Reservoir quality grading method considering dynamic and static parameters
Through numerical simulation model and principal component analysis technology, combined with cluster analysis, a reservoir quality evaluation system with dynamic and static parameters was established, which solved the difficulty and accuracy of the acquisition of reservoir quality grading indicators in the existing technology, and achieved reservoir quality The efficient and accurate classification of the oil field provides a reliable evaluation basis for oil field development.
Patent Information
- Application Number
- CN202411119809.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-08-02
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-12
AI Technical Summary
The prior art has problems with the difficulty of obtaining indicators and relying on human experience for reservoir quality grading methods, resulting in low grading accuracy and efficiency.
A numerical simulation model is used to combine principal component analysis and cluster analysis to establish a multi-dimensional evaluation index system for dynamic and static parameters, calculate the grid comprehensive score through principal component analysis, and divide reservoir quality categories using cluster analysis.
It realizes simple, accurate and efficient grading of reservoir quality, reduces the influence of engineers' subjective judgment, and provides a reliable evaluation basis for oil field development.
Smart Images

Figure CN118933721A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of oil development, and in particular to a reservoir quality classification method considering dynamic and static parameters. Background Art
[0002] In oil development technology, it is necessary to classify the quality of reservoirs in consideration of economic benefits.
[0003] The inventor of this application found in the research that the existing methods for grading reservoir quality include grading methods based on a single indicator, which may not be the best indication of reservoir development effect; the existing methods for grading reservoir quality also include multi-indicator evaluation methods, such as considering the mainstream throat radius, starting pressure gradient, viscosity mineral content, crude oil viscosity and other indicators to grade the reservoir of low permeability fault block reservoirs. There are two problems with this evaluation method. First, the selected indicators are difficult to obtain and require a large number of indoor mine field experiments; second, the evaluation and grading boundaries are artificially set, and their accuracy depends largely on the experience of engineers.
[0004] In summary, there is no simple and efficient reservoir quality classification method in the prior art, and further research is necessary. Summary of the invention
[0005] In response to the above problems, the purpose of this application is to provide a reservoir quality grading method that takes dynamic and static parameters into consideration, selects reasonable indicators to establish an evaluation index system, and performs a quantifiable grading method for the evaluation results, which can be simple, feasible, accurate and efficiently implementable reservoir quality grading.
[0006] To achieve the above objectives, this application adopts the following technical solutions:
[0007] In a first aspect, the present application provides a reservoir quality classification method considering dynamic and static parameters, the method comprising:
[0008] Establishing a numerical simulation model of the reservoir to be evaluated, wherein the numerical simulation model has a plurality of grids arranged in relative positions in the reservoir to be evaluated, and each grid corresponds to a parameter value of a preset category parameter;
[0009] Calculating a comprehensive score of each grid based on principal component analysis according to parameter values of preset category parameters corresponding to each grid of the numerical simulation model;
[0010] Based on the comprehensive score of each grid, cluster analysis is used to divide the grids into a preset number of reservoir quality categories.
[0011] In one implementation of the present invention, the preset category parameters include two types of parameter indicators: production potential related parameters and flow potential related parameters;
[0012] The production potential related parameters include oil saturation S o , reserve abundance N, thickness H, porosity
[0013] The flow potential related parameters include formation pressure P, permeability K, water-oil viscosity ratio μ R , grid connectivity number N.
[0014] In one implementation of the present invention, the calculating of the comprehensive score of each grid based on principal component analysis includes:
[0015] According to the parameter values of the production potential related parameters and the flow potential related parameters of each grid, an initial parameter matrix A is constructed;
[0016] Standardize the element values in the initial parameter matrix A;
[0017] According to the standardized values, calculate the correlation matrix R between the parameter indicators;
[0018] Calculate the eigenvalues and eigenvectors of the features according to the correlation matrix R, and construct the principal components based on the eigenvectors;
[0019] Select a set number of principal components to calculate the composite score for each grid.
[0020] In one implementation of the present invention, the initial parameter matrix A=(a ij ) n×8 ; Among them, i takes the value of [1,n], n is the number of grids; j takes the value of [1,8], corresponding to the number of parameters related to production potential and flow potential.
[0021] In one implementation of the present invention, the standardization process includes:
[0022] Standard value The implementation process meets:
[0023]
[0024] in: a' j 、s j is the sample mean and sample standard deviation of the jth indicator, a j is the parameter value of the jth indicator; correspondingly, is the standardized indicator vector, and its expression is:
[0025]
[0026] In one implementation of the present invention, the correlation coefficient matrix R=(r kj ) 8×8 ,satisfy:
[0027]
[0028] Where: r kk =1, r kj =r jk is the correlation coefficient between the kth index and the jth index.
[0029] In one implementation of the present invention, the eigenvalue λ of the correlation coefficient matrix R is calculated j , the eigenvalue satisfies λ1≥λ2≥···≥λ8≥0, and the corresponding standardized eigenvectors μ1, μ2,···,μ8 are calculated at the same time, where μ j =[μ 1j , μ 2j ,···,μ nj ] T ;
[0030] The eigenvectors form 8 new index vectors, and the index vectors satisfy:
[0031]
[0032] Among them: y1 is the first principal component, y2 is the second principal component, and so on. j is the jth principal component.
[0033] In one implementation of the present invention, p (p≤8) principal components are selected and the eigenvalue λ is calculated. j Information contribution rate and cumulative contribution rate:
[0034] Principal component y j The information contribution rate is:
[0035]
[0036] Principal components y1, y2,···,y p The cumulative contribution rate is calculated as follows:
[0037]
[0038] Choose the number of principal components: when α p When ≥0.95, the corresponding minimum p value is the number of principal components;
[0039] Calculate the comprehensive score Z for each grid i :
[0040]
[0041] The reservoir scoring results are normalized, and the processing process meets the following requirements:
[0042]
[0043] In one implementation of the present invention, the cluster analysis processing step includes:
[0044] (1) Determine the number of clusters k and select k samples as the initial cluster centers;
[0045] (2) Calculate the Euclidean distance d from all grid scoring results to each cluster center i , since the scoring result of each grid is a one-dimensional constant, the Euclidean distance d i can be simplified to:
[0046] d i =Z i -Z c
[0047] Where i is the grid number; Z c is any given cluster center;
[0048] (3) According to the principle of minimum Euclidean distance, samples are assigned to clusters, the mean of sample points in each cluster is calculated, and the cluster center is updated;
[0049] (4) Repeat steps (2) and (3). When the set number of iterations is reached or the mean vector of the cluster no longer changes, the model is built and the clustering algorithm results are output.
[0050] In one implementation of the present invention, the cluster number is set to 2, which is divided into general reservoirs and dominant reservoirs.
[0051] The present invention adopts the above technical solution, which has the following advantages:
[0052] (1) A multi-dimensional evaluation index system for reservoir quality was established from the two aspects of production potential and flow potential, and the principal component analysis method was used to achieve a comprehensive evaluation of reservoir quality;
[0053] (2) The evaluation results are graded using the k-means cluster analysis method, avoiding the uncertainty caused by the subjective grading of individual engineers;
[0054] (3) The present invention can achieve reservoir quality classification more accurately and efficiently, providing a basis for oil field development potential evaluation and development plan formulation. BRIEF DESCRIPTION OF THE DRAWINGS
[0055] Figure 1 Schematic diagram of the grid connectivity number in an embodiment of the present invention;
[0056] Figure 2 Schematic diagram of comprehensive evaluation results of grid principal components in an embodiment of the present invention;
[0057] Figure 3 It is a schematic diagram of the reservoir potential classification result in the embodiment of the present application. DETAILED DESCRIPTION
[0058] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.
[0059] In view of the existing problems. The present invention accordingly provides a reservoir quality grading method that takes into account dynamic and static parameters, including: establishing a numerical simulation model of the reservoir to be evaluated, wherein the numerical simulation model has a plurality of grids arranged according to relative positions in the reservoir to be evaluated, and each grid corresponds to a parameter value of a preset category parameter; according to the parameter value of the preset category parameter corresponding to each grid in the numerical simulation model, a comprehensive score of each grid is calculated based on principal component analysis; according to the comprehensive score of each grid, cluster analysis is used to divide the grids into a preset number of reservoir quality categories. The present invention selects reasonable indicators to establish an evaluation index system, and performs a quantifiable grading method on the evaluation results, which can simply and efficiently realize reservoir quality grading.
[0060] The above method and process are described below in more detailed embodiments in conjunction with more drawings of the present invention.
[0061] The present invention adopts the following steps to implement the technical solution:
[0062] Step 1. Establish a numerical simulation model.
[0063] Step 2. Determine the parameters for evaluating the reservoir quality, mainly considering the production potential and flow potential.
[0064] (1) Production potential related parameters: mainly oil saturation S o , reserve abundance N, thickness H, porosity These parameters can be directly derived from the established numerical simulation model.
[0065] (2) Flow potential related parameters: mainly including formation pressure P, permeability K, water-oil viscosity ratio μ R , the number of grid connections N. Among them, P, K, μ RIt can be directly derived from the numerical simulation model. N is defined as the number of grids connected to the grid, and its value is:
[0066] ① The grid is located at the corner of the model, and N is 1;
[0067] ② The grid is located at the edge of the model, and N is 2 or 3;
[0068] ③The grid is located in the middle of the model, and N is 4.
[0069] Generally speaking, the larger N is, the better the reservoir grid connectivity is. The schematic diagram of the grid connectivity number is as follows: Figure 1 Indicates.
[0070] Step 3. Use the principal component analysis method to score the quality of each grid: As described in step 1, the principal component analysis evaluation index x j There are 8 grids in total, so j is [1,8]; i represents the number of each grid, and there are n grids in total, so i is [1,n]. The quality characteristic values of the i-th grid are denoted as a ij , the entire reservoir constitutes the matrix A = (a ij ) n×8 The detailed calculation process is as follows:
[0071] (1) Standardize the raw data obtained from statistics. The implementation process meets:
[0072]
[0073] in: a' j 、s j is the sample mean and sample standard deviation of the jth indicator. Correspondingly, is the standardized indicator vector, and its expression is:
[0074]
[0075] (2) Calculate the correlation matrix R between the indicators, and its correlation coefficient matrix R = (r kj ) 8×8 ,satisfy:
[0076]
[0077] Where: r kk =1, r kj =r jk is the correlation coefficient between the kth index and the jth index.
[0078] (3) Calculate the eigenvalues and eigenvectors. Calculate the eigenvalues λ of the correlation coefficient matrix R j, the eigenvalue satisfies λ1≥λ2≥···≥λ8≥0, and the corresponding standardized eigenvectors μ1, μ2,···,μ8 are calculated at the same time, where μ j =[μ 1j , μ 2j ,···,μ nj ] T .
[0079] The eigenvectors form 8 new index vectors, and the index vectors satisfy:
[0080]
[0081] Among them: y1 is the first principal component, y2 is the second principal component, and so on. j is the jth principal component.
[0082] (4) Select p (p≤8) principal components and calculate the comprehensive evaluation value.
[0083] ① Calculate the eigenvalue λ j Information contribution rate and cumulative contribution rate.
[0084] Principal component y j The information contribution rate is:
[0085]
[0086] Principal components y1, y2,···,y p The cumulative contribution rate is calculated as follows:
[0087]
[0088] ②Select the number of principal components: When α p When ≥0.95, the corresponding minimum p-value is the number of principal components.
[0089] ③Calculate the comprehensive score Z of each grid i :
[0090]
[0091] ④Normalize the reservoir scoring results, and the processing process meets the following requirements:
[0092]
[0093] Results After principal component analysis, the scoring results of each grid in the block are obtained, as shown in the schematic diagram Figure 2 .
[0094] Step 4. Use the cluster analysis algorithm to classify the scoring results in step 2.
[0095] (1) Determine the number of clusters k and select k samples as the initial cluster centers. Here, the reservoirs are divided into two categories: dominant reservoirs and general reservoirs, and k is 2;
[0096] (2) Calculate the Euclidean distance d from all grid scoring results to each cluster center i , since the scoring result of each grid is a one-dimensional constant, the Euclidean distance d i can be simplified to:
[0097] d i =Z i -Z c (9)
[0098] Where i is the grid number; Z c is any given cluster center;
[0099] (3) According to the “Euclidean distance minimum principle”, samples are assigned to clusters, the mean of sample points in each cluster is calculated, and the cluster center is updated;
[0100] (4) Repeat steps (2) and (3). When the set number of iterations is reached or the mean vector of the cluster no longer changes, the model is built and the clustering algorithm results are output.
[0101] According to the clustering results, the reservoir quality is divided into two categories. The schematic diagram of the classification results is as follows Figure 3 .
[0102] In the several embodiments provided by the present invention, it should be understood that the disclosed method can be implemented in other ways. For example, the embodiments described above are merely illustrative.
[0103] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor (Processor) to perform some steps of the above-mentioned method of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.
[0104] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. A reservoir quality classification method considering dynamic and static parameters, characterized in that: The method comprises: Establishing a numerical simulation model of the reservoir to be evaluated, wherein the numerical simulation model has a plurality of grids arranged in relative positions in the reservoir to be evaluated, and each grid corresponds to a parameter value of a preset category parameter; Calculate the comprehensive score of each grid based on principal component analysis according to the parameter values of the preset category parameters corresponding to each grid of the numerical simulation model; Based on the comprehensive score of each grid, cluster analysis is used to divide the grids into a preset number of reservoir quality categories.
2. The reservoir quality classification method considering dynamic and static parameters according to claim 1, characterized in that: The preset category parameters include two types of parameter indicators: production potential related parameters and flow potential related parameters; The production potential related parameters include oil saturation S o , reserve abundance N, thickness H, porosity The flow potential related parameters include formation pressure P, permeability K, water-oil viscosity ratio μ R , grid connectivity number N.
3. The reservoir quality classification method considering dynamic and static parameters according to claim 2 is characterized in that: The comprehensive score of each grid is calculated based on principal component analysis, including: According to the parameter values of the production potential related parameters and the flow potential related parameters of each grid, an initial parameter matrix A is constructed; Standardize the element values in the initial parameter matrix A; According to the standardized values, the correlation matrix R between the parameter indicators is calculated; Calculate the eigenvalues and eigenvectors of the features according to the correlation matrix R, and construct the principal components based on the eigenvectors; Select a set number of principal components to calculate the composite score for each grid.
4. The reservoir quality classification method considering dynamic and static parameters according to claim 3 is characterized in that: The initial parameter matrix A=(a ij ) n×8 ; Among them, i takes the value of [1,n], n is the number of grids; j takes the value of [1,8], corresponding to the number of parameters related to production potential and flow potential.
5. The reservoir quality classification method considering dynamic and static parameters according to claim 4, characterized in that: The standardization process comprises: Standard value The implementation process meets: in: a' j 、s j is the sample mean and sample standard deviation of the jth indicator, a j is the parameter value of the jth indicator; correspondingly, is the standardized indicator vector, and its expression is:
6. The reservoir quality classification method according to claim 5, characterized in that: The correlation coefficient matrix R = (r kj ) 8×8 ,satisfy: Where: r kk =1, r kj =r jk is the correlation coefficient between the kth index and the jth index.
7. The reservoir quality classification method considering dynamic and static parameters according to claim 6, characterized in that: Calculate the eigenvalue λ of the correlation coefficient matrix R j , the eigenvalue satisfies λ1≥λ2≥···≥λ8≥0, and the corresponding standardized eigenvectors μ1, μ2,···,μ8 are calculated at the same time, where μ j =[μ 1j , μ 2j ,···,μ nj ] T ; The eigenvectors form 8 new index vectors, and the index vectors satisfy: Among them: y1 is the first principal component, y2 is the second principal component, and so on. j is the jth principal component.
8. The reservoir quality classification method considering dynamic and static parameters according to claim 7, characterized in that: Select p (p≤8) principal components and calculate the eigenvalue λ j Information contribution rate and cumulative contribution rate: Principal component y j The information contribution rate is: Principal components y1, y2,···,y p The cumulative contribution rate is calculated as follows: Choose the number of principal components: when α p When ≥0.95, the corresponding minimum p value is the number of principal components; Calculate the comprehensive score Z for each grid i : The reservoir scoring results are normalized, and the processing process meets the following requirements:
9. The reservoir quality classification method considering dynamic and static parameters according to claim 8, characterized in that: The processing steps of cluster analysis include: (1) Determine the number of clusters k and select k samples as the initial cluster centers; (2) Calculate the Euclidean distance d from all grid scoring results to each cluster center i , since the scoring result of each grid is a one-dimensional constant, the Euclidean distance d i can be simplified to: d i =Z i -Z c Where i is the grid number; Z c is any given cluster center; (3) According to the principle of minimum Euclidean distance, samples are assigned to clusters, the mean of sample points in each cluster is calculated, and the cluster center is updated; (4) Repeat steps (2) and (3). When the set number of iterations is reached or the mean vector of the cluster no longer changes, the model is built and the clustering algorithm results are output.
10. The reservoir quality classification method considering dynamic and static parameters according to claim 9, characterized in that: The cluster number is set to 2, which is divided into general reservoirs and dominant reservoirs.
