A method for dividing the bottom interface of the surface karst zone based on dynamic time series matching technology
Through dynamic timing matching technology, the logging data of standard wells and target wells is used to generate matching maps, which solves the problem of inaccurate identification of surface karst zones in a single well, and achieves efficient and accurate automatic identification and division, improving identification accuracy and operation efficiency.
Patent Information
- Application Number
- CN202210527988.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-16
- Publication Date
- 2024-12-03
- Estimated Expiration
- 2042-05-16
AI Technical Summary
The prior art is difficult to accurately identify and divide the surface karst zone in a single well that lacks imaging logging data, resulting in limited identification accuracy and it is difficult to accurately divide the surface karst zone.
Dynamic timing matching technology (DTW) is adopted, and the automatic matching program is compiled, and the matching map is generated using the conventional logging response values of standard wells and target wells, and the comparison results are automatically output to achieve the promotion and identification of the bottom interface of the karst belt between the surface layer between single wells. The specific steps include reading single well data, performing logarithmic processing and maximum-minimum standardization, calculating cumulative distances, and generating matching path maps to divide the surface karst belt bottom interface.
It achieves relatively accurate matching results, is simple to operate, can automatically batch processing, is efficient and has strong operability.
Smart Images

Figure CN114895371B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of quantitative characterization of carbonate reservoirs, and in particular to a method for dividing a bottom interface of a surface karst zone based on a dynamic time series matching technology. Background Art
[0002] Carbonate reservoirs are very important oil and gas reservoirs, and karst fracture-cave reservoirs are one of the important types. Due to differences in hydrodynamic conditions, paleokarst fracture-cave reservoirs can be divided from top to bottom into surface karst zones, vertical seepage zones, horizontal undercurrent zones, and deep slow-flow zones. Surface reservoirs are developed in the surface karst zone, which is a favorable location for further increasing reserves and production of fracture-cave reservoirs. Its development thickness is generally within 40m below the unconformity surface. Accurately identifying karst zoning and finely depicting reservoir structure are important research topics at present, which have guiding significance for the actual development of oil and gas fields.
[0003] At present, the karst zoning of single wells is mainly based on conventional logging and imaging logging data. Through the interval changes of logging response values and the various types of reservoir spaces reflected by downhole images, geological researchers manually interpret the karst zoning, which is inefficient. For single wells that lack imaging logging data, the recognition accuracy is limited, and it is often difficult to accurately divide the surface karst zone. Summary of the invention
[0004] In order to solve the above problems, the present invention provides a method for dividing the bottom interface of the surface karst zone based on the dynamic time series matching technology (DTW), which is used to solve the problem of inaccurate identification of the surface karst zone in a single well with missing imaging logging data. By compiling an automatic matching program and automatically outputting the comparison results, the generalized identification of the bottom interface of the surface karst zone between single wells is realized. It mainly includes the following steps:
[0005] S1: Acquire single well data of standard well and single well data of target well;
[0006] S2: using a dynamic time series matching program to read the above single well data and generate a standard well-target well matching map;
[0007] S3: According to the atlas generated in step S2, the depth position of the matching comparison between the target single well and the standard single well is found, and the bottom interface of the surface karst zone of the target single well is divided.
[0008] Furthermore, the specific implementation steps of step S2 are as follows:
[0009] S21: Read the standard well data, and use the vector P of the sequence of its conventional logging response values along the depth as a reference template for dynamic timing matching:
[0010] X i =(GR i ,CALi ,RD i ,RS i ,AC i ,DEN i ,CNL i )
[0011] P=(X1,X2,....,X i ,...,X m )
[0012] i=1,2,...,m;
[0013] Among them, X i are logging response values at different depths, and m represents the number of samples along the depth of the 7 conventional logging response values of the standard well;
[0014] S22: Read the target well data, and use the vector Q to represent the sequence of its conventional logging response values along the depth, and use it as a test template for dynamic timing matching, Y j are the logging response values at different depths:
[0015] Y j =(GR j ,CAL j ,RD j ,RS j ,AC j ,DEN j ,CNL j )
[0016] Q=(Y1,Y2,....,Y j ,...,Y n )
[0017] j=1,2,...,m;
[0018] After selecting the sequence of logging response values of the standard well and the target well along the depth, the values of RD and RS in the sequence samples are logarithmically processed;
[0019] S23: normalizing all 7 types of logging values in the sequence sample using the maximum-minimum normalization method;
[0020] S24: Based on the processed data, construct an m×m matrix M∈R m×m , the distance between any two points in the matrix is calculated, and the cumulative distance is further calculated. According to the cumulative distance, the well logging curve matching path diagram with the smallest distance when the sequences P' and Q' are matched together is obtained. This diagram is the standard well-target well matching map.
[0021] Furthermore, the maximum-minimum normalization process is to perform a linear transformation on the original data. Taking the GR value as an example, the minimum and maximum values of the GR logging response are GRmin and GRmax, respectively. An original value of attribute A is mapped to the interval (0, 1) through the maximum-minimum normalization, and the calculation formula is:
[0022]
[0023] Among them, GR i is the original GR logging response value, GR i ' is the normalized value of GR logging response value, GR max ,GR min They represent the maximum response value and the minimum response value of the original GR logging respectively.
[0024] Furthermore, the distance between any two points in the matrix is calculated using the Euclidean distance calculation method shown below:
[0025] d(X' i ,Y' j )=(X' i -Y' j ) 2
[0026] Among them, d(X' i ,Y' j ) represents two points X' in the matrix M i and Y' j The distance when aligning.
[0027] Furthermore, the calculation formula of the accumulated distance D(i,j) is as follows:
[0028]
[0029] Where D(X' i ,Y' j ) is the sequence P' and Q' at point X' i and Y' i The cumulative distance of all previous points, It is the minimum distance when matching along the directions of points (i-1, j-1), (i-1, j) and (i, j-1).
[0030] The technical solution provided by the present invention has the following beneficial effects: the matching result of the method is relatively accurate, the operation method is easy, and automatic batch comparison processing can be realized, the efficiency is high, and the operability is strong. BRIEF DESCRIPTION OF THE DRAWINGS
[0031] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:
[0032] Figure 1 It is a flow chart of a method for dividing the bottom interface of the surface karst zone based on dynamic time series matching technology in an embodiment of the present invention.
[0033] Figure 2 It is a comparison result diagram of the standard well-target well division in an embodiment of the present invention. DETAILED DESCRIPTION
[0034] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described in detail with reference to the accompanying drawings.
[0035] An embodiment of the present invention provides a method for dividing the bottom interface of a surface karst zone based on a dynamic time series matching technology.
[0036] Please refer to Figure 1 , Figure 1 1 is a flow chart of a method for dividing the bottom interface of the surface karst zone based on the dynamic time series matching technology in an embodiment of the present invention, which specifically includes the following steps:
[0037] (1) Find a single well in the study area that has a clear result of dividing the bottom interface of the surface karst zone and use it as a standard well. The conventional logging data of the standard well required for the study is sorted into an Excel table to obtain the table of the following 7 conventional logging data of the standard well unconformity surface as shown in Table 1:
[0038]
[0039] According to the depth, natural gamma (GR), well diameter (CAL), deep lateral (RD), shallow lateral (RS), acoustic wave (AC), density (DEN), and neutron (CNL), the depth section is from the unconformity surface to the lower 45m. The depth of the unconformity surface of the well is 5823.5m, and the sorted depth section is 5823.5m-5868.5m.
[0040] (2) Find the single well in the study area that needs to be divided into the bottom interface of the surface karst zone, that is, the target well, and organize its conventional logging data in the same way as step (1). Organize its conventional logging data into an Excel spreadsheet and obtain the following:
[0041] Table 2 shows the table of the following 7 conventional logging data of the target well unconformity surface:
[0042]
[0043] According to the depth, GR, CAL, RD, RS, AC, DEN, and CNL, the depth section is from the unconformity surface to the lower 45m. The depth of the unconformity surface of the well is 5810.0m, and the sorted depth section is 5810.0m-5855.0m.
[0044] (3) The dynamic time series matching program reads the well data and calculates and generates the standard well-target well matching map. The specific principle is as follows:
[0045] (3.1) Read the standard well data, and use the vector P to represent the sequence of its conventional logging response values along the depth, which is used as the reference template for dynamic timing matching. i are the logging response values at different depths:
[0046] X i =(GR i ,CAL i ,RD i ,RS i ,AC i ,DEN i ,CNL i )
[0047] P=(X1,X2,....,X i ,...,X m )
[0048] i=1,2,...,m;
[0049] Wherein, m represents the number of samples along the depth of the 7 conventional logging response values of the standard well;
[0050] (3.2) Read the target well data, and represent the sequence of its conventional logging response values along the depth with vector Q, which is used as the test template for dynamic timing matching, Y j are the logging response values at different depths:
[0051] Y j =(GR j ,CAL j ,RD j ,RS j ,AC j ,DEN j ,CNL j )
[0052] Q=(Y1,Y2,....,Y j ,...,Y n );
[0053] j=1,2,...,m;
[0054] After selecting the sequence of logging response values of the standard well and the target well along the depth, the values of RD and RS in the sequence samples are logarithmically processed. Taking the RD value as an example, the calculation formula is:
[0055] RD new = = log(RD i )
[0056] (3.3) All 7 types of logging values in the sequence samples are normalized using the maximum-minimum normalization method. Specifically, the maximum-minimum normalization is to perform a linear transformation on the original data. Taking the GR value as an example, the minimum and maximum values of the GR logging response are GR min and GR max , an original value of attribute A is mapped to the interval (0, 1) through maximum-minimum normalization, and the calculation formula is:
[0057]
[0058] Among them, GR i is the original GR logging response value, GR i ' is the normalized value of GR logging response.
[0059] (3.4) We need to construct an m×m matrix M∈R m×m , where the matrix element m i,j Represents X' i and Y' i The distance between two points d(X' i ,Y' j ), where the distance is Euclidean distance, which is calculated as follows:
[0060] d(X' i ,Y' j )=(X' i -Y' j ) 2
[0061] Where each element m in the matrix M i,j Both represent point X' i and Y' i The distance during alignment, the matching process is the alignment process, and the two points correspond one to one.
[0062] (3.5) It is necessary to find a path through several grid points in the matrix M, that is, the optimal matching path, so that after the path passes through the grid points, the two sequences P' and Q' can be matched together to the maximum extent, so that the matching distance between the two sequences P' and Q' is minimized. The sequences P' and Q' are obtained by taking the logarithm and normalizing the original sequences P and Q.
[0063] In order to make the dynamic timing matching process converge, the grid of the matrix M through which the optimal matching path passes must be continuous, monotonically increasing, and start at the grid (1,1) and end at the grid (m,m).
[0064] To complete this step, we need to define the cumulative distances, that is, from the starting point (1,1), each time the two sequences P' and Q' match a point, we need to calculate the cumulative distance of all previous points. After reaching the end point (m,m), this cumulative distance is the final cumulative distance, which represents the similarity between sequences P' and Q'. The calculation formula of the cumulative distance D(i,j) is as follows:
[0065]
[0066] Where D(X' i ,Y' j ) is the sequence P' and Q' at point X' i and Y' j The cumulative distance of all previous points, It is the minimum distance when matching along the directions of (i-1,j-1), (i-1,j) and (i,j-1).
[0067] (4) Figure 2 As shown, according to the well logging curve matching path diagram (i.e. Figure 2 The optimal dynamic matching path diagram in the figure is used to find the depth position of the matching comparison between the target single well and the standard single well, and to divide the bottom interface of the surface karst zone of the target single well. The specific process is as follows:
[0068] Find the depth corresponding to the bottom interface of the surface layer on the standard well; find the point where the depth is located in the standard well-target well matching map; find the depth corresponding to the target well based on the point, which is the depth of the bottom interface of the surface karst zone. In this embodiment, it is known that the depth of the bottom interface of the surface karst zone of the standard well is 5838.5m, which corresponds to point A in the matching map. The depth of the target well corresponding to this point is 5825.0m. Based on this, the depth of the bottom interface of the surface karst zone of the target well is divided into 5825.0m.
[0069] The beneficial effects of the present invention are: the matching result of the method is relatively accurate, the operation method is easy, and automatic batch comparison processing can be realized, the efficiency is high, and the operability is strong.
[0070] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims
1. A method for dividing the bottom interface of the surface karst zone based on dynamic time series matching technology, characterized in that: The following steps are involved: S1: Acquire single well data of standard well and single well data of target well; S2: using a dynamic time series matching program to read the above single well data and generate a standard well-target well matching map; The specific implementation steps of step S2 are as follows: S21: Read the standard well data, and use the vector P of the sequence of its conventional logging response values along the depth as a reference template for dynamic timing matching: X i =(GR i ,CAL i ,RD i ,RS i ,AC i ,DEN i ,CNL i ) P=(X1,X2,….,X i ,…,X m ) i=1,2,...,m; Among them, X i are logging response values at different depths, GRi, CALi, RDi, RSi, ACi, DENi, CNLi represent the natural gamma, well diameter, deep lateral, shallow lateral, acoustic wave, density, and neutron in the reference template respectively, and m represents the number of samples along the depth of the 7 conventional logging response values of the standard well; S22: Read the target well data, and use the vector Q to represent the sequence of its conventional logging response values along the depth, and use it as a test template for dynamic timing matching, Y j are the logging response values at different depths: Y j =(GR j ,CAL j ,RD j ,RS j ,AC j ,DEN j ,CNL j ) Q=(Y1,Y2,....,Y j ,...,Y n ) j=1,2,...,m; Among them, GRj, CALj, RDj, RSj, ACj, DENj, CNLj represent the corresponding natural gamma, well diameter, deep lateral, shallow lateral, acoustic wave, density, and neutron at different depths of logging response values respectively; After selecting the sequence of logging response values of the standard well and the target well along the depth, the values of RD and RS in the sequence are logarithmically processed; S23: All 7 types of logging values in the sequence samples are normalized using the maximum-minimum normalization method; S24: Based on the processed data, construct an m×m matrix M∈R m×m , calculate the distance between any two points in the matrix, further calculate the cumulative distance, the cumulative distance is the final cumulative distance, and the well logging curve matching path diagram with the minimum distance when the sequences P' and Q' are matched together is obtained according to the cumulative distance. This diagram is the standard well-target well matching map; S3: According to the atlas generated in step S2, the depth position of the matching comparison between the target single well and the standard single well is found, and the bottom interface of the surface karst zone of the target single well is divided.
2. The method for dividing the bottom interface of the surface karst zone based on the dynamic time series matching technology according to claim 1, characterized in that: The maximum-minimum normalization process is to perform a linear transformation on the original data. The minimum and maximum values of the GR logging response are GRmin and GRmax respectively. An original value of attribute A is mapped to the interval (0, 1) through the maximum-minimum normalization. The calculation formula is: Among them, GR i is the original GR logging response value, GR i ' is the normalized value of GR logging response value, GR max ,GR min They represent the maximum response value and the minimum response value of the original GR logging respectively.
3. The method for dividing the bottom interface of the surface karst zone based on dynamic time series matching technology according to claim 1, characterized in that: The distance between any two points in the matrix is calculated using the Euclidean distance calculation method as shown below: d(X' i ,Y' j )=(X' i -Y' j ) 2 Among them, d(X' i ,Y' j ) represents two points X' in the matrix M i and Y' j The distance when aligning.
4. The method for dividing the bottom interface of the surface karst zone based on the dynamic time series matching technology according to claim 1, characterized in that: The calculation formula of the cumulative distance D(i,j) is as follows: Where D(X' i ,Y' j ) is the sequence P' and Q' at point X' i and Y' i The cumulative distance of all previous points, It is the minimum distance when matching along the directions of points (i-1, j-1), (i-1, j) and (i, j-1).
