresponding to the target well to be optimized through the drilling parameter range; use the drilling data of the optimal adjacent well to optimize the drilling parameters of the target well to be optimized to obtain the optimized drilling parameters. The overall solution fully exploits the deep geological correlation between wells at the same layer, conducts more accurate drilling parameter optimization analysis, and performs multi-dimensional clustering of logging curves of vertical wells in the target block through unsupervised learning methods. On this basis, the mechanical drilling rate of each section of similar wells is analyzed in combination with the mechanical specific energy theory, and the optimal drilling parameter range for each layer is further determined, which is verified by typical wells. The present invention combines the mechanical specific energy theory to carry out targeted optimization of the drilling parameters of each layer of the target well to be optimized, effectively improves the drilling efficiency based on the concept of geological engineering integration, and provides strong technical and data support for the drilling field. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 It is a schematic flow chart of a drilling parameter optimization method based on cluster analysis according to an embodiment of the present invention.

[0024] Figure 2 It is a schematic diagram of a drilling parameter optimization process according to a specific embodiment of the present invention.

[0025] Figure 3 It is a schematic diagram of the steps of multi-dimensional clustering of logging curves based on the PCA-AP algorithm according to a specific embodiment of the present invention.

[0026] Figure 4A It is a schematic diagram of a standardized logging curve according to a specific embodiment of the present invention.

[0027] Figure 4B It is a schematic diagram of a standardized comprehensive logging curve (after dimensionality reduction) according to a specific embodiment of the present invention.

[0028] Figure 5It is a schematic diagram of the comprehensive logging curve clustering result according to a specific embodiment of the present invention.

[0029] Fig. 6A It is a schematic diagram of the mechanical drilling rate results after smoothing of the Y11 well in a specific embodiment of the present invention.

[0030] Figure 6B It is a schematic diagram of the mechanical specific energy results of the Y11 well after smoothing according to a specific embodiment of the present invention.

[0031] Figure 7 This is a comparison chart of mechanical drilling speeds of similar wells in the Xujiahe Formation Y11 according to a specific embodiment of the present invention.

[0032] Figure 8 This is a comparison diagram of mechanical specific energy of similar wells of Xujiahe Formation Y11 according to a specific embodiment of the present invention.

[0033] Fig. 9 It is a diagram of mechanical drilling speed and mechanical specific energy prediction results of a specific embodiment of the present invention.

[0034] Fig. 10A It is a diagram showing the optimization effect of drilling parameters (mechanical drilling rate) of the Y11 well according to a specific embodiment of the present invention.

[0035] Fig. 10B It is a diagram showing the optimization effect of the drilling parameters (mechanical specific energy) of the Y11 well according to a specific embodiment of the present invention.

[0036] Fig.11 It is a schematic diagram of the architecture of a drilling parameter optimization device based on cluster analysis according to an embodiment of the present invention.

[0037] Fig.12 It is a schematic diagram of the structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The principles and spirit of the present invention will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are provided only to enable those skilled in the art to better understand and implement the present invention, and are not intended to limit the scope of the present invention in any way. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete, and to fully convey the scope of the present disclosure to those skilled in the art.

[0039] Those skilled in the art will appreciate that the embodiments of the present invention may be implemented as a system, device, apparatus, method or computer program product. Therefore, the present disclosure may be specifically implemented in the following forms, namely: complete hardware, complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0040] According to an embodiment of the present invention, a method and device for optimizing drilling parameters based on cluster analysis are proposed, which relate to the technical field of drilling construction data analysis.

[0041] The principle and spirit of the present invention are explained in detail below with reference to several representative embodiments of the present invention.

[0042] Figure 1 FIG. 1 is a flow chart of a drilling parameter optimization method based on cluster analysis according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0043] S101, for a target well to be optimized, obtaining well logging data and drilling data of adjacent wells in the target block;

[0044] S102, performing dimensionality reduction processing on the logging data of the target well to be optimized and the logging data of adjacent wells to obtain a logging curve after dimensionality reduction processing;

[0045] S103, performing cluster analysis based on the well logging curves to determine similar wells of the same category as the target well to be optimized;

[0046] S104, based on the mechanical specific energy and mechanical drilling speed of similar wells, the drilling parameter range corresponding to each layer is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range;

[0047] S105, optimizing the drilling parameters of the target well to be optimized by using the drilling data of the optimal adjacent well to obtain optimized drilling parameters.

[0048] This application uses principal component analysis combined with the nearest neighbor propagation method to perform multi-dimensional clustering of logging curves, fully explores the deep geological correlation between wells in the same layer, and combines the mechanical specific energy theory to perform targeted optimization of the drilling parameters of each layer of the target well to be optimized, effectively improving the drilling efficiency based on the concept of geological and engineering integration.

[0049] The present invention uses the PCA-AP algorithm to cluster the well logging curves in multiple dimensions, thereby further optimizing the drilling parameters.

[0050] Principal component analysis (PCA) is a feature extraction method that maps n-dimensional features into k-dimensional orthogonal features. The new k-dimensional features retain most of the original information and do not overlap each other. The process of further reducing the k-dimensional features to one dimension is to perform a weighted average of the coefficients in each principal component score model according to the variance contribution of the k principal components. Before clustering, the well logging data is first subjected to principal component analysis dimensionality reduction.

[0051] Affinity propagation (AP) is an objective unsupervised learning algorithm that considers each data sample as a potential cluster center, then establishes a similarity matrix between all samples, and then determines the final cluster center through attraction and belonging. Compared with traditional clustering algorithms, the AP algorithm does not require manual setting of cluster centers and the number of clusters, and has higher computational efficiency and better clustering results when processing multidimensional large-scale data sets. In summary, the PCA-AP algorithm combining principal component analysis with neighbor propagation is used to realize multi-dimensional clustering of well logging curves of vertical wells in the target block.

[0052] In order to explain the above-mentioned drilling parameter optimization method based on cluster analysis more clearly, each step is described in detail below.

[0053] In one embodiment, S101, for a target well to be optimized, obtaining well logging data and drilling data of adjacent wells in the target block includes:

[0054] For different adjacent wells, the logging data and drilling data are calibrated according to the same depth; wherein the logging data at least includes acoustic wave time difference, well diameter, compensated neutron, density, natural gamma, and resistivity; and the drilling data at least includes drilling speed, drilling pressure, and rotation speed.

[0055] In one embodiment, S102, performing dimensionality reduction processing on the well logging data of the target well to be optimized and the well logging data of the adjacent wells to obtain the well logging curve after dimensionality reduction processing includes:

[0056] For the well logging data, the n-dimensional features of the well logging data are mapped into k-dimensional orthogonal features; wherein n>k, n and k are positive integers, and the information in the k-dimensional orthogonal features does not overlap;

[0057] The k-dimensional features are subjected to multiple dimensionality reduction processing by feature extraction to obtain the corresponding logging curve reduced to one dimension; wherein, when the dimensionality reduction processing is performed by feature extraction, the coefficients in each principal component score model are weighted averaged according to the variance contribution of the k principal components to obtain the dimensionality reduction result.

[0058] In one embodiment, S103, performing cluster analysis based on the well logging curve to determine similar wells of the same category as the target well to be optimized includes:

[0059] S1031, constructing a scoring model according to the well logging curve, the principal component variance contribution rate and the principal component scoring model coefficient, the relationship is:

[0060]

[0061] In the formula, w i,jis the coefficient of the jth logging data in the scoring model of the i-th vertical well sample; W i l is the variance contribution rate of the lth principal component of the i-th vertical well sample; is the coefficient of the jth logging data in the lth principal component of the i-th vertical well sample; F i is the scoring model of the i-th vertical well sample, as the well logging curve of the i-th vertical well sample; k represents the feature dimension; N is the number of vertical wells; n L is the number of logging data; X i,j represents the jth logging data of the i-th vertical well sample;

[0062] S1032, initialize the attraction matrix r(i, k) and the attribution matrix a(i, k) to zero matrices, set the maximum number of iterations T, the number of consecutive invariant times of the cluster center C, and the damping coefficient λ, the current number of iterations t = 0, and initialize the similarity matrix:

[0063]

[0064] In the formula, s t (i,k) represents the similarity between the i-th vertical well sample and the k-th vertical well sample. t (k,k) takes the median of the kth row in the matrix; F i,j represents the jth element of the logging curve of the i-th vertical well sample; F k,j represents the jth element of the combined logging curve of the kth vertical well sample; m represents the number of samples;

[0065] S1033, in each iteration, let t increase by 1, and update the attraction matrix r(i, k) and the attribution matrix a(i, k):

[0066] r t (i,k)=s t (i,k)-max{a t (i,j)+s t (i,j)}, j∈{1,2,...,N|j≠k}; (4)

[0067]

[0068]

[0069] In the formula, r t (i,k), a t (i,k),s t (i, k) is the attraction matrix, belonging matrix, and similarity matrix updated after the tth iteration; a t (i,j),s t(i, j) is the degree of attribution and similarity between the i-th vertical well sample and the j-th vertical well sample in the degree of attribution matrix and similarity matrix; r t (k,k), a t (k,k) takes the median of the kth row in the attraction matrix and the attribution matrix; r(j,k) represents the attraction between the jth vertical well sample and the kth vertical well sample in the attraction matrix;

[0070] S1034, introduce the damping coefficient λ and calculate the decision matrix c(i,k):

[0071] r t (i,k)=λr t-1 (i,k)+(1-λ)r t (i,k); (7)

[0072] a t (i,k)=λa t-1 (i,k)+(1-λ)a t (i,k); (8)

[0073] c t (i,k)=r t (i,k)+a t (i,k); (9)

[0074] In the formula, c t (i, k) represents the decision matrix obtained after the tth iteration; a t-1 (i,k),r t-1 (i, k) represents the attraction matrix and belonging matrix updated after the t-1th iteration;

[0075] S1035, after multiple iterations (repeating S1033 to S1034), until the number of iterations t = T or the cluster center remains unchanged for C consecutive times;

[0076] According to the decision matrix c(i, k) obtained by iterative updating, the vertical well samples corresponding to the column index where the elements on the diagonal are greater than 0 are taken as cluster centers, and the vertical well samples corresponding to the row index where all the elements greater than 0 corresponding to the column are taken as the same category.

[0077] In one embodiment, S104, based on the mechanical specific energy and mechanical drilling rate of similar wells, the drilling parameter range corresponding to each layer is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range, including:

[0078] For the mechanical specific energy and mechanical drilling speed of drilling, the relationship is:

[0079]

[0080] In the formula, MSE total Indicates mechanical specific energy; WOB surf is the drilling pressure; D is the drill bit diameter; TQ surf is torque; RPM surf is the rotation speed; ROP is the mechanical drilling rate; K t is the screw torque coefficient; ΔP is the screw pressure difference; K n is the screw speed coefficient; Q is the drilling fluid displacement;

[0081] The mechanical arm can be corrected, and the corrected mechanical specific energy is:

[0082] MSE adj =E f MSE total ; (11)

[0083] In the formula, MSE adj is the corrected mechanical specific energy; E f is the dimensionless correction factor;

[0084] According to the wells of the same category corresponding to the target well to be optimized, the mechanical specific energy of the wells of the same category is calculated, and the drilling efficiency of the wells of the same category in each layer is determined in combination with the mechanical drilling speed;

[0085] The drilling parameter range is determined according to the drilling efficiency, mechanical specific energy and mechanical drilling speed, and the optimal adjacent wells corresponding to the target well to be optimized are selected.

[0086] In one embodiment, S105, optimizing the drilling parameters of the target well to be optimized using the drilling data of the optimal adjacent well to obtain the optimized drilling parameters includes:

[0087] Determine the torque calculation value based on the optimized drilling parameters:

[0088]

[0089] Where, TQ is the calculated torque value; μ is the dimensionless sliding friction coefficient; WOB' is the optimized drilling pressure; D is the drill bit diameter;

[0090] According to the optimized drilling pressure, speed and torque calculation values, a nonlinear regression model of mechanical drilling speed is established, the mechanical specific energy is calculated, and the optimized mechanical drilling speed and mechanical specific energy are compared with those before optimization to determine the comparison results before and after optimization;

[0091] According to the optimized mechanical drilling speed and mechanical specific energy, the optimization effect of drilling parameters is verified, and the calculation formula is:

[0092]

[0093] Where MAPE is the mean absolute percentage error; N' is the number of samples used in the validation; y i * is the i-th predicted value, corresponding to the mechanical drilling speed and mechanical specific energy predicted after optimization; y i It is the i-th true value, corresponding to the mechanical drilling speed and mechanical specific energy obtained from actual drilling construction.

[0094] It should be noted that, although the operations of the method of the present invention are described in a specific order in the above embodiments and the accompanying drawings, this does not require or imply that the operations must be performed in the specific order, or that all the operations shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0095] Embodiment 1:

[0096] Take a target block as an example. The target block has logging data X of N vertical wells. L ={X iL (m L ,n L )|i=1,2,…,N} and drilling data X D ={X iD (m D ,n D )|i=1,2,…,N}. X iL (m L ,n L ) is the i-th sample in the logging data, X iD (m D ,n D ) is the i-th sample in the drilling data.

[0097] The logging data and drilling data of different wells are calibrated at the same depth, and the number of samples m in each well is unified to obtain a complete data set X = {X iL (m,n L ),XiD(m,n D )|i=1,2,···,N}.

[0098] Since the logging data X of each well L ={X iL (m L ,n L )|i=1,2,…,N} is the number containing n L The matrix of logging parameters is directly represented by N m×n LThe matrix as the input of the clustering algorithm is relatively complex, so the logging parameters can be feature extracted and dimension reduced, and the logging data of N wells can be converted into N m×1 vectors, and then clustered. Figure 2 , which is a schematic diagram of the drilling parameter optimization process of a specific embodiment of the present invention. By analyzing the logging data, through dimensionality reduction processing (PCA) and cluster analysis (AP), similar wells of the same category of the target well to be optimized are obtained, and then the optimal adjacent wells are determined by analyzing the mechanical drilling speed and mechanical specific energy, and the drilling parameters of the adjacent wells are used to optimize the drilling parameters of the target well to be optimized.

[0099] Assume that the logging data X of the vertical well in the target block is L ={X iL (m L ,n L )|i=1,2,…,N}, and the principal component analysis was performed to extract k principal components Z with eigenvalues ​​greater than 1. i =a i,1 X1+a i,2 X2+…+a i,nL X nL (i=1,2,…,k), reference Figure 3 , is a schematic diagram of the steps of multi-dimensional clustering of well logging curves based on PCA-AP algorithm according to a specific embodiment of the present invention, such as Figure 3 As shown, the specific steps are:

[0100] Step 1: Calculate the comprehensive score model based on the principal component variance contribution rate and the principal component score model coefficient:

[0101]

[0102] In the formula, w i,j is the coefficient of the jth logging data in the scoring model of the i-th vertical well sample; W i l is the variance contribution rate of the lth principal component of the i-th vertical well sample; is the coefficient of the jth logging data in the lth principal component of the i-th vertical well sample; F i is the score model of the i-th vertical well sample, as the well logging curve of the i-th vertical well sample;

[0103] Step 2: Initialize the attraction matrix r(i, k) and the attribution matrix a(i, k) to zero matrices, set the maximum number of iterations T, the number of consecutive invariant cluster centers C, and the damping coefficient λ, the current number of iterations t = 0, and initialize the similarity matrix s:

[0104]

[0105] In the formula, s t(i,k) represents the similarity between the i-th vertical well sample and the k-th vertical well sample. t (k,k) takes the median of the kth row in the matrix; F i,j represents the jth element of the logging curve of the i-th vertical well sample; F k,j represents the jth element of the combined logging curve of the kth vertical well sample; m represents the number of samples.

[0106] Step 3, let t = t + 1, update the attraction matrix r (i, k) and the attribution matrix a (i, k):

[0107] r t (i,k)=s t (i,k)-max{a t (i,j)+s t (i,j)}, j∈{1,2,...,Nj≠k}; (4)

[0108]

[0109] In the formula, r t (i,k), a t (i,k),s t (i, k) is the attraction matrix, belonging matrix, and similarity matrix updated after the tth iteration; a t (i,j),s t (i, j) is the degree of attribution and similarity between the i-th vertical well sample and the j-th vertical well sample in the degree of attribution matrix and similarity matrix; r t (k,k), a t (k,k) takes the median of the kth row in the attraction matrix and the attribution matrix; r(j,k) represents the attraction between the jth vertical well sample and the kth vertical well sample in the attraction matrix.

[0110] Step 4, introduce the damping coefficient λ and calculate the decision matrix c(i,k):

[0111] r t (i,k)=λr t-1 (i,k)+(1-λ)r t (i,k); (7)

[0112] a t (i,k)=λa t-1 (i,k)+(1-λ)a t (i,k); (8)

[0113] c t (i,k)=r t (i,k)+a t (i,k); (9)

[0114] Step 5: Repeat steps 3 and 4 until t=T or the cluster center remains unchanged for C consecutive times.

[0115] The vertical well samples corresponding to the column index of the elements greater than 0 on the diagonal of the decision matrix c(i,k) are the cluster centers, and the vertical well samples corresponding to the row index of all elements greater than 0 in the column are of the same type. The above steps can realize the multi-dimensional clustering of the logging curves of the vertical wells in the target block, laying the foundation for subsequent analysis and optimization.

[0116] When analyzing and optimizing drilling parameters, the mechanical specific energy model is used. The mechanical specific energy is used to describe the mechanical energy required by the drill bit to break a unit volume of rock under the action of drilling pressure and torque. The standardized relationship of the mechanical specific energy model can be:

[0117]

[0118] In the formula, MSE total is the total mechanical specific energy, MPa; WOB surf is the drilling pressure, N; D is the drill diameter, mm; TQ surf is the torque, N·m; RPM surf is the rotation speed, r / min; ROP is the mechanical drilling speed, m / h; K t is the screw torque coefficient, N·m / MPa; ΔP is the screw pressure difference, MPa; K n is the screw speed coefficient, r / L; Q is the drilling fluid displacement, L / min.

[0119] In the actual drilling process, due to the influence of unfavorable factors such as friction, formation pressure, and drill string vibration, the rock breaking efficiency of the drill bit is generally only 30% to 40%. Therefore, it is usually necessary to add a correction factor Ef before the mechanical specific energy model. The corrected mechanical specific energy model is:

[0120] MSE adj =E f MSE total ; (11)

[0121] In the formula, MSE adj is the corrected mechanical specific energy, MPa; E f It is a dimensionless correction coefficient, generally taken as 0.35.

[0122] Based on the above processing process, for the case where there are N-1 neighboring well samples around the vertical well Yt to be optimized in the target block, the PCA-AP algorithm is used to reduce the dimension and cluster the logging data of these N vertical wells, and then the mechanical specific energy of the drilling process of similar wells is calculated, and then the drilling efficiency of similar wells in each layer is analyzed in combination with the mechanical drilling rate. Finally, the drilling parameter range with low mechanical specific energy and fast mechanical drilling rate is compared and optimized, providing a basis for optimizing the drilling parameters of the target well to be optimized.

[0123] To verify the optimization effect of drilling parameters, the torque is first calculated:

[0124]

[0125] Where TQ is the calculated torque value, kN·m; μ is the dimensionless sliding friction coefficient, which is 0.25 for roller bit and 0.5 for PDC bit; WOB' is the optimized bit pressure, kN; and D is the drill bit diameter.

[0126] Then, the nonlinear regression model of mechanical drilling speed is established with the optimized drilling pressure and rotation speed and the calculated torque as input, and the mechanical specific energy is calculated. Finally, the optimized mechanical drilling speed and mechanical specific energy are compared with the values ​​before optimization.

[0127] In order to measure the prediction effect of the prediction model of mechanical drilling speed and mechanical specific energy in the test set, the mean absolute percentage error (MAPE) is used as the evaluation index. The calculation formula of MAPE is as follows:

[0128]

[0129] Where MAPE is the mean absolute percentage error; N' is the number of samples used in the validation; is the i-th predicted value, corresponding to the mechanical drilling speed and mechanical specific energy predicted after optimization; y i It is the i-th true value, corresponding to the mechanical drilling speed and mechanical specific energy obtained from actual drilling construction.

[0130] Embodiment 2:

[0131] Taking Well Y11 as an example, the target well reservoir is located in the Longmaxi Formation of the Tiangongtang Structure in the Sichuan Basin, with a formation pressure coefficient of 1.15-1.74 and a completed well depth of 3410m. The formation is mainly composed of mudstone, sandstone and limestone. The Changxing-Longtan Formation (2010-2370m) contains tuffaceous siltstone with strong rock abrasiveness, and the Maokou Formation (2370-2660m) contains a small amount of flint with poor drillability.

[0132] In order to study the optimization of drilling parameters of the target well, samples from 10 neighboring wells around the well were collected. The logging parameters were acoustic time difference AC, wellbore CAL, compensated neutron CNL, density DEN, natural gamma GR, and resistivity RT. Based on the PCA-AP algorithm, the logging parameters were first standardized and dimensionally reduced. The logging curves before and after dimension reduction were compared. Figure 4A and Figure 4B As shown, Figure 4A This is a schematic diagram of a standardized logging curve according to a specific embodiment of the present invention. Figure 4B It is a schematic diagram of a standardized comprehensive logging curve (after dimensionality reduction) according to a specific embodiment of the present invention.

[0133] from Figure 4A and Figure 4B It can be seen that the comprehensive logging curve retains most of the original trend of the logging curve, and the dimensionality reduction effect is good. The following is based on the PCA-AP algorithm to cluster the comprehensive logging curves of 11 wells after dimensionality reduction. The maximum number of iterations is set to 1000, the damping factor is set to 0.9, and the maximum number of consecutive invariant times of the cluster center is set to 50. The clustering results are as follows: Figure 5 As shown, Figure 5 It is a schematic diagram of the comprehensive logging curve clustering result according to a specific embodiment of the present invention.

[0134] like Figure 5 As shown in the figure, the PCA-AP algorithm clusters the 11 vertical wells in the target block into 3 categories, with cluster centers Y4, Y6, and Y9 respectively. The cluster center of the target well Y11 is Y9, and the same wells also include Y2 and Y7. The following calculates the mechanical specific energy of the same wells and combines the mechanical drilling rate to optimize the drilling parameter range of each layer. Taking the target well Y11 as an example, the mechanical specific energy and mechanical drilling rate of the well are as follows: Fig. 6A and 6B As shown, Fig. 6A This is a schematic diagram of the mechanical drilling rate results of the Y11 well after smoothing in a specific embodiment of the present invention. Figure 6B It is a schematic diagram of the mechanical specific energy results of the Y11 well after smoothing according to a specific embodiment of the present invention.

[0135] The smoothed mechanical drilling speed and mechanical specific energy retain the original trend of the data and are more intuitive, such as Fig. 6A and 6B As shown in the figure, the mechanical drilling speed of Well Y11 is slow and the mechanical specific energy is high in the two sections of 500-900m and 2100-2700m, and the drilling efficiency is low in a large section. Among them, the well depth of 500-900m corresponds to the Xujiahe Formation, and the well depth of 2100-2700m includes the Changxing-Longtan Formation and the Maokou Formation.

[0136] The mechanical drilling speed and mechanical specific energy of three similar wells Y2, Y7 and Y9 around the Y11 well were compared in these layers. Taking the Xujiahe Formation as an example, the results are as follows: Figure 7 and Figure 8 As shown, Figure 7 This is a comparison chart of mechanical drilling speed of the Xujiahe Formation Y11 well of the same type according to a specific embodiment of the present invention. Figure 8 This is a comparison diagram of mechanical specific energy of similar wells in the Xujiahe Formation Y11 according to a specific embodiment of the present invention.

[0137] from Figure 7 and Figure 8 It can be seen that among the four wells, the target well Y11 has the slowest mechanical drilling speed in the Xujiahe Formation and the highest mechanical specific energy. In contrast, the Y7 well has the fastest mechanical drilling speed, the lowest mechanical specific energy, and the highest drilling efficiency among its peers, making it the optimal adjacent well of Y11. The mechanical drilling speed and mechanical specific energy of the four wells in the Changxing-Longtan Formation and the Maokou Formation were analyzed in the same way, and the optimal adjacent wells of the target well and their corresponding drilling parameters were found. The specific results are shown in Table 1.

[0138] Table 1 Optimal adjacent well drilling parameters results

[0139]

[0140] As shown in Table 1, based on the drilling parameters of the optimal adjacent wells in the same formation, the drilling pressure and drilling speed should be appropriately increased in the Xujiahe Formation for the target well Y11, the drilling pressure should be appropriately increased and the drilling speed should be slightly increased in the Changxing-Longtan Formation, and the rotation speed should be reduced in the Maokou Formation with poor drillability to reduce drill bit wear while appropriately increasing the drilling pressure to obtain a faster mechanical drilling speed.

[0141] According to the above optimization principles, the drilling parameters of the Y11 well in three layers were adjusted. To verify the optimization effect, a Gaussian regression model of mechanical drilling speed was established using the shallow (depth less than 2370m) drilling parameters of the Longtan Formation of the Y11 well. The model was verified with the data of the Maokou Formation. The prediction effect is as follows: Fig. 9 As shown in the figure, the effect after optimization is as follows Fig. 10A and Fig. 10B As shown, Fig. 9 : is a mechanical drilling rate and mechanical specific energy prediction result diagram of a specific embodiment of the present invention, ROP is the mechanical drilling rate, MSE is the mechanical specific energy, MAPE is the mean absolute percentage error, Fig. 10A This is a diagram showing the optimization effect of drilling parameters (mechanical drilling speed) of the Y11 well according to a specific embodiment of the present invention. Fig. 10B This is a diagram showing the optimization effect of the Y11 well drilling parameters (mechanical specific energy) according to a specific embodiment of the present invention.

[0142] from Fig. 9 It can be seen that the prediction results of mechanical penetration rate and mechanical specific energy are basically consistent with the true value trend, and the predicted MAPE are 13.40% and 5.78% respectively.

[0143] Fig. 10A and Fig. 10BIt shows that after the optimization of drilling parameters of Well Y11, the average mechanical drilling speed increased by 11.27% and the average mechanical specific energy decreased by 12.90%. Among them, the Xujiahe Formation and Changxing-Longtan Formation have better drillability and lower rock strength, so the improvement of drilling efficiency is more significant than that of the Maokou Formation. For new wells to be drilled in the target block, the logging data of the current drilled layers can be clustered, and the drilling parameters of the optimal adjacent wells can be used to optimize the subsequent potential layers with poor drillability.

[0144] In summary, the drilling parameter optimization method based on cluster analysis of the present invention can achieve at least the following effects after optimizing drilling parameters:

[0145] 1. The PCA-AP algorithm is used to realize multi-dimensional clustering of logging curves. The dimension reduction process retains a large amount of original information. The clustering process automatically generates cluster centers and converges quickly.

[0146] 2. For drilling well Y11, the drillability of the Xujiahe Formation and Changxing-Longtan Formation is relatively good, and the drilling pressure and rotation speed should be appropriately increased. The drillability of the Maokou Formation is relatively poor, and the rotation speed should be slightly increased while reducing the drilling pressure.

[0147] 3. The drilling parameters of Well Y11 were optimized based on the optimal drilling parameter range of adjacent wells. The average mechanical drilling speed after optimization increased by 11.27% and the average mechanical specific energy decreased by 12.90% compared with before optimization, thus improving the drilling efficiency.

[0148] After introducing the method of the exemplary embodiment of the present invention, next, refer to Fig.11 A drilling parameter optimization device based on cluster analysis according to an exemplary embodiment of the present invention is introduced.

[0149] The implementation of the drilling parameter optimization device based on cluster analysis can refer to the implementation of the above method, and the repeated parts will not be repeated. The term "module" or "unit" used below can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0150] Based on the same inventive concept, the present invention also proposes a drilling parameter optimization device based on cluster analysis, such as Fig.11 As shown, the device comprises:

[0151] The data acquisition module 1110 is used to acquire the logging data and drilling data of the adjacent wells in the target block where the target well to be optimized is located;

[0152] A dimension reduction processing module 1120 is used to perform dimension reduction processing on the well logging data of the target well to be optimized and the well logging data of the adjacent wells to obtain a well logging curve after dimension reduction processing;

[0153] A cluster analysis module 1130 is used to perform cluster analysis based on the well logging curves to determine similar wells of the same category as the target well to be optimized;

[0154] The data analysis module 1140 is used to screen out the drilling parameter range corresponding to each layer in the drilling data according to the mechanical specific energy and mechanical drilling speed of similar wells, and determine the optimal adjacent well corresponding to the target well to be optimized through the drilling parameter range;

[0155] The parameter optimization module 1150 is used to optimize the drilling parameters of the target well to be optimized by using the drilling data of the optimal adjacent well to obtain the optimized drilling parameters.

[0156] It should be noted that, although several modules of the drilling parameter optimization device based on cluster analysis are mentioned in the above detailed description, such division is only exemplary and not mandatory. In fact, according to an embodiment of the present invention, the features and functions of two or more modules described above can be embodied in one module. Conversely, the features and functions of one module described above can be further divided into multiple modules for embodiment.

[0157] Based on the above invention concept, Fig.12 As shown, the present invention also proposes a computer device 1200, including a memory 1210, a processor 1220, and a computer program 1230 stored in the memory 1210 and executable on the processor 1220, wherein the processor 1220 implements the aforementioned drilling parameter optimization method based on cluster analysis when executing the computer program 1230.

[0158] Based on the aforementioned inventive concept, the present invention proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the aforementioned drilling parameter optimization method based on cluster analysis is implemented.

[0159] Based on the aforementioned inventive concept, the present invention proposes a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, a drilling parameter optimization method based on cluster analysis is implemented.

[0160] The drilling parameter optimization method and device based on cluster analysis proposed in the present invention obtain the logging data and drilling data of the adjacent wells in the target block where the target well to be optimized is located; perform dimensionality reduction processing based on the logging data of the target well to be optimized and the logging data of the adjacent wells to obtain the logging curve after dimensionality reduction processing; perform cluster analysis based on the logging curve to determine the same type of wells as the target well to be optimized; based on the mechanical specific energy and mechanical drilling speed of the same type of wells, screen out the drilling parameter range corresponding to each layer in the drilling data, and determine the optimal adjacent well corresponding to the target well to be optimized through the drilling parameter range; use the drilling data of the optimal adjacent well to optimize the drilling parameters of the target well to be optimized to obtain the optimized drilling parameters. The overall solution fully exploits the deep geological correlation between wells at the same layer, conducts more accurate drilling parameter optimization analysis, and performs multi-dimensional clustering of logging curves of vertical wells in the target block through unsupervised learning methods. On this basis, the mechanical drilling rate of each section of similar wells is analyzed in combination with the mechanical specific energy theory, and the optimal drilling parameter range for each layer is further determined, which is verified by typical wells. The present invention combines the mechanical specific energy theory to carry out targeted optimization of the drilling parameters of each layer of the target well to be optimized, effectively improves the drilling efficiency based on the concept of geological engineering integration, and provides strong technical and data support for the drilling field.

[0161] The acquisition, storage, use, and processing of data in the technical solution of this application comply with the relevant provisions of laws and regulations.

[0162] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0163] The present invention is described with reference to flowcharts and / or block diagrams of methods and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0164] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0165] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0166] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A drilling parameter optimization method based on cluster analysis, characterized in that: include: For the target well to be optimized, the logging data and drilling data of the adjacent wells in the target block are obtained; Performing dimensionality reduction processing on the logging data of the target well to be optimized and the logging data of adjacent wells to obtain a logging curve after dimensionality reduction processing; Performing cluster analysis based on the well logging curves to determine similar wells of the same category as the target well to be optimized; According to the mechanical specific energy and mechanical drilling speed of similar wells, the drilling parameter range corresponding to each layer is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range; The drilling data of the optimal adjacent well is used to optimize the drilling parameters of the target well to be optimized, so as to obtain optimized drilling parameters.

2. The drilling parameter optimization method based on cluster analysis according to claim 1, characterized in that: For the target well to be optimized, the logging data and drilling data of the adjacent wells in the target block are obtained, including: For different adjacent wells, the logging data and drilling data are calibrated according to the same depth; wherein the logging data at least includes acoustic wave time difference, well diameter, compensated neutron, density, natural gamma, and resistivity; and the drilling data at least includes drilling speed, drilling pressure, and rotation speed.

3. The drilling parameter optimization method based on cluster analysis according to claim 1, characterized in that: Performing dimension reduction processing on the logging data of the target well to be optimized and the logging data of the adjacent wells to obtain the logging curve after dimension reduction processing, including: For the well logging data, the n-dimensional features of the well logging data are mapped into k-dimensional orthogonal features; wherein n>k, n and k are positive integers, and the information in the k-dimensional orthogonal features does not overlap; The k-dimensional features are subjected to multiple dimensionality reduction processing by feature extraction to obtain the corresponding logging curve reduced to one dimension; wherein, when the dimensionality reduction processing is performed by feature extraction, the coefficients in each principal component score model are weighted averaged according to the variance contribution of the k principal components to obtain the dimensionality reduction result.

4. The drilling parameter optimization method based on cluster analysis according to claim 1, characterized in that: Performing cluster analysis based on the well logging curves to determine similar wells of the same category as the target well to be optimized includes: The scoring model is constructed according to the logging curve, the principal component variance contribution rate and the principal component scoring model coefficient, and the relationship is: In the formula, w i,j is the coefficient of the jth well logging data in the scoring model of the i-th vertical well sample; W i l is the variance contribution rate of the lth principal component of the i-th vertical well sample; is the coefficient of the jth logging data in the lth principal component of the i-th vertical well sample; F i is the scoring model of the i-th vertical well sample, as the well logging curve of the i-th vertical well sample; k represents the feature dimension; N is the number of vertical wells; n L is the number of logging data; X i,j represents the jth logging data of the i-th vertical well sample; Initialize the attraction matrix r(i, k) and the attribution matrix a(i, k) to zero matrices, set the maximum number of iterations T, the number of consecutive invariant cluster centers C, and the damping coefficient λ, the current number of iterations t = 0, and initialize the similarity matrix: In the formula, s t (i,k) represents the similarity between the i-th vertical well sample and the k-th vertical well sample. t (k,k) takes the median of the kth row in the matrix; F i,j represents the jth element of the logging curve of the i-th vertical well sample; F k,j represents the jth element of the combined logging curve of the kth vertical well sample; m represents the number of samples; In each iteration, let t increase by 1 and update the attraction matrix r(i,k) and the attribution matrix a(i,k): r t (i,k)=s t (i,k)-max{a t (i,j)+s t (i,j)},j∈{1,2,...,Nj≠k};(4) In the formula, r t (i,k), a t (i,k),s t (i, k) is the attraction matrix, belonging matrix, and similarity matrix updated after the tth iteration; a t (i,j),s t (i, j) is the degree of attribution and similarity between the i-th vertical well sample and the j-th vertical well sample in the degree of attribution matrix and similarity matrix; r t (k,k), a t (k,k) takes the median of the kth row in the attraction matrix and the attribution matrix; r(j,k) represents the attraction between the jth vertical well sample and the kth vertical well sample in the attraction matrix; Introduce the damping coefficient λ and calculate the decision matrix c(i,k): r t (i,k)=λr t-1 (i,k)+(1-λ)r t (i,k);(7) a t (i,k)=λa t-1 (i,k)+(1-λ)a t (i,k);(8) c t (i,k)=r t (i,k)+a t (i,k);(9) In the formula, c t (i, k) represents the decision matrix obtained after the tth iteration; a t-1 (i,k),r t-1 (i, k) represents the attraction matrix and belonging matrix updated after the t-1th iteration; After multiple iterations, the update is continued until the number of iterations t = T or the cluster center remains unchanged for C consecutive times; According to the decision matrix c(i, k) obtained by iterative updating, the vertical well samples corresponding to the column index where the elements on the diagonal are greater than 0 are taken as cluster centers, and the vertical well samples corresponding to the row index where all the elements greater than 0 corresponding to the column are taken as the same category.

5. The drilling parameter optimization method based on cluster analysis according to claim 1, characterized in that: According to the mechanical specific energy and mechanical drilling speed of similar wells, the drilling parameter range corresponding to each layer is screened out from the drilling data, and the optimal adjacent well corresponding to the target well to be optimized is determined through the drilling parameter range, including: For the mechanical specific energy and mechanical drilling speed of drilling, the relationship is: In the formula, MSE total Indicates mechanical specific energy; WOB surf is the drilling pressure; D is the drill bit diameter; TQ surf is torque; RPM surf is the rotation speed; ROP is the mechanical drilling rate; K t is the screw torque coefficient; ΔP is the screw pressure difference; K n is the screw speed coefficient; Q is the drilling fluid displacement; The mechanical arm can be corrected, and the corrected mechanical specific energy is: MSE adj =E f MSE total ;(11) In the formula, MSE adj is the corrected mechanical specific energy; E f is the dimensionless correction factor; According to the wells of the same category corresponding to the target well to be optimized, the mechanical specific energy of the wells of the same category is calculated, and the drilling efficiency of the wells of the same category in each layer is determined in combination with the mechanical drilling speed; The drilling parameter range is determined according to the drilling efficiency, mechanical specific energy and mechanical drilling speed, and the optimal adjacent wells corresponding to the target well to be optimized are selected.

6. The drilling parameter optimization method based on cluster analysis according to claim 1, characterized in that: The drilling data of the optimal adjacent well is used to optimize the drilling parameters of the target well to be optimized, and the optimized drilling parameters are obtained, including: Determine the torque calculation value based on the optimized drilling parameters: Where, TQ is the calculated torque value; μ is the dimensionless sliding friction coefficient; WOB' is the optimized drilling pressure; D is the drill bit diameter; According to the optimized drilling pressure, speed and torque calculation values, a nonlinear regression model of mechanical drilling speed is established, the mechanical specific energy is calculated, and the optimized mechanical drilling speed and mechanical specific energy are compared with those before optimization to determine the comparison results before and after optimization; According to the optimized mechanical drilling speed and mechanical specific energy, the optimization effect of drilling parameters is verified, and the calculation formula is: Where MAPE is the mean absolute percentage error; N' is the number of samples used in the validation; y i * is the i-th predicted value, corresponding to the mechanical drilling speed and mechanical specific energy predicted after optimization; y i It is the i-th true value, corresponding to the mechanical drilling speed and mechanical specific energy obtained from actual drilling construction.

7. A drilling parameter optimization device based on cluster analysis, characterized in that: include: A data acquisition module is used to acquire the logging data and drilling data of the adjacent wells in the target block where the target well to be optimized is located; A dimension reduction processing module is used to perform dimension reduction processing on the well logging data of the target well to be optimized and the well logging data of the adjacent wells to obtain a well logging curve after dimension reduction processing; A cluster analysis module, used for performing cluster analysis according to the well logging curves to determine similar wells of the same category as the target well to be optimized; A data analysis module is used to screen out the drilling parameter range corresponding to each layer in the drilling data according to the mechanical specific energy and mechanical drilling speed of similar wells, and determine the optimal adjacent well corresponding to the target well to be optimized through the drilling parameter range; The parameter optimization module is used to optimize the drilling parameters of the target well to be optimized by using the drilling data of the optimal adjacent well to obtain the optimized drilling parameters.

8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 6 is implemented.

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 a processor, the method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images