Method, device and medium for dividing CO2 drive dynamic flow units in low-permeability reservoirs

Abstract

The present application provides a method, device and medium for dividing CO2 drive dynamic flow units in low-permeability reservoirs. The method includes: determining an object to be analyzed in a target layer of a reservoir to be divided; performing normalization processing on the object to be analyzed to obtain a normalized object; performing first clustering processing on the normalized objects according to the numerical similarity between the normalized objects to determine an initial target normalized object among the normalized objects and an initial target object to be analyzed corresponding to the initial target normalized object among the objects to be analyzed; performing screening processing on the initial target object to be analyzed to determine a target object to be analyzed; determining a comprehensive object of a single well in the reservoir to be divided based on the target object to be analyzed; performing second clustering processing on the comprehensive object of the single well to obtain a clustering result, and determining flow units in the reservoir to be divided and interval boundaries of the flow units based on the average daily oil production of the single well and the clustering result. This method improves the accuracy of reservoir flow unit division.

Description

Technical Field

[0001] The present application relates to the technical field of oil and gas field exploration, and particularly to a method, device and medium for dividing CO2-driven dynamic flow units in low-permeability reservoirs. Background Art

[0002] In low-permeability reservoirs, due to their low reservoir permeability and complex pore structures, it is difficult for fluids to flow in the reservoirs. Traditional development methods are difficult to fully exploit their potential, with low water flooding displacement efficiency and difficult to achieve economic benefits. Therefore, CO2 flooding is usually used as the displacement medium to reduce the viscosity in low-permeability reservoirs and improve the oil production efficiency of the reservoirs. However, when injecting gas for development in low-permeability reservoirs based on CO2 as the displacement medium, as the development time goes by, the seepage characteristics of crude oil in the reservoir will change accordingly. In the practice of reservoir development, flow units constitute the most basic components. They refer to those reservoir units that show consistency in rock characteristics and physical properties. Each flow unit maps a specific sedimentary background and its corresponding fluid permeability attributes. Based on this, accurate identification and division of flow units in the reservoir can provide crucial guidance and basis for the efficient development of the reservoir.

[0003] In the prior art, the methods for dividing flow units are mainly divided into two types: one is the reservoir parameter analysis method mainly based on mathematical means; the other is the reservoir hierarchical analysis method mainly based on geological research. Among them, the FlowZone Indicator (FZI) method is the most popular method for dividing flow units at present. This method establishes the relationship between FZI and reservoir quality index and porosity index based on the Kozeny-Carman equation. First, the reservoir quality index and porosity index are calculated respectively using the porosity and permeability of core physical property experimental data, and then the flow units are divided through the double logarithmic coordinate cross plot of the reservoir quality index and porosity index.

[0004] However, the existing methods have good division effects for pore-type reservoirs, but there are problems with poor division effects of flow units for other reservoirs with changing formation characteristics during the development process. Summary of the Invention

[0005] The embodiments of the present application provide a method, device and medium for dividing CO2-driven dynamic flow units in low-permeability reservoirs to solve the problem of poor division effects of flow units existing in the existing methods and improve the accuracy of flow unit division.

[0006] In a first aspect, the embodiments of the present application provide a method for dividing CO2-driven dynamic flow units in low-permeability reservoirs, including:

[0007] Determine the objects to be analyzed for the target layer in the reservoir to be divided. The objects to be analyzed include at least three of reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation factor, storage coefficient, inter-well flow capacity index, and reservoir quality factor;

[0008] Perform normalization processing on the objects to be analyzed for the target layer in the reservoir to be divided to obtain normalized objects;

[0009] According to the numerical similarity between the normalized objects, perform the first clustering process on the normalized objects to determine the initial target normalized objects in the normalized objects and the corresponding initial target objects to be analyzed in the objects to be analyzed;

[0010] Perform screening processing on the initial target objects to be analyzed to determine the target objects to be analyzed. The target objects to be analyzed meet the requirements of the preset analysis object type and the importance degree of the analysis object;

[0011] Based on the target objects to be analyzed for the target layer in the reservoir to be divided, determine the comprehensive objects of a single well in the reservoir to be divided. A single well represents a test well in the reservoir to be divided, and the test well includes multiple target layers;

[0012] Perform the second clustering process on the comprehensive objects of the single well to obtain a clustering result, and based on the average daily oil production of the single well and the clustering result, determine the flow units in the reservoir to be divided and the interval boundaries of the flow units. The clustering result includes the clustering center and the clustering center value corresponding to the clustering center.

[0013] In a possible implementation manner, performing normalization processing on the objects to be analyzed for the target layer in the reservoir to be divided to obtain normalized objects includes:

[0014] Determine the numerical values of the objects to be analyzed for the target layer in the reservoir to be divided;

[0015] Analyze the numerical values of the objects to be analyzed to determine the numerical distribution characteristics of the objects to be analyzed;

[0016] According to the numerical distribution characteristics, perform normalization processing on the objects to be analyzed to obtain normalized objects.

[0017] In a possible implementation manner, according to the numerical similarity between the normalized objects, performing the first clustering process on the normalized objects to determine the initial target normalized objects in the normalized objects and the corresponding initial target objects to be analyzed in the objects to be analyzed includes:

[0018] According to the numerical similarity between the normalized objects, perform clustering recombination processing on the normalized objects to obtain the normalized objects after clustering recombination processing;

[0019] Take the normalized object after clustering and recombination processing as the normalized object, and repeat the step of performing clustering and recombination processing on the normalized object according to the numerical similarity between the normalized objects until the numerical similarity between the normalized objects after clustering and recombination processing meets the requirements of a preset numerical similarity threshold, and then determine the initial target normalized object in the normalized object and the initial target object to be analyzed corresponding to the initial target normalized object in the object to be analyzed.

[0020] In a possible implementation manner, performing clustering and recombination processing on the normalized object according to the numerical similarity between the normalized objects to obtain the normalized object after clustering and recombination processing includes:

[0021] Determine the numerical dimension of the normalized object, where the data dimension is determined according to the number of target layers in the reservoir to be divided;

[0022] According to the numerical dimension, determine the Euclidean distance between the normalized objects;

[0023] According to the Euclidean distance, determine the numerical similarity between the normalized objects;

[0024] Perform clustering and recombination processing on the normalized object according to the numerical similarity between the normalized objects to obtain the normalized object after clustering and recombination processing.

[0025] In a possible implementation manner, performing clustering and recombination processing on the normalized object according to the numerical similarity between the normalized objects to obtain the normalized object after clustering and recombination processing includes:

[0026] According to the numerical similarity between the normalized objects, determine the first normalized object, the second normalized object, and other normalized objects, where the numerical similarity between the first normalized object and the second normalized object meets the requirements of a preset numerical similarity recombination;

[0027] Obtain the recombined normalized object according to the first normalized object and the second normalized object;

[0028] Obtain the normalized object after clustering and recombination processing according to the recombined normalized object and other normalized objects.

[0029] In a possible implementation manner, performing screening processing on the initial target object to be analyzed to determine the target object to be analyzed, where the target object to be analyzed meets the requirements of a preset analysis object type and the importance degree requirement of the analysis object, includes:

[0030] According to the requirements of the preset type of analysis object, determine the combined objects in the initial target objects to be analyzed. The combined objects are analysis objects obtained by performing combined calculation processing on two or more single objects. The single objects are analysis objects obtained by detecting test data on the target layer in the reservoir to be divided;

[0031] According to the influence degree of the single object on the flow unit in the reservoir to be divided, screen the combined objects to determine the target objects to be analyzed.

[0032] In a possible implementation manner, based on the average daily oil production per well and the clustering result, determine the flow units in the reservoir to be divided and the interval boundaries of the flow units, including:

[0033] Obtain the average daily oil production per well of the single well in the reservoir to be divided;

[0034] According to the cluster centers in the clustering result and the corresponding cluster center values, divide the single wells in the reservoir to be divided to determine the initial flow units in the reservoir to be divided;

[0035] Based on the average daily oil production per well, adjust the initial flow units to obtain the flow units in the reservoir to be divided and the interval boundaries of the flow units. The interval boundaries represent the interval boundaries corresponding to the comprehensive object of the single well and the average daily oil production per well in the flow unit.

[0036] In a second aspect, an apparatus for dividing the CO2 drive dynamic flow units in a low-permeability oil reservoir provided by an embodiment of the present application includes:

[0037] A first determination module, configured to determine the objects to be analyzed of the target layer in the reservoir to be divided. The objects to be analyzed include at least three of reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation factor, storage coefficient, inter-well flow capacity index, and reservoir quality factor;

[0038] A preprocessing module, configured to perform normalization processing on the objects to be analyzed of the target layer in the reservoir to be divided to obtain normalized objects;

[0039] A first clustering module, configured to perform first clustering processing on the normalized objects according to the numerical similarity between the normalized objects to determine the initial target normalized objects in the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects in the objects to be analyzed;

[0040] A screening module, configured to perform screening processing on the initial target objects to be analyzed to determine the target objects to be analyzed, and the target objects to be analyzed meet the requirements of the preset type of analysis object and the importance degree of the analysis object;

[0041] A second determination module, configured to determine a comprehensive object of a single well in the reservoir to be divided based on a target object to be analyzed of a target layer in the reservoir to be divided, where the single well represents a test well in the reservoir to be divided, and the test well includes multiple target layers;

[0042] A second clustering module, configured to perform second clustering processing on the comprehensive object of the single well to obtain a clustering result, and determine flow units in the reservoir to be divided and interval boundaries of the flow units based on the average daily oil production of the single well and the clustering result, where the clustering result includes a clustering center and a clustering center value corresponding to the clustering center.

[0043] In a third aspect, an embodiment of the present application provides an electronic device, including: a memory and a processor;

[0044] The memory stores computer-executable instructions;

[0045] The processor executes the computer-executable instructions stored in the memory to implement the first aspect and / or various possible implementation manners of the first aspect as described above.

[0046] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed, they are used to implement the first aspect and / or various possible implementation manners of the first aspect as described above.

[0047] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, and when the computer program is executed, it implements the first aspect and / or various possible implementation manners of the first aspect as described above.

[0048] The method, device and medium for dividing CO2 drive dynamic flow units in low-permeability reservoirs provided by the embodiments of the present application take the factors affecting the flow characteristics of reservoirs as objects to be analyzed, such as reservoir depth, porosity, permeability, etc. These objects to be analyzed jointly determine the physical characteristics and flow capacity of the reservoir; perform normalization processing on these objects to be analyzed to eliminate the influence of the dimension and numerical range between objects; through the first clustering process, screen out the target objects to be analyzed that have an important impact on the division of flow units from the objects to be analyzed. Then, based on these target objects to be analyzed, determine the comprehensive objects of each single well in the reservoir to be divided, and perform the second clustering process on these comprehensive objects to obtain the clustering result; further, based on the average daily oil production of a single well and the clustering result, determine the flow units in the reservoir to be divided and the interval boundaries of the flow units, which can identify the similarity of each object to be analyzed of the internal flow characteristics of the reservoir, so as to select the objects to be analyzed that meet the preset requirements to divide the reservoir flow units, making the selected target objects to be analyzed more accurately reflect the complexity of the reservoir, thereby improving the accuracy of the division of reservoir flow units, making the resource allocation for different types of flow units more reasonable. At the same time, it can continuously dynamically update and optimize the division result of the flow unit according to the updated data of the object to be analyzed and the average daily oil production of a single well, providing a strong scientific basis for reservoir management and development. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings herein are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0050] Figure 1 It is a schematic flow chart of a method for dividing CO2 drive dynamic flow units in low-permeability reservoirs provided by the present application;

[0051] Figure 2 It is a schematic flow chart of another method for dividing CO2 drive dynamic flow units in low-permeability reservoirs provided by the present application;

[0052] Figure 3 It is a schematic diagram of the result of the first clustering process among the normalized objects provided by the present application;

[0053] Figure 4 It is a correlation trend chart between the comprehensive formation coefficient of a single well and the average daily oil production of a single well provided by the present application;

[0054] Figure 5 It is a correlation trend chart between the comprehensive storage coefficient of a single well and the average daily oil production of a single well provided by the present application;

[0055] Figure 6 It is a schematic diagram of the clustering result corresponding to the division of three types of flow units provided by the present application;

[0056] Figure 7 Schematic diagram of flow unit division for different types of single wells provided in this application under the condition of dynamic change of daily oil production

[0057] Figure 8 Schematic diagram of distribution characteristics of three types of flow units in the reservoir provided in this application

[0058] Figure 9 Schematic diagram of the structure of the CO2 drive dynamic flow unit division device for low - permeability oil reservoirs provided in this application

[0059] Figure 10 Schematic diagram of the structure of the electronic device provided in this application

[0060] Through the above - mentioned drawings, specific embodiments of this application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments Detailed implementation manners

[0061] Here, the exemplary embodiments will be described in detail, and the examples are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementation manners described in the following exemplary embodiments do not represent all implementation manners consistent with this application. On the contrary, they are merely examples of devices and methods consistent with some aspects of this application as detailed in the appended claims

[0062] In the prior art, the Flow Zone Indicator (FZI) method is one of the most popular methods for dividing flow units. This method establishes the relationship between FZI, reservoir quality index, and porosity index based on the Kozeny-Carman equation. First, the reservoir quality index and porosity index are calculated using the porosity and permeability from core physical property experiment data, and then the flow units are divided through a double logarithmic coordinate cross-plot of the reservoir quality index and porosity index. Although this method has a good effect on dividing pore-type reservoirs, FZI depends on the porosity and permeability data of physical property experiment data. Since when the CO2 flooding technology is used for developing low-permeability reservoirs, the factors affecting the reservoir flow characteristics change significantly compared with water flooding development, obviously this method has problems of insufficient factor analysis and inaccurate division in this case. Moreover, the reservoir will undergo dynamic changes during the development process, such as changes in physical properties, etc. The FZI method is mainly based on static data and may not be able to accurately reflect the impact of these dynamic changes on the division of flow units. Therefore, this method has a problem of poor flow unit division effect for reservoirs with changing formation physical properties and parameters during the development process, especially for reservoirs after CO2 injection flooding in low-permeability reservoirs.

[0063] To solve the above problems, the embodiments of the present application provide a method, device, and medium for dividing dynamic flow units in a low-permeability reservoir by CO2 flooding. Factors affecting the reservoir flow characteristics can be used as objects to be analyzed, such as reservoir depth, porosity, permeability, etc. These objects to be analyzed jointly determine the physical properties and flow capacity of the low-permeability reservoir. Among them, since the dimensions and numerical ranges of the objects to be analyzed may be different, normalization processing is adopted to ensure the comparability of each object in cluster analysis. Then, based on the normalized objects to be analyzed, through the first clustering process, the similarity between the objects to be analyzed is determined, and the target objects to be analyzed that have an important impact on the division of flow units are selected from them, so as to achieve the purpose of improving the division accuracy while reducing the division error caused by similar objects to be analyzed. Then, based on these selected target objects to be analyzed, the comprehensive objects of each single well in the reservoir to be divided are determined, and the second clustering process is performed on these comprehensive objects to obtain the clustering result. Then, based on the average daily oil production per well and the clustering result, the flow units of the reservoir are divided to determine the flow units in the reservoir to be divided and the interval boundaries of each flow unit. Among them, the selected target objects to be analyzed have a small similarity, can more accurately reflect the complexity of the reservoir, and thus can more accurately divide the flow units of the reservoir. At the same time, by combining the average daily oil production per well and the comprehensive objects of the single well, based on multi-dimensional comprehensive evaluation, the division effect of the flow units in the reservoir to be divided is better.

[0064] The following uses specific embodiments to elaborate in detail on the technical solution of the present application and how the technical solution of the present application solves the above technical problems. The following several specific embodiments can be combined with each other, and the same or similar concepts or processes may not be elaborated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0065] The execution subject of the method for dividing CO2 drive dynamic flow units in low-permeability reservoirs provided by the embodiments of the present application can be a server. Among them, the server can be a device such as a mobile phone, a computer, or a tablet. The embodiments of the present application do not make special restrictions on the implementation manner of the execution subject, as long as the execution subject can determine the objects to be analyzed in the target layer of the reservoir to be divided, and the objects to be analyzed include at least three of reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation factor, storage coefficient, inter-well flow capacity index, and reservoir quality factor; perform normalization processing on the objects to be analyzed in the target layer of the reservoir to be divided to obtain normalized objects; perform the first clustering process on the normalized objects according to the numerical similarity between the normalized objects to determine the initial target normalized objects among the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects among the objects to be analyzed; perform screening processing on the initial target objects to be analyzed to determine the target objects to be analyzed, and the target objects to be analyzed meet the requirements of the preset analysis object type and the requirements of the importance degree of the analysis object; based on the target objects to be analyzed in the target layer of the reservoir to be divided, determine the comprehensive objects of a single well in the reservoir to be divided, and a single well represents a test well in the reservoir to be divided, and multiple target layers are included in the test well; perform the second clustering process on the comprehensive objects of the single well to obtain a clustering result, and based on the average daily oil production of the single well and the clustering result, determine the flow units in the reservoir to be divided and the interval boundaries of the flow units, and the clustering result includes the clustering center and the clustering center value corresponding to the clustering center.

[0066] Among them, CO2 flooding can refer to carbon dioxide (CO2) substances injected underground during the reservoir development process to displace crude oil and push it towards the production well.

[0067] Low-permeability reservoir formations can refer to those oil and gas reservoir formations with relatively low permeability. Such reservoirs are usually difficult to effectively develop using traditional extraction methods. Its characteristics are low porosity and low permeability, which result in poor fluid flow ability in the reservoir and greater difficulty in oil and gas extraction. In the present application, a reservoir or a reservoir formation can refer to a geological reservoir that can store and transport oil and gas.

[0068] Figure 1The flowchart shows a method for dividing CO2 drive dynamic flow units in low-permeability reservoirs provided in this application. The execution entity of this method can be a server storing the method for dividing CO2 drive dynamic flow units in low-permeability reservoirs or other servers, and there is no special limitation here in this embodiment. For example, Figure 1 as shown, this method may include:

[0069] S101. Determine the objects to be analyzed in the target layer of the reservoir to be divided. The objects to be analyzed include at least three of reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation coefficient, storage coefficient, inter-well flow capacity index, and reservoir quality factor.

[0070] Among them, the target layer may refer to each formation test data sampling area in the reservoir. For example, in the entire reservoir area, there are 100 test wells, and each test well includes 10 data sampling layers, then the target layer includes 1000 sampling areas.

[0071] In some embodiments, the method for determining the objects to be analyzed in the target layer of the reservoir to be divided may include: obtaining the actual reservoir data of the target layer in the reservoir to be divided; analyzing the actual reservoir data of the target layer according to the reservoir storage characteristics of the reservoir to be divided, and determining the objects to be analyzed related to the division of flow units in the target layer.

[0072] S102. Perform normalization processing on the objects to be analyzed in the target layer of the reservoir to be divided to obtain normalized objects.

[0073] Among them, the purpose of performing normalization processing on each object to be analyzed is to eliminate the differences in dimension and numerical range between different objects to be analyzed, and ensure that each object to be analyzed has the same importance in the clustering process.

[0074] The normalization method can be methods such as min-max standardization, Z-score standardization, normalization to unit length, etc. There is no special limitation on the selection of the normalization method here. For example, the values of all objects to be analyzed can be regarded as points in a multi-dimensional space, and then these points can be connected by edges, and the similarity between each object to be analyzed can be determined by comparing the distances of each edge.

[0075] Among them, in the embodiment of this application, the method for performing normalization processing on the objects to be analyzed in the target layer of the reservoir to be divided to obtain normalized objects may include:

[0076] Determine the numerical values of the objects to be analyzed in the target layer of the reservoir to be divided;

[0077] Analyze the numerical values of the objects to be analyzed to determine the numerical distribution characteristics of the objects to be analyzed;

[0078] Normalize the object to be analyzed according to the numerical distribution characteristics to obtain a normalized object.

[0079] Among them, the numerical distribution characteristics can characterize the differences and distribution among the values of the objects to be analyzed in each target layer. According to the numerical distribution characteristics, the influence of abnormal data in the values on the subsequent clustering analysis can be reduced, and the error of data processing can be reduced.

[0080] S103. Perform a first clustering process on the normalized objects according to the numerical similarity between the normalized objects, and determine the initial target normalized objects in the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects in the objects to be analyzed.

[0081] Among them, the numerical similarity can refer to the degree of similarity between the objects to be analyzed, and it can be used to characterize that the objects to be analyzed have similar geological meanings or formation conditions.

[0082] The first clustering process can be a process based on the Euclidean distance clustering algorithm. The Euclidean distance is the straight-line distance between two points in a multi-dimensional space, and it can be used to measure the similarity or difference between data points.

[0083] Among them, in the embodiments of the present application, the method for performing a first clustering process on the normalized objects according to the numerical similarity between the normalized objects and determining the initial target normalized objects in the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects in the objects to be analyzed may include:

[0084] Perform a clustering recombination process on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after the clustering recombination process;

[0085] Use the normalized objects after the clustering recombination process as the normalized objects, and repeat the step of performing a clustering recombination process on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after the clustering recombination process until the numerical similarity between the normalized objects after the clustering recombination process meets the requirements of a preset numerical similarity threshold, and then determine the initial target normalized objects in the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects in the objects to be analyzed.

[0086] Among them, the preset numerical similarity threshold requirement can be a value set in advance or a value range. In some embodiments, the method for determining the initial target normalized object in the normalized object and the initial target object to be analyzed corresponding to the initial target normalized object in the object to be analyzed may further include: using the normalized object after clustering and recombination processing as the normalized object, and repeatedly performing the step of clustering and recombination processing on the normalized object according to the numerical similarity between the normalized objects until all the normalized objects after clustering and recombination processing are grouped into one group to obtain a clustering and recombination pedigree diagram, where the clustering pedigree diagram represents the magnitude of the numerical similarity between the normalized objects in different groups obtained after each clustering and recombination processing and the normalized objects corresponding to each group during the recombination process; analyzing the clustering and recombination pedigree diagram to determine the initial target normalized object in the normalized object and the initial target object to be analyzed corresponding to the initial target normalized object in the object to be analyzed.

[0087] For example, for the normalized objects A, B, C, D, E, and F, the clustering and recombination pedigree diagram may include the normalized objects after the first clustering and recombination process: AC, B, D, E, F, and the corresponding numerical similarity of the clustering and recombination is 1; the normalized objects after the second clustering and recombination process: AC, B, D, EF, and the corresponding numerical similarity of the clustering and recombination is 4; the normalized objects after the third clustering and recombination process: ACB, D, EF, and the corresponding numerical similarity of the clustering and recombination is 9; the normalized objects after the fourth clustering and recombination process: ACBD, EF, and the corresponding numerical similarity of the clustering and recombination is 18; the normalized objects after the fifth clustering and recombination process: ACBDEF, and the corresponding numerical similarity of the clustering and recombination is 20. Among them, after the first clustering and recombination process, five groups of normalized objects can be obtained, and one group includes two normalized objects A and C. Similarly, after the third clustering and recombination process, three groups of normalized objects can be obtained, one group includes three normalized objects A, C, and B, another group includes one normalized object D, and another group includes two normalized objects E and F. Among them, in some embodiments, by analyzing the span of the numerical similarity intervals of the normalized objects after different numbers of clustering and recombination processes, the recombined normalized objects with a larger span of the numerical similarity intervals in the clustering and recombination process can be selected as the initial target normalized objects to ensure that the similarity between the objects to be analyzed in each group is low, while the similarity between the objects to be analyzed in the group is high, and to avoid the situation of high similarity when screening the initial target normalized objects, thereby improving the accuracy of the subsequent flow unit division. For example, the span of the numerical similarity interval corresponding to the first clustering and recombination and the second clustering and recombination is 4 - 1 = 3; the span of the numerical similarity interval corresponding to the second clustering and recombination and the third clustering and recombination is 9 - 4 = 5; the span of the numerical similarity interval corresponding to the third clustering and recombination and the fourth clustering and recombination is 18 - 9 = 9; the span of the numerical similarity interval corresponding to the fourth clustering and recombination and the fifth clustering and recombination is 20 - 18 = 2. The span of the numerical similarity interval corresponding to the third clustering and recombination and the fourth clustering and recombination is significantly larger than others, indicating that there is a large difference and low similarity between the normalized objects in each group after the third clustering and recombination. Therefore, the normalized objects after the third clustering and recombination process: ACB, D, EF are selected as the initial target normalized objects.

[0088] Among them, in the embodiments of the present application, a method for clustering and recombining normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after the clustering and recombination process may include:

[0089] Determine the numerical dimension of the normalized objects, and the data dimension is determined according to the number of target layers in the reservoir to be divided;

[0090] According to the numerical dimension, determine the Euclidean distance between the normalized objects;

[0091] Determine the numerical similarity between the normalized objects according to the Euclidean distance;

[0092] Perform clustering and recombination processing on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after the clustering and recombination processing.

[0093] Among them, the Euclidean distance satisfies:

[0094]

[0095] Among them, d ( x,y ) is the Euclidean distance; x is the value of an object to be analyzed; y is the value of another object to be analyzed; i is the numerical dimension, i = 1, 2, 3,..., n.

[0096] Among them, the numerical similarity can be the Euclidean distance or the index data for judging similarity obtained according to the Euclidean distance.

[0097] Among them, in the embodiments of the present application, the method for performing clustering and recombination processing on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after the clustering and recombination processing may include:

[0098] Determine a first normalized object, a second normalized object, and other normalized objects according to the numerical similarity between the normalized objects, and the numerical similarity between the first normalized object and the second normalized object meets the preset numerical similarity recombination requirement;

[0099] Obtain a recombined normalized object according to the first normalized object and the second normalized object;

[0100] Obtain the normalized objects after the clustering and recombination processing according to the recombined normalized object and the other normalized objects.

[0101] Among them, the preset numerical similarity recombination requirement may refer to recombining the object to be analyzed with the smallest numerical similarity. For example, given the normalized objects A, B, C, D, E, and F, according to the numerical similarity between them, it can be known that the numerical similarity between A and C is the smallest, which is 1. Then A and C are grouped into a set to recombine the normalized object AC. After the clustering recombination process, the normalized objects are AC, B, D, E, and F. Then, determine the numerical similarity between the normalized objects after the clustering recombination process. It can be known that the numerical similarity between E and F is the smallest, which is 4. Then E and F are grouped into a set to recombine the normalized object EF. After the clustering recombination process, the normalized objects are AC, B, D, and EF. Determine the numerical similarity again. It can be known that the numerical similarity between AC and B is the smallest, which is 9. Then AC and B are grouped into a set to recombine the normalized object ACB. After the clustering recombination process, the normalized objects are ACB, D, and EF. Determine the numerical similarity again. It can be known that the numerical similarity between ACB and D is the smallest, which is 18. Then ACB and D are grouped into a set to recombine the normalized object ACBD. After the clustering recombination process, the normalized objects are ACBD and EF. Determine the numerical similarity again. It can be known that the numerical similarity between ACBD and EF is the smallest, which is 20. Then ACBD and EF are grouped into a set to recombine the normalized object ACBDEF.

[0102] Among them, the recombined normalized object may include multiple normalized objects with the same recombination label. The similarity degree between the normalized objects with the same recombination label is higher than the similarity degree between this normalized object and the normalized objects corresponding to other recombination labels. For example, the similarity degree between A and C is higher than the similarity degree between A and B or A and D.

[0103] In some embodiments, as Figure 3 shown, the normalized objects include reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation coefficient, storage coefficient, inter-well flow capacity index, and reservoir quality factor. After the clustering recombination process, they are divided into three categories, that is, the inter-well flow capacity index and the formation coefficient have the same first-type recombination label; the reservoir quality factor, pore throat radius, permeability, and storage coefficient have the same second-type recombination label; the reservoir depth, porosity, oil saturation, and effective thickness have the same third-type recombination label.

[0104] S104. Perform a screening process on the initial target object to be analyzed to determine the target object to be analyzed, and the target object to be analyzed meets the preset requirements for the type of analysis object and the importance level of the analysis object.

[0105] Among them, the preset requirements for the type of analysis object can refer to selecting different types of objects to be analyzed to improve the accuracy of flow unit division. For example, the type of analysis object can include single-type objects and combined-type objects. The combined-type objects can effectively reduce the data errors caused by the single-type objects being too large or too small. Therefore, when screening, the requirements for the type of analysis object can be set to only screen combined-type objects. For example, in the second type of recombined label, the storage coefficient or reservoir quality factor is considered for selection; in the third type of recombined label, only single-type objects are included, so the target object to be analyzed can be not selected in the third group.

[0106] The requirements for the importance degree of the analysis object can refer to determining the influence degree of each object to be analyzed on the flow unit division according to the change of the physical property characteristics of the reservoir to be divided, and selecting the object to be analyzed with a high influence degree is beneficial to improving the accuracy of flow unit division. For example, for a reservoir with CO2 flooding development, after a long period of development, the long-term gas flooding will cause the permeability of different regions in the reservoir to change, and the lower limit of the effective thickness of the reservoir will also change. Therefore, when one combined-type object contains the effective thickness of the reservoir and the other combined-type object does not contain the effective thickness of the reservoir, because the change of the effective thickness of the reservoir has a greater impact on the division of the flow unit, the combined-type object containing the effective thickness of the reservoir is selected as the target object to be analyzed.

[0107] Among them, in the embodiments of the present application, the method for screening the initial target object to be analyzed to determine the target object to be analyzed, where the target object to be analyzed meets the preset requirements for the type of analysis object and the importance degree of the analysis object can include:

[0108] According to the preset requirements for the type of analysis object, determine the combined-type objects among the initial target objects to be analyzed. The combined-type object is an analysis object obtained by performing combined calculation processing on two or more single-type objects. The single-type object is an analysis object obtained by detecting test data on the target layer in the reservoir to be divided;

[0109] According to the influence degree of the single-type object on the flow unit in the reservoir to be divided, perform screening processing on the combined-type object to determine the target object to be analyzed.

[0110] In some embodiments, the single-type objects in the low-permeability reservoir with CO2 flooding development include pore throat radius, permeability, reservoir depth, porosity, oil saturation, and effective thickness. Among them, the effective thickness can be the available effective thickness of the CO2 flooding reservoir, that is, during the CO2 flooding process, the reservoir thickness that can be effectively utilized and mobilized, reflecting the formation thickness where the crude oil that can be effectively displaced and recovered by CO2 in the reservoir is located; the combined-type objects include the inter-well flow capacity index, formation coefficient, reservoir quality factor, and storage coefficient. Among them:

[0111] The inter-well flow capacity index satisfies:

[0112] IFCI = (K1 × h1) / (K2 × h2);

[0113] Where, IFCI is the inter-well flow capacity index; K1 is the permeability of the low-permeability layer; K2 is the permeability of the high-permeability layer; h1 is the effective thickness of the low-permeability layer; h2 is the effective thickness of the high-permeability layer.

[0114] The formation factor satisfies:

[0115] C1 = kh;

[0116] Where, C1 is the formation factor; k is the permeability; h is the effective thickness.

[0117] The reservoir quality factor satisfies:

[0118]

[0119] Where, RQI is the reservoir quality factor; φ is the porosity.

[0120] The storage coefficient satisfies:

[0121] C2 = φhs o ;

[0122] Where, C2 is the storage coefficient; s o is the oil saturation.

[0123] For the reservoir of CO2 flooding development, after a long period of development, the lower limit of the effective thickness of the reservoir will also change. Therefore, when the formation factor and the inter-well flow capacity index have the same first type of recombination label, both the formation factor and the inter-well flow capacity index are combined objects. Analyzing the influence degree of dividing the flow units by treating them as single objects, both of them include the effective thickness in the single object, but the inter-well flow capacity index is the ratio of the effective thickness between the low-permeability layer and the high-permeability layer, which cannot reflect the flow capacity of each layer. And the division of flow units needs to be refined to each layer. Therefore, the formation factor that can evaluate the single-layer permeability and effective thickness is selected as the target object to be analyzed, making the division more accurate.

[0124] In the second type of recombination label, the storage coefficient comprehensively considers the changes of oil saturation and porosity, and the reservoir quality factor characterizes the quality of the seepage of the reservoir, that is, the change relationship between the reservoir permeability and the reservoir porosity. In the actual environment, due to the serious reservoir heterogeneity, the data of the high-permeability layer is nearly a thousand times higher than that of the low-permeability layer, resulting in a large difference in permeability between the reservoirs with basically the same porosity. Therefore, it is more widely applicable to select the storage coefficient to classify the flow units.

[0125] S105. Determine the comprehensive object of a single well in the reservoir to be divided based on the target object to be analyzed in the target layer of the reservoir to be divided. A single well represents a test well in the reservoir to be divided, and the test well includes multiple target layers.

[0126] In this step, calculate the target object to be analyzed according to a preset formula, and then perform weighted summation or other mathematical operations on the calculated target object to be analyzed to obtain the corresponding comprehensive object of the single well. Among them, the comprehensive object of the single well corresponds to each target object to be analyzed.

[0127] In some embodiments, when the selected target objects to be analyzed are formation coefficient and storage coefficient, the comprehensive object of the single well is the comprehensive formation coefficient of the single well and the comprehensive storage coefficient of the single well, or the comprehensive formation coefficient of the single well and the comprehensive storage coefficient of the single well after further normalization. Since the variability between the formation coefficients or storage coefficients of different single wells may be very large. For example, the coefficient values of some single wells may have a large span, while others are relatively stable. Normalization processing can reduce this heteroscedasticity and make the coefficients of each single well have the same importance in the calculation. Among them:

[0128] The comprehensive formation coefficient of the single well satisfies:

[0129]

[0130] where C 1total is the comprehensive formation coefficient of the single well; P1 is the normalized comprehensive formation coefficient of the single well; C 1i is the formation coefficient of the i-th target layer in the single well; C 1j is the comprehensive formation coefficient of the j-th single well; n is the number of target layers in the single well; m is the number of single wells.

[0131] The comprehensive storage coefficient of the single well satisfies:

[0132]

[0133] where C 2total is the comprehensive storage coefficient of the single well; P2 is the normalized comprehensive storage coefficient of the single well; C 2i is the storage coefficient of the i-th target layer in the single well; C 2j is the comprehensive storage coefficient of the j-th single well.

[0134] S106. Perform second clustering processing on the comprehensive object of the single well to obtain a clustering result, and determine the flow units in the reservoir to be divided and the interval boundaries of the flow units based on the average daily oil production of the single well and the clustering result. The clustering result includes the cluster center and the cluster center value corresponding to the cluster center.

[0135] In this step, the second clustering process can be performed based on the K-Means clustering method. The number of cluster centers can represent the number of flow units divided, and the cluster center values can be used to determine whether the flow units divided by the clustering results corresponding to the cluster centers are reasonable. If the span between two cluster center values is large, it indicates that there is a division error in the clustering results corresponding to the cluster center, and the number of cluster centers needs to be further subdivided.

[0136] In some embodiments, the method for performing the second clustering process on the comprehensive object of a single well to obtain the clustering results may include:

[0137] Performing the second clustering process on the comprehensive object of a single well to obtain initial clustering results, where there are more than two initial clustering results and the number of cluster centers in different initial clustering results is different;

[0138] According to the preset requirements for the span interval of cluster center values, screening the cluster center values in the initial clustering results to determine the clustering results that meet the requirements for the span interval of cluster center values in the initial clustering results.

[0139] Among them, the preset requirements for the span interval of cluster center values can be that each cluster center corresponding to the comprehensive object of a single well needs to meet the requirement of uniform interval span, and cannot differ too little or too much. For example, when the number of cluster centers is 2, the formation factor of the first cluster center is 35.8, and the formation factor of the second cluster center is 1.9. The span of these two cluster centers is large, indicating that it does not meet the span interval requirements; another example is when the number of cluster centers is 3, the formation factor of the first cluster center is 35.1, the formation factor of the second cluster center is 30.6, and the formation factor of the third cluster center is 5.3. The span between the first and second cluster centers is small, while the span between the third and second cluster centers is large, indicating that it does not meet the span interval requirements.

[0140] Exemplarily, when the selected comprehensive object of a single well is the comprehensive formation factor of a single well and the comprehensive reservoir coefficient of a single well, based on the K-Means clustering method, the number of cluster centers K is designed as 2, 3, and 4 respectively, and the above three initial clustering results are obtained by using the K-means clustering algorithm of SPSS software. Based on data analysis, it is determined that when K = 2, the span of the comprehensive formation factor of a single well and the comprehensive reservoir coefficient of a single well between the cluster centers of the initial clustering results is large, and the seepage characteristics of the reservoir cannot be finely characterized; when K = 4, the comprehensive formation factor interval of a single well between the third and fourth cluster centers is nearly coincident, and the difference in seepage between various flow units cannot be reflected; when K = 3, the span interval between the cluster centers is uniform, and the seepage capacity and reservoir capacity between the flow units of the reservoir can be better reflected. Therefore, the clustering results are determined as the number of cluster centers and the cluster center values when K = 3.

[0141] Among them, in the embodiments of the present application, based on the average daily oil production per well and the clustering results, the method for determining the flow units in the reservoir to be divided and the interval boundaries of the flow units may include:

[0142] Obtain the average daily oil production per well of a single well in the reservoir to be divided;

[0143] According to the cluster centers in the clustering results and the cluster center values corresponding to the cluster centers, divide the single wells in the reservoir to be divided to determine the initial flow units in the reservoir to be divided;

[0144] Based on the average daily oil production per well, adjust the initial flow units to obtain the flow units in the reservoir to be divided and the interval boundaries of the flow units. The interval boundaries represent the interval boundaries corresponding to the comprehensive object of the single well and the average daily oil production per well in the flow unit.

[0145] In this step, the average daily oil production per well can be obtained by collecting the production data of the single well. The production data usually includes the daily oil production, liquid production, water cut, etc. Then, for each single well, calculate its corresponding average daily oil production. For example, divide the total oil production of the single well within a preset time period by the number of days to obtain the average daily oil production.

[0146] The number of cluster centers can determine the number of flow unit divisions, and the cluster center values corresponding to the cluster centers can determine the interval boundaries of the comprehensive objects of the single wells corresponding to each cluster center, that is, the interval boundaries of the flow units. Among them, the interval boundaries are determined according to the numerical intervals of the comprehensive objects of the single wells in the clustering results. For example, when K = 3, the division result is 3 types of initial flow units.

[0147] The method for adjusting the initial flow units based on the average daily oil production per well can be to adjust the initial flow units by merging or splitting based on the correlation trend analysis between the single well comprehensive formation coefficient / single well comprehensive reservoir coefficient and the average daily oil production per well, so as to obtain the flow units in the final reservoir to be divided and the interval boundaries of the flow units; it can also be to perform multi-dimensional clustering on the average daily oil production per well and the comprehensive object of the single well based on clustering algorithms such as hierarchical clustering and K-Means, so as to obtain the flow units in the final reservoir to be divided and the interval boundaries of the flow units. This method can better reflect the actual situation of the reservoir, thereby providing a better strategy for oilfield development and management.

[0148] In some examples, such as Figure 4 and Figure 5As shown, for wells with higher single-well comprehensive reservoir coefficients and single-well comprehensive formation coefficients, the average daily oil production level per well is also higher, indicating a positive correlation trend among the single-well comprehensive formation coefficient, the single-well comprehensive reservoir coefficient, and the average daily oil production per well. Therefore, based on the initial flow units, by comparing the single-well comprehensive formation coefficient, the single-well comprehensive reservoir coefficient, and the average daily oil production per well, units with significant differences in daily oil production are identified, and adjustment strategies are formulated to adjust the interval boundaries of the unit. For example, for units with significant differences in daily oil production, consider splitting them into smaller units to make the daily oil production within each new unit more uniform; merge low-production units with similar daily oil production to form new low-production flow units.

[0149] When the comprehensive objects of the selected single well are the single-well comprehensive formation coefficient and the single-well comprehensive reservoir coefficient, and when the clustering result determines the number of clustering centers and the clustering center values for K = 3, the flow units are finally divided into 3 categories according to the clustering result in combination with the average daily oil production. And based on the single-well comprehensive formation coefficient, the single-well comprehensive reservoir coefficient, and the average daily oil production, the interval boundaries of each flow interval in the reservoir to be divided are established. As shown in Table 1, the seepage capacity and reservoir capacity of the first-class flow units are the best, those of the second-class flow units are the second best, and those of the third-class flow units are the worst.

[0150] Table 1: Division results of flow units corresponding to K = 3.

[0151] Single-well flow unit Single-well comprehensive formation coefficient Single-well comprehensive storage coefficient Average daily oil production of a single well Type I flow unit 0.4-1 0.5-1 ＞1 Type II flow unit 0-0.4 0.4-0.8 0.5-1 Type III flow unit 0-0.4 0-0.4 0-0.5

[0152] The method for dividing CO2-driven dynamic flow units in low-permeability oil reservoirs provided by the embodiments of the present application takes the factors affecting the flow characteristics of the reservoir as the objects to be analyzed, reflecting the physical properties and flow capabilities of the reservoir after CO2 flooding in low-permeability oil reservoirs, and using normalization processing to ensure the comparability of each object to be analyzed in cluster analysis. Then, based on the objects to be analyzed after normalization processing, through the first clustering process, the similarity between the objects to be analyzed is determined, and the target objects to be analyzed that have an important impact on the division of flow units are screened out, so as to achieve the purpose of improving the division accuracy while reducing the division error caused by similar objects to be analyzed. Among them, the similarity between the screened target objects to be analyzed is small, which can more accurately reflect the complexity of the reservoir, and achieve the purpose of accurately dividing flow units with fewer target objects to be analyzed, thereby improving the accuracy of dividing flow units in the reservoir with a single well as the dimension.

[0153] Figure 2 It is a schematic flow chart of another method for dividing CO2-driven dynamic flow units in low-permeability oil reservoirs provided by the embodiments of the present application. As Figure 2 shown, the method may include:

[0154] S201. According to the storage characteristics of low-permeability reservoirs in CO2 flooding reservoirs, screen out the influencing and evaluation parameters related to flow units. Among them, the single-type parameters include: pore throat radius, permeability, reservoir depth, porosity, oil saturation, and effective thickness; the combined-type parameters include: inter-well flow capacity index, formation coefficient, reservoir quality factor, and storage coefficient.

[0155] Among them, considering that the heterogeneity within the reservoir flow unit is relatively serious, if we want to accurately divide the flow unit, we first need to clarify the seepage law of the flow unit in the reservoir after CO2 flooding development in low-permeability reservoirs. A flow unit is a reservoir unit with similar lithology and physical properties within the reservoir body. Each flow unit usually represents a specific sedimentary environment and seepage characteristics. Therefore, by analyzing the actual data on the reservoir site, all the parameters that can characterize the flow unit characteristics are screened out, so as to select relatively independent and representative parameters as the classification indicators of the flow unit for subsequent use.

[0156] Among them, after screening out the parameters, it is also necessary to calculate the values of single-type parameters and combined-type parameters in each detection area of the research block of the flow unit to be divided. The detection areas include the well areas in the research block and the reservoir areas in each well.

[0157] S202. Classify all similar parameters based on the Euclidean distance, and screen out relatively independent parameters from numerous parameters to optimize the comprehensive evaluation index for flow unit classification.

[0158] Among them, the distance coefficient is one of the most commonly used classification statistics in hierarchical clustering analysis. The Euclidean distance clustering algorithm is used to calculate the similarity between the similar parameters of the flow unit. In some embodiments, the Euclidean distance clustering algorithm includes: initializing n classes, regarding each parameter as a class. If the calculated Euclidean distance is smaller, the properties of the influencing parameters of the flow unit are closer, then the parameters with relatively small Euclidean distances are classified into a new class; calculate the distances from the new class to other classes, select the smallest distance among them, and merge these two classes into a new class. Repeat the calculation for each parameter according to the above method until finally merged into one class; determine the classification of parameters from multiple combined categories. As Figure 3 shown, the various parameters of the flow unit are classified through the similarity of the minimum distance. Its principle is: regard the data of all parameters as points in space. These points can be connected by edges. The edge weight value between two points with a relatively large distance is lower, while the edge weight value between two points with a relatively small distance is higher. By cutting the graph composed of all data points, make the sum of the edge weights between different subgraphs after cutting as low as possible, and the sum of the edge weights within the subgraph as high as possible, so as to achieve the purpose of clustering. In the figure, the various parameters of the flow unit are divided into three types through the similarity of the minimum distance:

[0159] ① Class I includes the inter-well flow capacity index and formation coefficient;

[0160] ② Class II includes the reservoir quality factor, pore throat radius, permeability, and storage coefficient;

[0161] ③ Class III includes porosity, oil saturation, reservoir depth, and effective thickness.

[0162] In some embodiments, due to the severe reservoir heterogeneity, there is a large gap in the data of each target layer of single-type parameters. Based on the sensitivity analysis of porosity and permeability in single-type parameters, it can be seen that the method of classifying flow units using single parameters such as porosity or permeability does not have wide applicability. Therefore, considering that combined parameters can effectively reduce the errors caused by single-type parameters being too large or too small, in this embodiment, parameters are selected from combined-type parameters as the evaluation parameters for reservoir flow units.

[0163] Among them, the formation coefficient reflects the magnitude of the seepage capacity of the reservoir, the storage coefficient reflects the magnitude of the storage capacity of the reservoir, the inter-well flow capacity index reflects the magnitude of the fluid flow capacity between reservoir layers, and the reservoir quality factor reflects the quality of the seepage of the reservoir.

[0164] By comparing and analyzing these four combined-type parameters, it can be seen that the three evaluation parameters of formation coefficient, reservoir quality factor, and inter-well flow capacity index can all reflect the strength of the flow capacity of the reservoir, and the storage coefficient comprehensively considers the changes in oil saturation and porosity. Considering that the lower limit of the effective thickness of the reservoir changes after CO2 flooding, combined with the magnitude of the oil saturation of the reservoir, the inter-well flow capacity index does not reflect the magnitude of the single-layer flow capacity. Therefore, the formation coefficient and the storage coefficient are selected as the comprehensive evaluation indicators to divide the flow units.

[0165] S203. Calculate and record the magnitudes of the evaluation indicators of each flow unit in the single-well reservoir, and determine the magnitude of the comprehensive evaluation indicator of the single well.

[0166] In this step, the formation coefficient C1 and storage coefficient C2 of all flow units in each reservoir of a single well in a low-permeability oil reservoir after CO2 flooding are calculated through a preset formula. And based on the preset formula, the comprehensive formation coefficient C of the single well 1total and the comprehensive storage coefficient C of the single well 2total are calculated. Furthermore, through the normalization method, the magnitudes of the normalized single-well comprehensive evaluation indicators, namely P1 and P2, are obtained.

[0167] S204. Based on the magnitude of the single-well comprehensive evaluation indicator and the magnitude of the average daily oil production, classify all flow units using the clustering method to obtain the flow unit classification results of each single well and the comprehensive evaluation indicator boundaries corresponding to the flow unit classification results.

[0168] In this step, in the prior art when using the cluster analysis method to divide flow units, usually the cluster analysis method is first applied to classify known samples, and then the stepwise analysis method is applied to discriminate and classify unknown samples, so as to obtain a discriminant function representing different flow units, thus completing the automatic classification of flow units within the reservoir. However, in the actual application process, most of them face the problem of a small number of known samples, and the results of cluster analysis plus discriminant analysis are not ideal. Therefore, in this embodiment, the test data within all ranges of the reservoir in the research block are recorded and the cluster analysis is completed, so as to reduce the processing steps of dividing the flow units.

[0169] After performing cluster analysis on the comprehensive evaluation indicators of a single well, an initial cluster classification result is obtained. Then, the correlation between the dynamic parameter (average daily oil production of a single well) and the single well evaluation indicators is analyzed, and further the relationship between the average daily oil production level of a single well, the comprehensive formation coefficient of a single well, and the comprehensive reservoir coefficient is analyzed. According to the initial cluster classification result, combined with the average daily oil production level of a single well, a dynamic flow unit division standard is established.

[0170] In one example, the formation coefficients and reservoir coefficients of each well and each reservoir in the block of the flow units to be divided are calculated and sorted out; the cluster analysis method uses the mature IBM SPSS Statistics data editor to process the sorted comprehensive formation coefficient P1 of a single well and the comprehensive reservoir coefficient P2 of a single well. First, the sorted data is imported into the SPSS database, the data to be cluster-analyzed is selected, and the Analysis - Classification - K-Means Cluster in the SPSS interface is opened, and the K value of the K-Means cluster (i.e., the number of categories into which the flow units are divided) is set. Relying on the data processing module, the classification result and the cluster intersection diagram of the comprehensive formation coefficient of a single well and the comprehensive reservoir coefficient of a single well can be automatically obtained; according to the K-Means cluster classification result, the span distribution of the comprehensive formation coefficient interval and the comprehensive reservoir coefficient interval corresponding to each single well should be uniform and representative, so as to select the K value that can reflect the distribution of flow units in the entire research block; according to the final K-Means cluster classification result, the boundary values of the comprehensive formation coefficient of a single well and the comprehensive reservoir coefficient corresponding to each type of flow unit can be obtained, and finally the initial division standard of each type of flow unit of each single well in the reservoir is obtained. Then, based on the dynamic parameter (average daily oil production of a single well), its correlations with the comprehensive formation coefficient of a single well and the comprehensive reservoir coefficient are analyzed respectively. As Figure 4 and Figure 5 shown, the comprehensive formation coefficient of a single well and the comprehensive reservoir coefficient of a single well are both positively correlated with the average daily oil production of a single well. Then, according to the actual demand of the average daily oil production of a single well, the initial division standard can be adjusted (such as splitting or merging or re-clustering) to obtain the final division standard.

[0171] For example, the single-well comprehensive formation coefficient and the single-well comprehensive reservoir coefficient are used as the comprehensive evaluation indexes of flow units, and the K-means clustering method is used to perform clustering analysis on the reservoir flow units in the CO2 block. The flow units are divided into 2 categories, 3 categories, and 4 categories respectively (i.e., K values = 2, 3, 4). The clustering center numbers corresponding to different categories of flow units are analyzed, and the clustering center result with a uniform span interval between each clustering center (when K value = 3) is selected to divide the flow units into three categories. Then, based on the average daily oil production of a single well, the actual oil production of each single well in these three categories of flow units is analyzed to obtain the final division standard through a preset method (such as re-clustering analysis or merging / splitting these three categories of flow units). Figure 6 FIG. is a schematic diagram of the clustering result corresponding to the division of three types of flow units provided by an embodiment of the present application, as Figure 6 shown. When the flow units are divided into three categories, the numerical interval boundaries of the single-well comprehensive formation coefficient and the single-well comprehensive reservoir coefficient in different categories can be determined to establish the division standard of the flow units.

[0172] Exemplarily, applying the flow unit classification results of each finally determined single well and the comprehensive evaluation index boundaries corresponding to the flow unit classification results (the division standard shown in Table 1) to the actual development now, the following can be obtained Figure 7 a schematic diagram of the flow unit division under the dynamic change of the daily oil production of different types of single wells as shown; as Figure 8 shown, a schematic diagram of the distribution characteristics of the three types of flow units in the reservoir. Through the analysis of these on-site parameters, it can be known that the seepage capacity and reservoir capacity of the first-type flow units are the best, those of the second-type flow units are the second, and those of the third-type flow units are the worst. The analysis result is the same as the analysis result of the division standard shown in Table 1, thus reflecting the division accuracy of the CO2 drive dynamic flow unit division method for low-permeability oil reservoirs.

[0173] The method for dividing the CO2-driven dynamic flow units in low-permeability reservoirs provided by the embodiments of the present application is to deepen the understanding of the internal fluid flow characteristics in the reservoir after the physical property lower limit of the oilfield developed by CO2 flooding changes. Among them, key reservoir physical properties are selected as evaluation indexes for flow unit division. Through clustering processing, parameters with important influences are initially identified, and then these parameters are deeply analyzed to determine the clustering centers and corresponding values. The dynamic parameter of the average daily oil production per well is introduced. Based on these clustering results and the average daily oil production per well, the reservoirs corresponding to single wells in the research block are accurately divided into flow units, which can finely identify the differences in internal flow characteristics of the reservoir, capture the actual oil production situation, obtain more reasonable division criteria, improve the pertinence and efficiency of reservoir development, optimize resource allocation, reduce development risks, and support scientific decision-making. This method has strong adaptability and flexibility, can be continuously updated and optimized according to new data, and provides a strong scientific basis for reservoir management and development.

[0174] Figure 9 As shown in the structural schematic diagram of the device for dividing the CO2-driven dynamic flow units in low-permeability reservoirs provided by the present application, Figure 9 As shown, the device 30 for dividing the CO2-driven dynamic flow units in low-permeability reservoirs provided by this embodiment includes:

[0175] A first determination module 301, configured to determine the objects to be analyzed of the target layer in the reservoir to be divided, and the objects to be analyzed include at least three of reservoir depth, porosity, permeability, effective thickness, oil saturation, pore throat radius, formation factor, storage coefficient, inter-well flow capacity index, and reservoir quality factor;

[0176] A preprocessing module 302, configured to perform normalization processing on the objects to be analyzed of the target layer in the reservoir to be divided to obtain normalized objects;

[0177] A first clustering module 303, configured to perform first clustering processing on the normalized objects according to the numerical similarity between the normalized objects, and determine the initial target normalized objects in the normalized objects and the initial target objects to be analyzed corresponding to the initial target normalized objects in the objects to be analyzed;

[0178] A screening module 304, configured to perform screening processing on the initial target objects to be analyzed to determine the target objects to be analyzed, and the target objects to be analyzed meet the requirements of the preset type of analysis object and the importance degree of the analysis object;

[0179] A second determination module 305, configured to determine the comprehensive objects of single wells in the reservoir to be divided based on the target objects to be analyzed of the target layer in the reservoir to be divided. A single well represents a test well in the reservoir to be divided, and multiple target layers are included in the test well;

[0180] The second clustering module 306 is configured to perform second clustering processing on the comprehensive objects of a single well to obtain a clustering result, and determine flow units in the reservoir to be divided and the interval boundaries of the flow units based on the average daily oil production of the single well and the clustering result. The clustering result includes cluster centers and the corresponding cluster center values corresponding to the cluster centers.

[0181] In a possible implementation manner, the preprocessing module 302 may also be configured to:

[0182] Determine the values of the objects to be analyzed in the target layer of the reservoir to be divided;

[0183] Analyze the values of the objects to be analyzed to determine the value distribution characteristics of the objects to be analyzed;

[0184] Perform normalization processing on the objects to be analyzed according to the value distribution characteristics to obtain normalized objects.

[0185] In a possible implementation manner, the first clustering module 303 may also be configured to:

[0186] Perform clustering and recombination processing on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after clustering and recombination processing;

[0187] Take the normalized objects after clustering and recombination processing as the normalized objects, and repeat the step of performing clustering and recombination processing on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after clustering and recombination processing until the numerical similarity between the normalized objects after clustering and recombination processing meets the requirements of a preset numerical similarity threshold, and then determine the initial target normalized objects in the normalized objects and the initial target objects to be analyzed in the objects to be analyzed corresponding to the initial target normalized objects.

[0188] In a possible implementation manner, the first clustering module 303 may also be configured to:

[0189] Determine the numerical dimension of the normalized objects, and the data dimension is determined according to the number of target layers in the reservoir to be divided;

[0190] Determine the Euclidean distance between the normalized objects according to the numerical dimension;

[0191] Determine the numerical similarity between the normalized objects according to the Euclidean distance;

[0192] Perform clustering and recombination processing on the normalized objects according to the numerical similarity between the normalized objects to obtain the normalized objects after clustering and recombination processing.

[0193] In a possible implementation manner, the first clustering module 303 may also be configured to:

[0194] Determine a first normalized object, a second normalized object, and other normalized objects according to the numerical similarity between the normalized objects. The numerical similarity between the first normalized object and the second normalized object meets the preset numerical similarity recombination requirement;

[0195] Obtain a recombined normalized object according to the first normalized object and the second normalized object;

[0196] Obtain the normalized object after clustering and recombination processing according to the recombined normalized object and other normalized objects.

[0197] In a possible implementation manner, the screening module 304 can also be used for:

[0198] Determine the combined objects in the initial target object to be analyzed according to the preset requirements for the type of analysis object. The combined objects are analysis objects obtained by performing combined calculation processing on two or more single objects, and the single objects are analysis objects obtained by detecting test data for the target layer in the reservoir to be divided;

[0199] Screen the combined objects according to the influence degree of the single objects on the flow units in the reservoir to be divided, and determine the target object to be analyzed.

[0200] In a possible implementation manner, the second determination module 305 can also be used for:

[0201] Obtain the average daily oil production per well of a single well in the reservoir to be divided;

[0202] Divide the single wells in the reservoir to be divided according to the cluster center in the clustering result and the cluster center value corresponding to the cluster center, and determine the initial flow units in the reservoir to be divided;

[0203] Based on the average daily oil production per well, adjust the initial flow units to obtain the flow units in the reservoir to be divided and the interval boundaries of the flow units. The interval boundaries represent the interval boundaries corresponding to the comprehensive object of the single well and the average daily oil production per well in the flow unit.

[0204] The low-permeability oil reservoir CO2 drive dynamic flow unit division device provided in this embodiment can execute the method provided in the above method embodiment, and its implementation principle and technical effect are similar, which will not be elaborated here in this embodiment.

[0205] Figure 10 It is a schematic structural diagram of the electronic device provided in this application. As Figure 10 shown, the electronic device 40 provided in this embodiment includes: at least one processor 401 and a memory 402. Optionally, the device 40 further includes a communication component 403. Among them, the processor 401, the memory 402, and the communication component 403 are connected through a bus 404.

[0206] In a specific implementation process, at least one processor 401 executes computer-executable instructions stored in a memory 402, so that at least one processor 401 executes the above-mentioned method.

[0207] For the specific implementation process of the processor 401, reference may be made to the above-mentioned method embodiments, and their implementation principles and technical effects are similar, so they will not be elaborated here in this embodiment.

[0208] In the above embodiments, it should be understood that the processor may be a central processing unit (Central Processing Unit, CPU for short), or may also be other general-purpose processors, digital signal processors (Digital Signal Processor, DSP for short), application-specific integrated circuits (Application Specific Integrated Circuit, ASIC for short), etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0209] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0210] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the buses in the drawings of this application are not limited to only one bus or one type of bus.

[0211] This application also provides a computer program product, including a computer program, which implements the above-mentioned method when executed by a processor.

[0212] This application also provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the processor executes the computer-executable instructions, the above-mentioned method is implemented.

[0213] The above-readable storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disk. The readable storage medium may be any available medium accessible by a general-purpose or special-purpose computer.

[0214] An exemplary readable storage medium is coupled to the processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium may also be a component of the processor. The processor and the readable storage medium may be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium may also exist as discrete components in a device.

[0215] The division of units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection between each other may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

[0216] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0217] In addition, each functional unit in various embodiments of the present invention may be integrated in a processing unit, may also exist separately as individual physical units, or two or more units may be integrated in one unit.

[0218] If a function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs and other various media that can store program codes.

[0219] Those of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When this program is executed, it executes the steps including the above method embodiments; and the aforementioned storage medium includes: ROMs, RAMs, magnetic disks, or optical discs and other various media that can store program codes.

[0220] Finally, it should be noted that: after considering the specification and practicing the invention disclosed herein, those skilled in the art will easily think of other implementation manners of the present invention. The present invention is intended to cover any variations, uses, or adaptive changes of the present invention. These variations, uses, or adaptive changes follow the general principles of the present invention and include the common general knowledge or conventional technical means in the technical field of the present invention that are not disclosed in the present invention. It is not limited to the exact structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present invention is only limited by the appended claims.

Claims

1. A method for dividing CO2-driven flow units in low permeability reservoirs, characterized in that: include: Determine the object to be analyzed of the target layer in the reservoir to be divided, the object to be analyzed includes a single object and a combined object, the combined object is an analysis object obtained by combining and calculating two or more single objects, the single object is an analysis object obtained by testing the target layer in the reservoir to be divided by test data, the single object includes reservoir depth, porosity, permeability, effective thickness, oil saturation and pore throat radius, the combined object includes formation coefficient, reservoir coefficient, well flow capacity index and reservoir quality factor; Normalizing the object to be analyzed in the target layer of the reservoir to be divided to obtain a normalized object; According to the numerical similarity between the normalized objects, a first clustering process is performed on the normalized objects to determine an initial target normalized object among the normalized objects and an initial target object to be analyzed among the objects to be analyzed that corresponds to the initial target normalized object; Screening the initial target objects to be analyzed to determine target objects to be analyzed, wherein the target objects to be analyzed meet preset requirements for the type of analysis objects and the importance of analysis objects; Based on the target object to be analyzed of the target layer in the reservoir to be divided, determining a comprehensive object of a single well in the reservoir to be divided, wherein the single well represents a test well in the reservoir to be divided, and the test well includes a plurality of target layers; A second clustering process is performed on the comprehensive object of the single well to obtain a clustering result, and based on the average daily oil production of the single well and the clustering result, the flow units in the reservoir to be divided and the interval boundaries of the flow units are determined. The clustering result includes a cluster center and a cluster center value corresponding to the cluster center.

2. The method according to claim 1, characterized in that The step of normalizing the object to be analyzed in the target layer in the reservoir to be divided to obtain a normalized object includes: Determine the value of the object to be analyzed in the target layer of the reservoir to be divided; Analyzing the value of the object to be analyzed to determine the value distribution characteristics of the object to be analyzed; According to the numerical distribution characteristics, the object to be analyzed is normalized to obtain a normalized object.

3. The method according to claim 1, characterized in that The step of performing a first clustering process on the normalized objects according to the numerical similarities between the normalized objects, and determining an initial target normalized object in the normalized objects, and an initial target object to be analyzed in the objects to be analyzed corresponding to the initial target normalized object, comprises: According to the numerical similarities between the normalized objects, clustering and reorganizing the normalized objects to obtain normalized objects after clustering and reorganizing; The normalized object after the clustering and reorganization processing is used as the normalized object, and the step of performing clustering and reorganization processing on the normalized object according to the numerical similarity between the normalized objects to obtain the normalized object after the clustering and reorganization processing is repeatedly performed, until the numerical similarity between the normalized objects after the clustering and reorganization processing meets the preset numerical similarity threshold requirement, and then the initial target normalized object in the normalized object and the initial target object to be analyzed corresponding to the initial target normalized object in the objects to be analyzed are determined.

4. The method according to claim 3, characterized in that: The step of performing clustering and reorganizing processing on the normalized objects according to the numerical similarities between the normalized objects to obtain normalized objects after clustering and reorganizing processing includes: Determining a numerical dimension of the normalized object, wherein the numerical dimension is determined according to the number of target layers in the reservoir to be divided; Determining the Euclidean distance between the normalized objects according to the numerical dimension; Determining the numerical similarity between the normalized objects according to the Euclidean distance; According to the numerical similarities between the normalized objects, clustering and reorganization processing is performed on the normalized objects to obtain normalized objects after clustering and reorganization processing.

5. The method according to claim 3, characterized in that: The step of performing clustering and reorganizing processing on the normalized objects according to the numerical similarities between the normalized objects to obtain normalized objects after clustering and reorganizing processing includes: Determining a first normalized object, a second normalized object, and other normalized objects according to the numerical similarity between the normalized objects, wherein the numerical similarity between the first normalized object and the second normalized object meets a preset numerical similarity recombination requirement; Obtaining a recombined normalized object according to the first normalized object and the second normalized object; According to the reorganized normalized object and the other normalized objects, a normalized object after clustering and reorganization processing is obtained.

6. The method according to claim 1, characterized in that The screening process of the initial target object to be analyzed to determine the target object to be analyzed, wherein the target object to be analyzed meets the preset analysis object type requirement and analysis object importance requirement, includes: According to the preset analysis object type requirements, determining the combination type objects in the initial target objects to be analyzed; According to the influence degree of the single type object on the flow unit in the reservoir to be divided, the combined type objects are screened to determine the target object to be analyzed.

7. The method according to claim 1, characterized in that The step of determining the flow units in the reservoir to be divided and the interval boundaries of the flow units based on the average daily oil production of the single well and the clustering result includes: Obtaining the average daily oil production of a single well in the reservoir to be divided; According to the cluster center in the clustering result and the cluster center value corresponding to the cluster center, the single well in the reservoir to be divided is divided to determine the initial flow unit in the reservoir to be divided; Based on the average daily oil production of the single well, the initial flow unit is adjusted to obtain the flow unit in the reservoir to be divided and the interval limit of the flow unit. The interval limit represents the comprehensive object of the single well in the flow unit and the interval limit corresponding to the average daily oil production of the single well.

8. A device for dividing flow units in a CO2-driven state in a low permeability reservoir, characterized in that: include: The first determination module is used to determine the object to be analyzed of the target layer in the reservoir to be divided, the object to be analyzed includes a single object and a combined object, the combined object is an analysis object obtained by combining and calculating two or more single objects, the single object is an analysis object obtained by testing the target layer in the reservoir to be divided by test data, the single object includes reservoir depth, porosity, permeability, effective thickness, oil saturation and pore throat radius, and the combined object includes formation coefficient, reservoir coefficient, well flow capacity index and reservoir quality factor; A preprocessing module, used for normalizing the object to be analyzed in the target layer in the reservoir to be divided to obtain a normalized object; A first clustering module, configured to perform a first clustering process on the normalized objects according to the numerical similarities between the normalized objects, and determine an initial target normalized object among the normalized objects, and an initial target object to be analyzed among the objects to be analyzed that corresponds to the initial target normalized object; A screening module is used to screen the initial target object to be analyzed to determine the target object to be analyzed, wherein the target object to be analyzed meets the preset analysis object type requirement and analysis object importance requirement; A second determination module is used to determine a comprehensive object of a single well in the reservoir to be divided based on a target object to be analyzed of the target layer in the reservoir to be divided, wherein the single well represents a test well in the reservoir to be divided, and the test well includes a plurality of target layers; The second clustering module is used to perform a second clustering process on the comprehensive object of the single well to obtain a clustering result, and based on the average daily oil production of the single well and the clustering result, determine the flow units in the reservoir to be divided and the interval boundaries of the flow units. The clustering result includes a cluster center and a cluster center value corresponding to the cluster center.

9. An electronic device, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed.

Images