A method and system for determining perforation position of oil layer

Through clustering and probabilistic modeling of multi-source data, combined with Markov model, the problem of low accuracy in the prediction of perforation layer of the oil layer is solved, and more efficient decision-making support for oil layer mining is achieved.

CN118797392BActive Publication Date: 2024-12-06XIAN MOKO XINGYE PETROLEUM ENG TECH CO LTD
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411273868.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-12
Publication Date
2024-12-06
Estimated Expiration
2044-09-12

AI Technical Summary

Technical Problem

In the prior art, the accuracy of the oil layer perforation layer prediction is not high, and there is a problem of human factors involved.

Method used

Clustering and probabilistic modeling is performed by acquiring multi-source data (electrical logging data, acoustic logging data, density logging data and neutron logging data), using the Markov model to predict potential levels, combining cluster probability and label The final potential level is calculated by counting the final potential level and selecting the layer with the greatest potential as the ideal perforation layer.

Benefits of technology

It improves the prediction accuracy and reliability of the perforation layer of the oil layer, reduces manual intervention, and improves the efficiency and economic benefits of oil well mining.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118797392B_ABST
    Figure CN118797392B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of data processing, and the present invention relates to a method and system for determining the perforation layer of an oil layer, comprising: obtaining multi-source data of the perforation layer of the current oil layer and the next oil layer and corresponding labels, clustering the multi-source data of the perforation layer of the current oil layer and the next oil layer respectively, obtaining corresponding multiple cluster clusters, and calculating the cluster probability that each multi-source data of the corresponding oil layer belongs to each cluster cluster, respectively calculating the potential level of the perforation layer of the current oil layer and the next oil layer, and using a Markov model to predict the predicted potential level of the perforation layer of the next oil layer, taking the average of the predicted potential level and the potential level of the next oil layer as the final potential level of the next perforation layer, and responding to the layer with the largest final potential level as the ideal perforation layer. The present invention improves the problem of the accuracy of predicting the perforation layer of an oil layer.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and more specifically, to a method and system for determining the perforation layer position of an oil layer. Background Art

[0002] In the process of oil production, determining the perforation layer of the oil layer is crucial to improving the production efficiency and economic benefits of the oil well. Reasonable determination of the perforation layer can not only effectively improve the recovery rate of the oil layer, but also significantly reduce operating costs, while improving the safety and stability of the oil well. The optimization of perforation technology is of great significance for maximizing the economic return of the oil well and reducing environmental impact. Therefore, improving the accuracy and reliability of perforation layer detection is a key technical problem in oil exploration and production.

[0003] With the development of science and technology, especially the significant progress in measurement technology and computing power, more and more intelligent and automated technologies have been introduced into the process of determining perforation positions. These technologies include advanced methods such as big data analysis, machine learning, and artificial intelligence. Through deep mining and comprehensive analysis of a large amount of measurement data, these technologies can effectively identify formation characteristics and oil layer distribution, thereby improving the accuracy of perforation position determination. For example, machine learning algorithms can establish accurate prediction models through training of historical data, help dynamically adjust perforation positions during actual drilling, and optimize the production effect of oil wells. Although intelligent and automated technologies have shown great potential in perforation position determination, their application still faces many challenges.

[0004] The Chinese patent application document with publication number CN106504107A discloses a method and device for determining the perforation layer of an oil layer. The invention obtains the logging data and mud logging data of the current well; identifies the low-resistance oil layer in the current well according to the preset low-resistance oil layer database and the logging data and mud logging data; selects the to-be-filled perforation layer of the low-resistance oil layer in the current well; calculates the production potential of the to-be-filled perforation layer of the low-resistance oil layer according to preset rules; and determines the perforation layer of the oil layer according to the calculated production potential. However, due to the intervention of human factors, this method has the problem of low accuracy in predicting the perforation layer of the oil layer. Summary of the invention

[0005] In order to solve the problem of low accuracy in predicting the perforation level of an oil layer raised in the above background technology, the present invention provides solutions in the following aspects.

[0006] In a first aspect, the present invention provides a method for determining a perforation level of an oil layer, comprising:

[0007] The multi-source data of the perforation layers of the current oil layer and the next oil layer and their corresponding labels are obtained, and the labels are divided into large, medium and small, indicating the suspected mining potential level corresponding to the multi-source data; the multi-source data of the perforation layers of the current oil layer and the next oil layer are clustered respectively to obtain the corresponding multiple cluster clusters, and the cluster probability that each multi-source data of the corresponding oil layer belongs to each cluster cluster is calculated; the potential level of the perforation layers of the current oil layer and the next oil layer is calculated respectively, and the potential level is positively correlated with the cluster probability corresponding to each cluster cluster of the corresponding oil layer and the number of the same label in each cluster cluster, and is negatively correlated with the number of all labels in each cluster cluster; based on the potential level of the perforation layer of the current oil layer, the predicted potential level of the perforation layer of the next oil layer is predicted by using the Markov model; the average of the predicted potential level and the potential level of the next oil layer is taken as the final potential level of the perforation layer of the next oil layer; the layer with the largest final potential level is the ideal perforation layer.

[0008] The above technical solution fully considers the characteristic distribution of multi-source data, obtains the potential level and predicted potential level of the next oil layer from two perspectives, comprehensively considers the potential level and predicted potential level to determine the final potential level, and comprehensively analyzes from multiple perspectives to improve the accuracy and reliability of the evaluation of the next oil layer, and selects the perforation layer of the oil layer corresponding to the optimal exploitation potential.

[0009] Furthermore, the potential level of the perforation layer of the current layer is calculated , the formula is:

[0010] ;

[0011] In the formula, Indicates The multi-source data of the perforation layer belongs to The cluster probability corresponding to the clusters, Indicates The first The number of tags, Indicates The number of all labels in the clusters, Represents the total number of clusters.

[0012] The above technical solution comprehensively considers the label distribution of different clusters by utilizing the potential level formula that combines cluster probability and label quantity, provides a detailed evaluation of the perforation layer potential, enhances the quantification ability of the perforation layer potential of the oil layer, and improves the accuracy of the prediction.

[0013] Furthermore, the multi-source data is clustered. The specific clustering process is as follows: constructing a covariance matrix based on the multi-source data and performing eigenvalue decomposition to obtain eigenvalues ​​and their corresponding eigenvectors; using principal component analysis to obtain the weight of each eigenvector; randomly selecting The feature vectors are used as the initial cluster centers, the weighted distance from each feature vector to the cluster center is calculated, and each feature vector is assigned to the cluster center closest to it. The center point of each cluster center is recalculated and used as the new cluster center until the movement of the cluster center is less than the set threshold α, where α∈[0.001,0.01]; based on the new cluster center, the multi-source data are clustered to obtain multiple clusters.

[0014] The above technical solution clusters the feature vectors by introducing the concept of weighted distance, ensuring that each feature vector is reasonably distributed according to its importance during the clustering process. By randomly selecting the initial cluster center and gradually iterating and optimizing, the position of the cluster center is continuously adjusted until the set threshold α is met. This process effectively reduces the possibility of local optimality and ensures the stability and accuracy of the final clustering results. Through this optimized clustering method, data clusters can be divided more accurately and the accuracy of oil layer perforation layer selection can be enhanced.

[0015] Furthermore, it also includes normalizing multi-source data.

[0016] Furthermore, the multi-source data includes: electrical logging data, sonic logging data, density logging data and neutron logging data.

[0017] Furthermore, the multi-source data of the current layer and the next oil layer perforation layer are obtained by using a micro resistivity logging tool, an array sonic logging tool, a gamma density logging tool and a compensated neutron logging tool.

[0018] Furthermore, the multi-source data is clustered using a K-means clustering algorithm.

[0019] The above technical solution uses K-means clustering algorithm to cluster multi-source data, which can efficiently process large-scale data, quickly complete clustering tasks, and improve the data grouping accuracy through iterative optimization. This makes the analysis of oil layer perforation layers faster and more accurate, thereby improving the efficiency and reliability of decision-making.

[0020] Furthermore, the cluster probability is: estimating the data distribution parameters of each cluster cluster by maximum likelihood; calculating the mean vector and covariance matrix of each cluster cluster, and constructing a Gaussian probability density function representing the cluster cluster data distribution; and calculating the cluster probability of each cluster cluster using the Gaussian probability density function.

[0021] The above technical solution can more accurately describe the data distribution of each cluster, and based on this, calculate the cluster probability of each cluster. Ultimately, with the help of accurate probability evaluation, the prediction accuracy of the perforation layer of the oil layer is improved, making the selection of the perforation layer of the oil layer more scientific and reasonable.

[0022] In a second aspect, the present invention provides a system for determining perforation levels of oil layers, comprising a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, any one of the above methods for determining perforation levels of oil layers is implemented.

[0023] The beneficial effects of the present invention are:

[0024] The present invention provides an accurate and efficient method for selecting perforation positions in oil layers. By effectively integrating multidimensional information from multiple data sources and combining advanced cluster analysis, probability modeling and sequence prediction algorithms, the accuracy of data processing and the reliability of the decision-making process are greatly improved. Through this comprehensive method, the potential characteristics of the perforation positions in the oil layer can be better captured, so that the accuracy of the prediction is significantly optimized, thereby improving the accuracy and scientificity of the perforation position selection under complex geological conditions. In addition, the present invention can also effectively reduce manual intervention in practical applications, improve operating efficiency, and provide more reliable technical support for oil layer development decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0026] Figure 1 is a flow chart schematically illustrating a method for determining a perforation level of an oil layer according to an embodiment of the present invention;

[0027] Figure 2 It is a block diagram schematically showing a system structure for determining the perforation level of an oil layer according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0029] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0030] An embodiment of a method for determining perforation levels of an oil layer.

[0031] like Figure 1 As shown, a flow chart of a method for determining the perforation level of an oil layer according to an embodiment of the present invention includes the following steps:

[0032] S1: Obtain multi-source data of the perforation layers of the current oil layer and the next oil layer and their corresponding labels.

[0033] Among them, the current oil layer and the next oil layer represent different depth ranges in the oil layer, and the multi-source data is used to evaluate the mining potential of the corresponding oil layer, wherein the multi-source data includes electrical logging data, acoustic logging data, density logging data, and neutron logging data. In this embodiment, the relationship between each multi-source data and the mining potential is obtained by marking the quality of each data in the multi-source data. The relationship is represented by a label, that is, the label is divided into large, medium, and small, indicating the suspected mining potential level corresponding to the multi-source data. For example, if the logging result shows that the electrical logging data is high, the label is large; if the logging result shows that the electrical logging data is moderate, the label is medium; if the logging result shows that the electrical logging data is low, the label is small.

[0034] In this embodiment, the division can be performed by artificial setting, and of course, the division can also be performed by statistical analysis or machine learning.

[0035] In one embodiment, due to the difference in units and magnitudes of different data sources, it is necessary to normalize the multi-source data. For example, the data normalization can use the standard score method to eliminate the influence of the scales of different data sources. The standardized multi-source data is used to construct a covariance matrix, and the covariance matrix is ​​subjected to eigenvalue decomposition to extract eigenvalues ​​and eigenvectors. The eigenvalues ​​represent the variance of each principal component, and the eigenvectors represent the directions of the principal components. After the eigenvalues ​​are sorted, the eigenvectors corresponding to the larger eigenvalues ​​represent the directions with the largest variances in the data. The principal components are selected according to the size of the eigenvalues, and the original multi-source data are projected onto the selected principal components to achieve dimensionality reduction, so as to extract key information from the complex multi-source data.

[0036] It also includes the use of microresistivity logging tools, array sonic logging tools, gamma density logging tools and compensated neutron logging tools to obtain multi-source data of the perforation layers of the current oil layer and the next oil layer.

[0037] Different logging data types reflect different properties of oil layers. Electrical logging data can reflect the conductivity or resistivity of the oil layer, acoustic logging data can be used to evaluate the sound velocity of the rock formation, density logging data reveals the density distribution of the formation, and neutron logging data provides information about the porosity of the formation. By collecting these complementary multi-source data, the comprehensive properties of the oil layer can be comprehensively evaluated from multiple dimensions, avoiding the one-sidedness that may be caused by a single data source.

[0038] S2: Cluster the multi-source data of the perforation layers of the current oil layer and the next oil layer to obtain corresponding multiple clusters.

[0039] In one embodiment, the K-means clustering algorithm can be used to divide the multi-source data into multiple clusters. The specific clustering process is: constructing a covariance matrix based on the multi-source data and performing eigenvalue decomposition to obtain eigenvalues ​​and their corresponding eigenvectors; using principal component analysis to obtain the weight of each eigenvector;

[0040] Using principal component analysis to extract feature vector weights can ensure that important features are given higher weights during the clustering process, making key features more influential in the clustering process; randomly selecting The feature vectors are used as the initial cluster center to avoid the local optimal problem that may be caused by the fixed selection of the initial center. The weighted distance from each feature vector to the cluster center is calculated. Through this weighted method, the importance of different features is considered when calculating the distance, ensuring the accuracy of clustering. Each feature vector is assigned to the nearest cluster center. Such a data allocation process can ensure that similar feature vectors are classified into the same cluster cluster. The center point of each cluster center is recalculated and used as the new cluster center until the movement of the cluster center is less than the set threshold α, where α∈[0.001,0.01]; based on the new cluster center, the multi-source data is clustered to obtain multiple clusters.

[0041] Using principal component analysis to project the original high-dimensional multi-source data onto a few principal components can effectively extract the most representative features of the data, thereby reducing the dimension of the data and reducing redundant information.

[0042] This embodiment combines the principal component analysis method with the K-means clustering algorithm, which not only improves the clustering accuracy and calculation efficiency, but also lays a foundation for the accurate evaluation of the perforation layer of the oil layer.

[0043] In another embodiment, a hierarchical clustering algorithm can be used to divide multi-source data into multiple clusters. The specific clustering process is: using agglomerative hierarchical clustering, each feature vector is first treated as a separate cluster, and then the two closest clusters are gradually merged until a predetermined number of clusters or a distance threshold is reached. Unlike the K-means clustering algorithm, hierarchical clustering does not require the number of clusters to be specified in advance, and by constructing a clustering tree, the clustering structure can be observed at different levels. This method can discover nested structures in data and is suitable for irregularly shaped cluster distributions.

[0044] S3: Calculate the cluster probability that each multi-source data of the corresponding oil layer belongs to each cluster cluster and the distribution of its corresponding label, and calculate the potential level of the perforation layer of the current oil layer.

[0045] In one embodiment, a Gaussian model may be used to model the multiple clusters to obtain the cluster probability of each cluster and the distribution of its corresponding labels. The specific process is as follows:

[0046] The data distribution parameters of each cluster are estimated by maximum likelihood; the mean vector and covariance matrix of each cluster are calculated, and a Gaussian probability density function representing the data distribution of the cluster is constructed; the cluster probability of each cluster is calculated using the Gaussian probability density function, and the number of labels in each cluster is counted.

[0047] Exemplarily, the result output by the Gaussian model is [0.45, 0.3, 0.25], where 0.45 represents the probability that the potential level of the perforation layer is large, 0.3 represents the probability that the potential level of the perforation layer is medium, and 0.25 represents the probability that the potential level of the perforation layer is small.

[0048] Specifically, calculate the potential level of the perforation layer of the current oil layer , the formula is:

[0049] ;

[0050] In the formula, Indicates The multi-source data of the perforation layer belongs to The cluster probability corresponding to the clusters, Indicates The first The number of tags, Indicates The number of all labels in the clusters, Represents the total number of clusters.

[0051] Exemplarily, assuming that the current oil layer is clustered into 3 clusters, the probability and number of labels of each cluster are as follows: the output result of cluster 1 is: [0.4, 0.3, 0.3], where the number of large labels is 4, the number of medium labels is 3, and the number of small labels is 3; the output result of cluster 2 is: [0.3, 0.4, 0.3], where the number of medium labels is 4, the number of large labels is 3, and the number of small labels is 3; the output result of cluster 3 is: [0.3, 0.3, 0.4], where the number of small labels is 4, the number of large labels is 3, and the number of medium labels is 3. Substituting into the formula, the potential level of the current oil layer perforation layer is 1.02.

[0052] S4: Based on the perforation layer potential of the current oil layer, the Markov model is used to predict the potential of the next oil layer, and the average of the predicted value and the actual potential is taken as the final potential level. Finally, the layer with the greatest potential is selected as the ideal perforation layer.

[0053] In one embodiment, the potential levels of the perforation layers of the current oil layer and the next oil layer are calculated based on the above potential level formula. Based on the potential level of the perforation layer of the current oil layer, the potential level of the perforation layer of the next oil layer is predicted using the Markov model. The Markov model is a common prediction model used to infer the future state based on the current state. Therefore, the Markov model can be used to predict the selection of the perforation layer, and the average of the predicted potential level and the potential level of the next oil layer is used as the final potential level of the next perforation layer. The layer with the largest potential level is the ideal perforation layer.

[0054] The present invention significantly improves the accuracy and scientificity of oil layer perforation position selection. It can not only process complex data sets more effectively and accurately predict layer potential, but also reduce errors and improve the reliability of decision-making by optimizing clustering and modeling methods, thereby optimizing oil layer development effects and economic benefits.

[0055] An embodiment of a system for determining perforation levels in an oil formation.

[0056] like Figure 2 As shown, a system structure block diagram of determining the perforation level of an oil layer according to an embodiment of the present invention includes a processor and a memory.

[0057] The present invention also provides a system for determining the perforation level of an oil layer. Figure 2 As shown, the system includes a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for determining the perforation layer of an oil layer according to the first aspect of the present invention is implemented.

[0058] The system for determining the perforation level of an oil layer also includes other components well known to those skilled in the art, such as a communication interface, and the configuration and functions of the components are known in the art, so they will not be described in detail here.

[0059] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that may be stored or otherwise maintained by such a computer-readable medium.

[0060] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0061] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A method for determining the perforation level of an oil layer, characterized in that: include: Obtain multi-source data of perforation layers of the current oil layer and the next oil layer and their corresponding labels, wherein the labels are classified into large, medium and small, indicating the suspected mining potential level corresponding to the multi-source data; Clustering the multi-source data of the perforation layer of the current oil layer and the next oil layer respectively to obtain a plurality of corresponding clusters, and calculating the cluster probability that each multi-source data of the corresponding oil layer belongs to each cluster; The potential levels of the perforation layers of the current oil layer and the next oil layer are calculated respectively. The potential levels are positively correlated with the cluster probabilities corresponding to the clusters of the corresponding oil layer and the number of the same labels in each cluster, and are negatively correlated with the number of all labels in each cluster; Calculate the potential level of the perforated layer in the current oil layer , the formula is: ; In the formula, Indicates The multi-source data of the perforation layer belongs to The cluster probability corresponding to the clusters, Indicates The first The number of tags, Indicates The number of all labels in the clusters, Represents the total number of clusters; Based on the potential level of the perforation layer of the current oil layer, the Markov model is used to predict the predicted potential level of the perforation layer of the next oil layer; The average of the predicted potential level and the potential level of the next oil layer is taken as the final potential level of the next perforated layer; The layer with the largest final potential level response is the ideal perforation layer.

2. A method for determining the perforation level of an oil layer according to claim 1, characterized in that: The multi-source data is clustered, and the specific clustering process is: Constructing a covariance matrix based on the multi-source data and performing eigenvalue decomposition to obtain eigenvalues ​​and their corresponding eigenvectors; The principal component analysis method is used to obtain the weight of each eigenvector; Random Selection The feature vectors are used as the initial cluster centers, the weighted distance from each feature vector to the cluster center is calculated, and each feature vector is assigned to the nearest cluster center. The center point of each cluster center is recalculated and used as the new cluster center until the movement of the cluster center is less than the set threshold α, where α∈[0.001,0.01]; Based on the new cluster center, the multi-source data is clustered to obtain multiple clusters.

3. The method for determining the perforation level of an oil layer according to claim 1, characterized in that: It also includes normalizing multi-source data.

4. A method for determining the perforation level of an oil layer according to claim 1, characterized in that: The multi-source data includes: electrical logging data, acoustic logging data, density logging data and neutron logging data.

5. The method for determining the perforation level of an oil layer according to claim 1, characterized in that: The multi-source data of the perforation layer of the current oil layer and the next oil layer are obtained by using a micro resistivity logging tool, an array sonic logging tool, a gamma density logging tool and a compensated neutron logging tool.

6. A method for determining the perforation level of an oil layer according to claim 1, characterized in that: The K-means clustering algorithm is used to cluster the multi-source data.

7. A method for determining the perforation level of an oil layer according to claim 1, characterized in that: The cluster probability is: Estimate the data distribution parameters of each cluster by maximum likelihood; Calculate the mean vector and covariance matrix of each cluster, and construct a Gaussian probability density function representing the distribution of cluster data; The Gaussian probability density function is used to calculate the cluster probability of each cluster.

8. A system for determining the perforation level of an oil layer, characterized in that: The method comprises a memory and a processor, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for determining the perforation level of an oil layer as claimed in any one of claims 1 to 7 is implemented.

Citation Information

Patent Citations

  • Method and device for determining the position of perforation layer of oil reservoir

    CN106504107A

  • Low-contrast hydrocarbon reservoir identification method, device, equipment and system

    CN111783847A

  • Method and device for determining cluster perforation position of tight oil horizontal well

    CN115110936A

  • Clustering optimal perforation interval method for oil reservoir chemical flooding

    CN116181292A

  • Perforation cluster position optimization design method, processor and machine readable storage medium

    CN117807720A