A method for fusing oilfield core data and well logging data

Abstract

The present invention relates to a method for processing oilfield exploration data, and in particular to a method for fusing oilfield core data and well logging data. The method of the present invention utilizes a density-based DBSCAN algorithm to perform cluster analysis on core data and well logging data and obtain clustering results, find a common well set, and then take each coring well in the common well set as a label class, and use a supervised learning K-nearest neighbor algorithm to subdivide the corresponding well logging class until all well logging data are classified, ensuring that each well logging data can be classified to the closest coring well, thereby realizing the fusion application of two scales and different density data, well logging data and core data. The method of the present invention can be widely used in the research of reservoir engineering and geological modeling, and effectively improve the accuracy and efficiency of geological research.

Description

Technical Field

[0001] The invention relates to an oilfield exploration data processing method, and in particular to an oilfield core data and well logging data fusion method. Background Art

[0002] Oilfield research is an interdisciplinary and multi-field research, involving geophysical exploration, geological research, drilling, logging, well logging and development testing. There are significant differences in data acquisition methods and technologies, resulting in different data representation forms. At the same time, in geological research, different levels of reservoirs, such as the division of geological units at different levels, form data of different scales. Oilfield core data and well logging data are two types of data with different representation forms, scales, spaces and existence densities. Core data is a multiple attribute value representation of the core of the geological object, and the data representation is mostly within the centimeter range in the vertical direction; well logging data is a multiple attribute representation of the well layer of the geological object, and the data representation is within the kilometer range in the vertical direction; the distribution of core and well logging data on the plane of the oilfield is basically distributed at a density ratio of 1:15 according to the number of wells. However, both types of data are the most direct reflection of the reservoir status, and both types of data are the most basic requirements for oilfield fusion applications. The core work of fusion is core homing.

[0003] Core retrieval refers to unifying the drilling depth of the cores taken from drilling to a certain standard depth. The standard depth currently used in scientific research generally refers to the logging depth.

[0004] Chinese patent application CN108227036A discloses a method for tracing the core of fine-grained sedimentary rocks, comprising the following steps: 1) data collection and data preparation; 2) establishing a lithology logging plate using sensitive logging curves; 3) selecting and calibrating typical lithology segments and lithology combination segments using logging lithology, and performing overall preliminary tracing of the core using centimeter-level fine descriptions of the core or comparative analysis of core scanning images; 4) performing local fine tracing of the preliminary tracing core using analytical data and centimeter-level fine descriptions of the core. The purpose of the invention is to provide a method for tracing the core of fine-grained sedimentary rocks based on the multi-parameter combination of "assay data-core description-core image-lithology combination-logging curves", so as to provide a practical technical system for reducing the exploration risk of unconventional oil and gas in fine-grained sedimentary rock formations and improving the exploration success rate of high-yield and high-efficiency wells.

[0005] Chinese patent application CN110134918A discloses a core automatic homing method based on the sliding window method; the method comprises a preliminary depth matching step, a sliding window construction step, a correlation coefficient calculation step and a screening step in sequence, and uses the sliding window method in combination with preliminary depth matching of logging data and experimental data measured by the core, and calculation of the correlation coefficient between the two when the porosity parameter slides along the logging curve to perform intelligent calculation processing for core automatic homing, thereby maximizing the reduction of the systematic error between the cable depth and the drill pipe depth, so as to improve the calculation efficiency and accuracy of core homing.

[0006] At present, the correspondence between lithology and electrical properties is mainly used to unify the cores on the well logging curve. This method is relatively cumbersome, greatly affected by human factors, and relatively poor in accuracy. Later, the ground gamma testing technology of cores was developed, which is a testing technology developed for the reuse and in-depth study of core data. The ground natural gamma ray homing of cores is to conduct ground natural gamma testing on the cores, and compare the measured curve with the well logging natural gamma curve for homing. It is less affected by human factors, has greatly improved accuracy, and is convenient and fast. However, there is a big disadvantage that the number of wells for ground core gamma scanning is very small, with a ratio of 1:80 to coring wells. With such a large ratio, it is not very meaningful to complete the core homing work and realize the fusion of the two data, because the ratio of the fusion of core gamma data and well logging data is 1:1200. Geological research is a global work, which requires obtaining sufficient reservoir physical property data reflected by the core and combining it with well logging data for research.

[0007] Therefore, it is necessary to find a method to generalize core data to all well logging data, so as to achieve a comprehensive integration of reservoir properties reflected by core data and reservoir states reflected by well logging data. Summary of the invention

[0008] The main purpose of the present invention is to provide a new method for fusing oilfield core data and logging data. Make full use of the existing full logging data and coring data of the oilfield, optimize the application of the density-based method DBSCAN algorithm and the supervised learning K-nearest neighbor algorithm, and establish a classification model for the fusion application of logging data and coring data based on the business condition constraints corresponding to the multi-attribute of reservoir depth, so as to form a complete set of coring data obtained from all logged single wells in the oilfield, making it an essential path for obtaining reservoir physical property data for geological research. The method of the present invention studies comprehensive data and can accurately reflect the current status of oil reservoirs and reservoirs.

[0009] To achieve the above object, the present invention adopts the following technical solutions:

[0010] The present invention provides a method for fusing oilfield core data and well logging data, which comprises the following steps:

[0011] 1) Establish a machine learning data training set, collect all the well logging data and core data, and establish a well logging data training set for cluster analysis and a core data training set for cluster analysis respectively;

[0012] 2) Based on the data training set, cluster analysis is performed using the DBSCAN algorithm to obtain a set of well logging data cluster analysis results and a set of core data cluster analysis results;

[0013] 3) determine the common clustering wells;

[0014] 4) Label each well in the common cluster well set;

[0015] 5) Use supervised K-nearest neighbor algorithm for further classification;

[0016] 6) Complete supervised learning classification of all logging data;

[0017] 7) Obtain core data that is closest to reservoir properties.

[0018] Furthermore, in step 1), all logging data are collected, common logging curve types are analyzed, data units are unified, data cleaning is completed, and the data is standardized and normalized to establish a logging data training set for cluster analysis.

[0019] Furthermore, in step 1), the core data includes porosity, permeability, saturation, carbon conventional experimental data, oil-water phase permeability data, wettability data, sensitivity analysis data, and mercury injection data; the data is cleaned and normalized to establish a core data training set for cluster analysis.

[0020] Furthermore, in step 2), based on the well logging data training set, feature engineering analysis is further performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the well logging data cluster analysis result set A = {a1, a2, ... a m}; Based on the core data training set, further feature engineering analysis is performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the core data cluster analysis result set B = {b1, b2, ... b n}.

[0021] Further, in step 3), the logging data a is obtained. i class, automatically find the coring data with common well numbers b j Class, establish a common well set C i ={c1,c2,...c k}, k <n,k<m。

[0022] Further, in step 4), take the well set C i ={c1,c2,...c kEach well in the} is taken as a class, and the label set is established as the label for further subdividing the logging data according to the core data in the next step.

[0023] Furthermore, the coordinate data of each well is used as a constraint condition, and the well logging a i Wells in the class press C i The label set is further classified using the supervised K-nearest neighbor algorithm to complete the logging a i The classification of all wells in the class is continuously iterated and optimized to reach the target value.

[0024] Further, steps 3) to 5) are repeated until the closest coring well is found for each well.

[0025] Compared with the prior art, the present invention has the following advantages:

[0026] (1) The method of the present invention can achieve real-time fusion of two types of data with different scales and distribution densities, namely, well logging data and core data. The method is not limited by test conditions, oil development stages, or traditional empirical formulas, and can achieve real-time fusion of the two types of data, thus realizing the on-demand integration of basic data for oilfield geological research.

[0027] (2) The method of the present invention can form a complete set of physical property data for all wells in the oil field. The method undergoes two classification learnings. The first unsupervised learning obtains a preliminary classification, and the second supervised learning obtains a more accurate classification. Professional constraints are added so that all wells in the oil field obtain the closest physical property parameter values, so that the physical property data of the oil field is expanded from points to surfaces, which can be used as a new data resource for the oil field. This solves the serious problem of missing reservoir physical properties, the most important data in the research of smart oil fields and smart reservoirs.

[0028] (3) The method of the present invention can be widely applied in the research of reservoir engineering and geological modeling, effectively improving the accuracy and efficiency of geological research. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0030] Figure 1 The present invention is a flowchart of a method for fusing oilfield core data and well logging data according to a specific embodiment of the present invention. DETAILED DESCRIPTION

[0031] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0032] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations and / or combinations thereof.

[0033] In order to enable those skilled in the art to more clearly understand the technical solution of the present invention, the technical solution of the present invention will be described in detail below in conjunction with specific embodiments.

[0034] Example 1

[0035] like Figure 1 As shown, the method for fusing oilfield core data and well logging data comprises the following steps:

[0036] Step 110. Establish a machine learning data training set. Collect all the well logging data, analyze the commonly used logging curve types, including SP, R4, RIID / RIIS, CAL, AC, GR, POR, etc., unify the data units, complete data cleaning, and standardize and normalize the data to establish a well logging data training set for cluster analysis; at the same time, analyze the core data, mainly various physical property data of the core, including porosity, permeability, saturation, carbon conventional experimental data, oil-water phase permeability data, wettability data, sensitivity analysis data, mercury injection data, etc., clean and normalize the data, and establish a core data training set for cluster analysis.

[0037] Step 120. Perform cluster analysis using the DBSCAN algorithm. Based on the well logging data training set, further feature engineering analysis is performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the well logging data cluster analysis result set A = {a1, a2, ... a m}; Based on the core data training set, further feature engineering analysis is performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the core data cluster analysis result set B = {b1, b2, ... b n}.

[0038] Step 130. Determine the common clustered wells. An automatic correspondence program needs to be written to obtain well logging data a i class, automatically find the coring data with common well numbers b j Class, establish a common well set C i ={c1,c2,...c k}, k <n,k<m。

[0039] Step 140. Label each well in the common cluster well set. Take well set Ci = {c1, c2, …… c k} Each well is regarded as a class, serving as a label for further subdivision of logging data according to core data in the next step, and a label set is established.

[0040] Step 150. Further classify using the supervised K-nearest neighbor algorithm. Use the coordinate data of each well as a constraint condition. For the wells in logging class a i The wells in class are classified according to C i The label set is further classified using the supervised K-nearest neighbor algorithm to complete the classification of all wells in logging class a i That is, the corresponding specific core wells. Evaluate the algorithm according to algorithm evaluation indicators such as EVS, MAE, MSE, and R2, and continuously perform iterative optimization to reach the target value.

[0041] Step 160. Complete the supervised learning classification of all logging data. Repeat the third step to the fifth step until all logging data is divided into N - T classes (where: 0 <= T < N), that is, find the closest core well for each logging well.

[0042] Step 170. Each well with logging data obtains the core data with the closest reservoir physical properties. The logging borrows the core data of the core well in the same classification to achieve the fusion of two types of data with different scales and different distribution densities.

[0043] In the method described in this embodiment, the fusion process of the two types of data is automated, while the traditional method is basically manual. The general implementation steps of the traditional method are: borrowing the paper materials of core logging; copying the text records of the lifting or lowering values of the core-taking tool from them. If there is no text record, it is also necessary to manually compare the difference between the logging depth and the core-taking depth on the core logging map to obtain the core repositioning values at different depths; then manually reposition the depth of each core sample; finally, correspond the sample repositioning depth with the logging depth to achieve the fusion of the two. Comparison of the analysis results of the method in this embodiment with the analysis results of the traditional method: The model prediction accuracy of the method in this embodiment reaches 92% of the traditional method, but the data fusion application efficiency is increased by 8 - 9 times.

[0044] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications made without departing from the spirit and principle of the present invention shall be equivalent replacement methods and are all included in the protection scope of the present invention.

Claims

1. A method for fusing oilfield core data and well logging data, characterized in that: The following steps are involved: 1) Establish a machine learning data training set, collect all the well logging data and core data, and establish a well logging data training set for cluster analysis and a core data training set for cluster analysis respectively; 2) Based on the data training set, cluster analysis is performed using the DBSCAN algorithm to obtain a set of well logging data cluster analysis results and a set of core data cluster analysis results; 3) determine the common clustering wells; 4) Label each well in the common cluster well set; 5) Use supervised K-nearest neighbor algorithm for further classification; 6) Complete supervised learning classification of all logging data; 7) Obtain core data that is closest to reservoir properties; In step 2), based on the well logging data training set, feature engineering analysis is further performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the well logging data cluster analysis result set A = {a1, a2, ... a m }; Based on the core data training set, further feature engineering analysis is performed, and cluster analysis is performed using the DBSCAN algorithm to obtain the core data cluster analysis result set B = {b1, b2, ... b n }; In step 3), determine the common clustered well set; write an automatic corresponding program to obtain the well logging data a i class, automatically find the coring data with common well numbers b j Class, establish a common well set C i ={c1,c2,...c k }, k <n,k<m; In step 5, a supervised K-nearest neighbor algorithm is used for further classification; the coordinate data of each well is used as a constraint condition, and the well logging a i Wells in the class press C i The label set is further classified using the supervised K-nearest neighbor algorithm to complete the logging a i The classification of all wells in the class, that is, the corresponding specific core wells, is evaluated by the algorithm evaluation indicators such as EVS, MAE, MSE, and R2, and iterative optimization is continuously performed to achieve the target value; In step 6, supervised learning classification is performed on all well logging data; steps 3) to 5) are repeated until the closest coring well is found for each well logging.

2. The method according to claim 1, characterized in that: In step 1), all logging data are collected, common logging curve types are analyzed, data units are unified, data cleaning is completed, and the data is standardized and normalized to establish a logging data training set for cluster analysis.

3. The method according to claim 1, characterized in that: In step 1), the core data includes porosity, permeability, saturation, carbon conventional experimental data, oil-water phase permeability data, wettability data, sensitivity analysis data, and mercury injection data; the data is cleaned and normalized to establish a core data training set for cluster analysis.

4. The method according to claim 1, characterized in that: In step 4), take well set C i ={c1,c2,...c k Each well in the} is taken as a class, and the label set is established as the label for further subdividing the logging data according to the core data in the next step.

Images