Quantitative evaluation and analysis method for anisotropism of reservoir

Abstract

The invention discloses a quantitative evaluation and analysis method for anisotropism of a reservoir, relates to the technical field of geological evaluation of oil and gas field development, and combines seismic data and logging data which can obtain parameters such as reservoir rock mechanics, reservoir quality and reservoir ground stress at present. The advantages of large seismic data coverage range and high logging data precision are utilized, meanwhile, the defects of low seismic data precision and small logging data coverage range are overcome, the combined target layer section is long, quantitative evaluation on the heterogeneity of the reservoir is achieved, and the method is suitable for mass production. The defects of difficult sample acquisition, low representativeness, single characterization capability, insufficient dynamic characteristic characterization and the like when an experimental method is adopted to evaluate the heterogeneity of the reservoir are avoided, and meanwhile, numerical realization of the characteristics of the whole reservoir is realized. Scientific basis is provided for segmentation, clustering and perforation of three-dimensional fracturing of oil and gas reservoirs and control of extension paths and forms of hydraulic fractures, and effectiveness and accuracy of reservoir transformation are improved.

Description

Technical Field

[0001] This invention relates to the field of geological evaluation technology for oil and gas field development, and more specifically to a method for quantitative evaluation and analysis of reservoir heterogeneity. Background Technology

[0002] Evaluating reservoir heterogeneity is fundamental to the design of oil and gas reservoir stimulation programs and the simulation of fracture propagation. Existing methods mainly rely on laboratory core experiments to determine permeability or porosity and use Lorentz curves to evaluate reservoir heterogeneity. However, these methods have the following problems:

[0003] (1) Difficulty in coring. It is necessary to test the porosity or permeability of a large number of reservoir cores. However, most wells, especially those in development blocks, no longer cor them, making it impossible to evaluate the heterogeneity of the reservoir where the new well is located through experimental cores.

[0004] (2) Low representativeness. The evaluation method based on indoor experiments is mainly based on core test data, but the length and range of core sampling are limited. It is impossible to core the entire horizontal section of the reservoir. Moreover, the core sampling usually comes from the vertical section. Therefore, for reservoirs with strong heterogeneity, core analysis cannot represent the heterogeneity of the reservoir and cannot obtain core data outside the wellbore, thus failing to characterize the heterogeneity of the entire reservoir.

[0005] (3) Single characterization. Porosity and permeability alone cannot characterize the heterogeneity of other reservoir parameters, such as stress heterogeneity and rock mechanical heterogeneity.

[0006] (4) Static data. The data obtained from the laboratory experiments are static data, and their characteristics may differ significantly from those under reservoir conditions.

[0007] (5) The process is complex and has low scalability. For example, by analyzing the entropy value obtained by logging signals, the degree of heterogeneity of different well sections can be qualitatively analyzed by comparing the entropy values ​​of different well sections. This method is complex and can only qualitatively evaluate the differences in heterogeneity of different well sections, but cannot perform quantitative analysis.

[0008] (6) Evaluation lag. For example, the method of obtaining empirical modes and Hilbert plots from fracturing operation curves to determine reservoir heterogeneity is based on the curves after fracturing. This evaluation lags and cannot provide support for reservoir stimulation scheme design and fracture numerical simulation.

[0009] (7) Limited scope and difficulty in data acquisition. For example, the entropy weight algorithm is used to calculate reservoir heterogeneity using permeability data.

[0010] (8) Cannot be directly applied to numerical implementation. All existing methods only evaluate reservoir heterogeneity and do not establish a relationship with the numerical implementation of reservoir heterogeneity characteristics. Therefore, they cannot directly apply the heterogeneity characteristics to reservoir stimulation scheme design and fracture propagation numerical simulation. Summary of the Invention

[0011] To overcome the shortcomings of the existing technologies, this invention discloses a quantitative evaluation and analysis method for reservoir heterogeneity. This invention utilizes seismic data, well logging data, Lorentz curves, and conversion methods for shape parameters in Gini coefficients and Weibull distributions to achieve quantitative evaluation of oil and gas reservoir heterogeneity and numerical realization of reservoir heterogeneity characteristics, including rock mechanical heterogeneity, reservoir physical property heterogeneity, and geostress heterogeneity. This provides a scientific basis for segmentation, clustering, perforation, and control of hydraulic fracture extension paths and morphology in "three-dimensional fracturing" of oil and gas reservoirs, improving the effectiveness and accuracy of reservoir stimulation.

[0012] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0013] A method for quantitative evaluation and analysis of reservoir heterogeneity includes the following steps:

[0014] I. Parameter Data Extraction

[0015] S1. Acquire logging data of the target reservoir and extract heterogeneous characteristic parameter data from the logging data;

[0016] Preferably, in step S1, parameters for evaluating the heterogeneity of the target reservoir are obtained through well logging, including reservoir rock mechanical parameters, reservoir physical property parameters, and reservoir in-situ stress parameters; wherein, the reservoir rock mechanical parameters include Young's modulus, Poisson's ratio, and fracture pressure; the reservoir physical property parameters include porosity, permeability, total organic carbon content, and clay content; and the reservoir in-situ stress parameters include maximum horizontal principal stress, minimum horizontal principal stress, and vertical stress.

[0017] II. Completeness Assessment and Missing Data Handling

[0018] S2. Perform a completeness check on the parameter data and complete any missing data identified.

[0019] Preferably, step S2 includes the following steps:

[0020] S21. Each parameter in the calculation parameter data should have a data volume;

[0021] Preferably, step S21 includes:

[0022]

[0023] Where Num0 is the required amount of data for the parameter; L is the total length from point A to point B in the horizontal well, in meters; D is the data interval for logging, in meters; and k is the parameter type, including rock mechanics parameter R, reservoir physical property parameter P, and geostress parameter S.

[0024] S22. Determine whether the required data volume for each parameter is equal to the actual logging data volume. If yes, proceed to step S26. If no, find and determine the missing well section through programming, and then proceed to step S23.

[0025] S23. Combine drilling and logging data to determine whether the well section with missing logging data and the well section without missing data belong to the same formation. If they belong to the same formation, proceed to step S24. If they do not belong to the same formation or the lithology is different, proceed to step S25.

[0026] S24. The logging data of the missing well section is improved by interpolation method, and then the process proceeds to step S26. The interpolation method includes linear regression interpolation, random forest interpolation, and multiple interpolation method.

[0027] S25. Calibrate the parameters based on the seismic data and the characteristics of the same layers in adjacent wells. Use the calibrated parameter values ​​as the parameter values ​​for the missing well section of the target well, and then proceed to step S26.

[0028] S26. Use the extreme value normalization method to transform the parameter data into a new data volume.

[0029] Preferably, the extreme value normalization method in step S26 includes:

[0030]

[0031] in, This is the normalized value of the i-th data point of the k-th class of parameters; P is the i-th data of class k parameters; k For all data of class k parameters; Max(P) k Min(P) represents the maximum value of all data in class k parameters; k ) represents the minimum value of all data for parameter type k; k is the parameter type; i is the data index.

[0032] Preferably, in step S26, the transformed data is not less than 0, and the spacing between each parameter point is consistent.

[0033] III. Generating the Lorentz Curve

[0034] S3. Using well logging data and parameter data, generate Lorentz curves;

[0035] Preferably, step S3 includes the following steps:

[0036] S31. Calculate the cumulative percentage of reservoir length covered by the well logging data.

[0037] Preferably, step S31 includes:

[0038]

[0039] Among them, L n L represents the length of the reservoir section corresponding to the logging parameters, in meters; L represents the total length of the target reservoir, in meters. is the cumulative percentage of reservoir interval length, and n is the total amount of logging parameter data.

[0040] S32, Calculate the cumulative percentage of logging parameters

[0041] Preferably, step S32 includes:

[0042]

[0043] Among them, K n This is the nth data point of parameter K; K represents different parameter types, including rock mechanics parameters, reservoir physical property parameters, and geostress parameters. The cumulative percentage of logging parameters; n is the total amount of data; k is the parameter type.

[0044] S33. Arrange all the cumulative percentages of reservoir length and parameter percentages in ascending order, and plot the Lorentz curve Q on the Lorentz curve diagram with the cumulative percentage of reservoir length as the horizontal axis and the cumulative percentage of parameter percentages as the vertical axis.

[0045] S34. In the Lorentz curve diagram, draw a straight line M passing through the two points A(0,0) and B(100,100).

[0046] IV. Analysis of Reservoir Heterogeneity

[0047] S4. Use the Lorentz curve to analyze reservoir heterogeneity;

[0048] Preferably, in step S4, the reservoir heterogeneity includes rock mechanical heterogeneity, reservoir quality heterogeneity, and stress heterogeneity.

[0049] Preferably, step S4 includes: fitting a linear relationship to the Lorentz curve Q through programming, and determining the slope k of each fitted line. m This includes the following steps:

[0050] S41. Let p be the total number of data in the Lorentz curve, and let n = m = i = 1, where n, m, and i are counts;

[0051] S42. Points in the Lorentz curve are represented as follows: The data set is represented as {X}, where n = n + 1;

[0052] S43. Determine if i equals 5. If not, go to step S42 and set i = i + 1; otherwise, proceed to the next step.

[0053] S44. Perform a linear fit on the data set {X} and calculate the goodness of fit R. 2 ;

[0054] S45. Add a new data point, i.e., n = n + 1, and incorporate the new data into a new dataset, i.e., {X} = {X}. n+1};

[0055] S46. Perform linear fitting on the data set {X} again and calculate the goodness of fit R. N 2 ;

[0056] S47. Determine the goodness of fit R. 2 and R N 2 relative difference ratio Is it less than 0.1? If yes, proceed to step S45; otherwise, proceed to the next step.

[0057] S48. Let i = 1, calculate the slope k of the curve formed by the data set {X}. m ;

[0058] S49. Determine if all data has been analyzed, i.e., determine if n equals p. If n = p, then summarize all slopes to form a slope data set {k}. m Otherwise, m = m + 1, and adjust to step S42.

[0059] Preferably, in step S4, the analysis of reservoir heterogeneity using the Lorentz curve includes:

[0060] If the slope k m The closer it is to 1, the better it indicates that k m The smaller the difference between all the corresponding data, the weaker the heterogeneity;

[0061] Adjacent slope difference ratio The closer to 0, the better k m and k m+1 The smaller the differences between all the corresponding data, the weaker the heterogeneity.

[0062] V. Quantitative evaluation of heterogeneity

[0063] S5. Use the Lorentz curve to determine the Gini coefficient, then use the Gini coefficient to determine the heterogeneity parameter in the Weibull distribution. After that, combine seismic and well logging data, use the Weibull distribution to assign reservoir values, and quantitatively evaluate the reservoir heterogeneity.

[0064] Preferably, step S5 includes the following steps:

[0065] S51. Calculate the area S1 of the triangle formed by points A(0,0), B(100,100), and C(100,0) in the Lorentz curve graph.

[0066] S52. Calculate the area S2 formed by the curve Q and the straight line M;

[0067] Preferably, in step S52, if the well section step length corresponding to the data is large, the trapezoidal integration method is used to calculate the area S2; if the well section step length corresponding to the data is small, the numerical integration method is used to calculate the area S2 through programming.

[0068] S53. Calculate the Gini coefficient using areas S1 and S2, where Gini = S2 / S1;

[0069] Preferably, in step S53, the Gini coefficient ranges from 0 to 1.0, and the correspondence between the Gini coefficient and the degree of parameter heterogeneity includes:

[0070] When 0 ≤ Gini coefficient ≤ 0.2, the degree of heterogeneity is extremely weak;

[0071] When 0.2 < Gini coefficient ≤ 0.3, the degree of heterogeneity is relatively weak;

[0072] When 0.3 < Gini coefficient ≤ 0.4, the degree of heterogeneity is moderate;

[0073] When 0.4 < Gini coefficient ≤ 0.5, the degree of heterogeneity is relatively strong;

[0074] When 0.5 < Gini coefficient ≤ 1.0, the degree of heterogeneity is extremely strong.

[0075] S54. Use the Gini coefficient to determine the non-homogeneity parameter m in the Weibull distribution;

[0076] Preferably, step S54 includes:

[0077] If any parameter has a Gini coefficient less than 0.2, then take m0 = 6 based on the smallest Gini coefficient;

[0078] If all the Gini coefficients of the parameters are greater than 0.2, then take m0 = 1.5 based on the largest Gini coefficient;

[0079] The non-homogeneous parameters of the remaining parameters are determined by the ratio of the Gini coefficient corresponding to m0 to the Gini coefficients of other parameters.

[0080] Step S54 concerns the relationship, or conversion method, between the Gini coefficient and the heterogeneity coefficient *m* in the Weibull distribution. This relationship is an empirical formula obtained through multiple calculations. For example, if the calculated Gini coefficient for Young's modulus is 0.33, the Gini coefficient for Poisson's ratio is 0.15, and the Gini coefficient for burst pressure is 0.41, and since the Poisson's ratio Gini coefficient of 0.15 is less than 0.2, then the heterogeneity parameter *m0* corresponding to Gini = 0.15 is 6. The ratio of Gini = 0.15 to Young's modulus Gini = 0.33 is 0.45, therefore the heterogeneity parameter *m* corresponding to Young's modulus is 0.45 * *m0* = 0.45 * 6 = 2. 7; For example, the calculated Gini coefficient for Young's modulus is 0.33, the Gini coefficient for Poisson's ratio is 0.26, and the Gini coefficient for fracture pressure is 0.41. Since all of these Gini coefficients are greater than 0.2, the heterogeneity parameter m0 corresponding to Gini = 0.41 is 1.5. The ratio of Gini = 0.41 to Young's modulus Gini = 0.33 is 1.24. Therefore, the heterogeneity coefficient m corresponding to Young's modulus is m = 1.24 * m0 = 1.24 * 1.5 = 1.86. After determining the heterogeneity parameter m, the Weibull distribution can be used for grid assignment to visualize the degree of reservoir heterogeneity, providing a basis for numerical simulation.

[0081] The non-mean parameter m ranges from 1.5 to 6.0.

[0082] S55. Using the scale of the 3D seismic data as the grid scale, assign the 3D seismic data values ​​to the corresponding grid as grid data values;

[0083] S56. The grid for each seismic scale is further subdivided, with the subdivision scale based on the logging accuracy or the accuracy required for later simulations. The Weibull distribution is used for parameter assignment, including: using the values ​​in the seismic grid as the scale parameter X in the Weibull distribution, using the logging accuracy as the dimension of the Weibull distribution matrix, and using the heterogeneity parameter m as the shape parameter in the Weibull distribution to assign values ​​to the entire reservoir, thereby quantitatively evaluating the reservoir heterogeneity.

[0084] Preferably, step S56 includes:

[0085] Establish the Weibull probability distribution function Where X is the parameter value corresponding to each seismic scale grid, x is the assigned parameter, and m is the non-homogeneous parameter;

[0086] The programmatically implements the Weibull probability distribution function, performing a sampling operation for each calculation. Each sampling yields a set of grid scales that can be further subdivided according to the logging or simulation requirements.

[0087] The sampled data is input into the rock mechanics simulator to obtain the rock stress-strain curve under the data conditions. It is then compared with the core stress-strain curve of the corresponding layer obtained in the experiment. If the characteristics of the rock stress-strain curve under the sampled data conditions are significantly different from those of the core stress-strain curve under the experimental conditions, then after multiple samplings, its mechanical properties are found to be consistent with those of the real core.

[0088] The beneficial effects of this invention are:

[0089] This invention combines seismic and well logging data, which are currently available for obtaining parameters such as reservoir rock mechanics, reservoir quality, and reservoir in-situ stress. It leverages the advantages of seismic data's wide coverage and well logging data's high accuracy while overcoming the disadvantages of low seismic data accuracy and limited well logging data coverage. By incorporating long target stratigraphic intervals, it achieves quantitative evaluation of reservoir heterogeneity, avoiding the difficulties in obtaining samples, low representativeness, limited characterization capabilities, and insufficient dynamic feature representation inherent in experimental methods for evaluating reservoir heterogeneity. Furthermore, it achieves numerical realization of the entire reservoir's characteristics. This provides a scientific basis for the segmentation, clustering, perforation, and control of hydraulic fracture extension paths and morphology in "three-dimensional fracturing" of oil and gas reservoirs, improving the effectiveness and precision of reservoir stimulation. Attached Figure Description

[0090] Figure 1 This is a flowchart of step 2 in an embodiment of the present invention;

[0091] Figure 2 This is a flowchart of the curve segmentation slope calculation process of the present invention.

[0092] Figure 3 This is a Lorentz curve diagram of the present invention;

[0093] Figure 4 This is a schematic diagram illustrating the assignment of values ​​to grid cells based on seismic scale in this invention;

[0094] Figure 5 The Weibull distribution is assigned a value in conjunction with the logging scale in this invention. Detailed Implementation

[0095] The following will provide a clear and complete description of the concept, specific structure, and technical effects of the present invention in conjunction with the embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention.

[0096] A method for quantitative evaluation and analysis of reservoir heterogeneity includes the following steps:

[0097] 1. Obtain the rock mechanical parameters (R), reservoir physical property parameters (P), in-situ stress parameters (S), or other parameters whose heterogeneity characteristics need to be evaluated through logging. The rock mechanical parameters include, but are not limited to, Young's modulus, Poisson's ratio, fracture pressure, etc.; the reservoir physical property parameters include, but are not limited to, porosity, permeability, TOC, clay content, etc.; the in-situ stress parameters include the maximum horizontal principal stress, the minimum horizontal principal stress, the vertical stress, etc.

[0098] 2. Judge and process all the parameters that need to be evaluated in step 1. Logging usually obtains the parameters (including all parameters at that point) of each point every 0.15 m, but due to equipment or other factors, there may be cases where the relevant parameters at some points are missing. Therefore, it is necessary to judge the data integrity and intervene in and improve the data at the missing points to ensure the data quality for heterogeneity analysis. As Figure 1 shown, this step includes:

[0099] (1) Judging the integrity of logging data.

[0100] ① Calculate the expected data volume of each parameter:

[0101]

[0102] Where: L is the total length (m) from point A to point B of the horizontal well; D is the data interval (m) of logging, which can be 0.15 m or other values; k is the parameter type, including rock mechanical parameters (R), reservoir physical property parameters (P), in-situ stress parameters (S), or other parameters whose heterogeneity characteristics need to be evaluated, etc.

[0103] ② Compare the relationship between the expected data volume (Num0) and the actual data volume (Num1). If Num1 < Num0, search for the missing well sections through programming.

[0104] (2) Completing the logging data.

[0105] ① Combine drilling and logging data to judge whether the well sections with missing logging data and the well sections without missing data belong to the same horizon. If they are in the same horizon, complete the logging data of the missing well sections through interpolation methods, including, but not limited to, linear regression interpolation, random forest interpolation, or multiple imputation methods, etc.

[0106] ② If the well sections with missing logging data and the well sections without missing data do not belong to the same horizon or have different lithologies, it is necessary to calibrate using seismic data and the parameter characteristics of the same horizon in adjacent wells, and use the calibrated parameter values as the parameter values of the missing well sections of the target well.

[0107] (3) Use the min-max normalization method to transform the original data (or the data improved according to step 2) into a new data body: Alternatively, other normalization methods can be used, but it is necessary to ensure that the transformed data is not less than 0. It is important to note that the spacing between all parameter points must be consistent, such as 0.15m or 0.3m between points.

[0108] 3. Calculate the cumulative percentage of reservoir length covered by the well logging data, as shown in the following formula:

[0109]

[0110] Among them, L n L represents the length of the reservoir section corresponding to the logging parameters, in meters; L represents the total length of the target reservoir, in meters. is the cumulative percentage of reservoir interval length, and n is the total amount of logging parameter data.

[0111] 4. Calculate the cumulative percentage of logging parameters, as shown in the following formula:

[0112]

[0113] Here, K represents different parameter types, such as R (rock mechanics parameters), P (reservoir physical property parameters), or S (geostress parameters), etc.

[0114] 5. All P L n P k n The values ​​are arranged in ascending order and expressed as a cumulative percentage of reservoir length (P). L n The x-axis represents the cumulative percentage of parameters (P). k n Plot curve Q with y as the vertical axis.

[0115] 6. In the curve plotted in step 5, draw a straight line M passing through points A(0,0) and B(100,100).

[0116] 7. First, reservoir heterogeneity is analyzed using curve analysis. Reservoir heterogeneity includes, but is not limited to, rock mechanical heterogeneity, reservoir quality heterogeneity, and stress heterogeneity. For example... Figure 2 and Figure 3 As shown, the analysis method is as follows:

[0117] (1) Fit the linear relationship of the data through programming and determine the slope k of each fitted line. m ,like Figure 2 As shown:

[0118] ① Let P be the total number of data points in the Lorentz curve, and let n = m = i = 1;

[0119] ② The points in the Lorentz curve are represented as The data set is represented as {X}, where n = n + 1;

[0120] ③ Determine if i equals 5. If not, go back to step ② and set i = i + 1; otherwise, proceed to the next step.

[0121] ④ Perform linear fitting on the data set {X} and calculate the goodness of fit R. 2 ;

[0122] ⑤ Add one data point at a time, i.e., n = n + 1, and incorporate the new data into a new dataset, i.e., {X} = {X} n+1};

[0123] ⑥ Perform linear fitting on the data set {X} again and calculate the goodness of fit R. N 2 ;

[0124] ⑦ Determine the goodness of fit R 2 and R N 2 The relative difference ratio |R 2 -R N 2 | / R 2 Is it less than 0.1? If yes, proceed to step ⑤; otherwise, proceed to the next step.

[0125] ⑧ Let i = 1, calculate the slope K of the curve formed by this data set. m ;

[0126] 9. Determine if all data has been analyzed, i.e., determine if n equals p. If n = p, then summarize all slopes to form a slope data set {K}. m Otherwise, m = m + 1, and adjust to step ②.

[0127] (2) The analysis method is as follows:

[0128] ①If the slope k m The closer it is to 1, the better it indicates that k m The smaller the difference between all the corresponding data, the weaker the heterogeneity;

[0129] ② Ratio of adjacent slope differences (|k m -k m+1 | / k m The closer k is to 0, the better. m and k m+1 The smaller the differences between all the corresponding data, the weaker the heterogeneity.

[0130] 8. The method for quantitatively evaluating heterogeneity is as follows: Calculate the area S1 of the triangle formed by points A(0,0), B(100,100), and C(100,0), where the area is a constant;

[0131] S1 = 5000

[0132] 9. Calculate the area S2 enclosed by curve Q and line M. If the well section step size corresponding to the data is large, the trapezoidal integration method can be used to calculate S2. If the well section step size corresponding to the data is small, the numerical integration method can be used to calculate it through programming.

[0133] 10. Calculate the Gini coefficient, Gini = S² / S¹, where the Gini coefficient ranges from 0 to 1.0. The relationship between the Gini coefficient and the degree of heterogeneity of the parameters is shown in the table below:

[0134] Gini coefficient Heterogeneity 0～0.2 extremely weak 0.2～0.3 Weak 0.3～0.4 medium 0.4～0.5 Strong 0.5～1.0 Extremely strong

[0135] 11. Numerical Realization of Reservoir Heterogeneity Characteristics: The heterogeneity parameter (m) in the Weibull distribution is determined to realize the numerical representation of the heterogeneity characteristics of reservoir parameters. The correspondence between the heterogeneity parameter in the Weibull distribution and the Gini coefficient is as follows:

[0136] (1) If any parameter has a Gini coefficient less than 0.2, take m0 = 6 as the minimum Gini coefficient;

[0137] (2) If the Gini coefficients of all parameters are greater than 0.2, take m0 = 1.5 based on the maximum Gini coefficient;

[0138] (3) The rest are determined based on the ratio of the Gini coefficient corresponding to m0 to the Gini coefficient of other parameters.

[0139] 12. Using the scale of the 3D seismic data as the grid scale, assign the 3D seismic data values ​​to the corresponding grids as grid data values, such as... Figure 4 As shown. Currently, seismic data has low precision, typically using data points at 10m or 5m intervals. This means that a certain parameter value for the reservoir is the same at either the 10m or 5m scale, which fails to reflect the heterogeneous characteristics of the reservoir. After assigning values ​​to the seismic data, proceed to the next step.

[0140] 13. Based on step 12, the grid for each seismic scale is further subdivided. The subdivision scale is determined by the logging accuracy or the accuracy required for later simulations. According to the results of step 11, the Weibull distribution is used for parameter assignment. The values ​​in the seismic grid are used as the scale parameter (X) in the Weibull distribution, the logging accuracy is used as the dimension of the Weibull distribution matrix, and the m value from step 11 is used as the shape parameter in the Weibull distribution for full reservoir assignment. This method integrates seismic and logging data to numerically represent the reservoir heterogeneity characteristics, maximizing the representation of both macroscopic and microscopic heterogeneity. The assignment method is as follows, and the assignment results are as follows. Figure 5 As shown:

[0141] (1) The Weibull probability distribution function is: Where X is the parameter value corresponding to each seismic scale grid, x is the assigned parameter (such as rock mechanics parameters, reservoir quality parameters, etc.), and m is the result of step 11;

[0142] (2) Implement the Weibull probability distribution in programming. Each operation is a sampling. Each sampling obtains a set of grid scales (dimensions) that meet the requirements of well logging or simulation in step 13.

[0143] (3) Input the data obtained from the sampling into the rock mechanics simulator and obtain the rock stress-strain curve under the data conditions. Compare it with the rock core stress-strain curve obtained from the experiment. If the characteristics of the rock stress-strain curve under the sampling data conditions are significantly different from the characteristics of the rock core stress-strain curve under the experimental conditions, multiple samplings are required to ensure that its mechanical properties are consistent with the real rock core.

[0144] The embodiments of the present invention have been described in detail above, but the present invention is not limited to the described embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and these equivalents or substitutions are all included within the scope defined by the claims of the present invention.

Claims

1. A method for quantitative evaluation and analysis of reservoir heterogeneity, characterized in that, Includes the following steps: S1. Acquire logging data of the target reservoir and extract heterogeneous characteristic parameter data from the logging data; S2. Perform a completeness check on the parameter data and complete any missing data identified. S3. Using well logging data and parameter data, generate Lorentz curves; S4. Use the Lorentz curve to analyze reservoir heterogeneity; S5. Use the Lorentz curve to determine the Gini coefficient, then use the Gini coefficient to determine the heterogeneity parameter in the Weibull distribution. After that, combine seismic and well logging data, use the Weibull distribution to assign reservoir values, and quantitatively evaluate the reservoir heterogeneity.

2. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 1, characterized in that, In step S1, parameters for evaluating the heterogeneity of the target reservoir are obtained through well logging, including reservoir rock mechanical parameters, reservoir physical property parameters, and reservoir in-situ stress parameters. The reservoir rock mechanical parameters include Young's modulus, Poisson's ratio, and fracture pressure; the reservoir physical property parameters include porosity, permeability, total organic carbon content, and clay content; and the reservoir in-situ stress parameters include maximum horizontal principal stress, minimum horizontal principal stress, and vertical stress.

3. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 1, characterized in that, Step S2 includes the following steps: S21. Each parameter in the calculation parameter data should have a data volume; S22. Determine whether the required data volume for each parameter is equal to the actual logging data volume. If yes, proceed to step S26. If no, find and determine the missing well section through programming, and then proceed to step S23. S23. Combine drilling and logging data to determine whether the well section with missing logging data and the well section without missing data belong to the same formation. If they belong to the same formation, proceed to step S24. If they do not belong to the same formation or the lithology is different, proceed to step S25. S24. The logging data of the missing well section is improved by interpolation method, and then the process proceeds to step S26. The interpolation method includes linear regression interpolation, random forest interpolation, and multiple interpolation method. S25. Calibrate the parameters based on the seismic data and the characteristics of the same layers in adjacent wells. Use the calibrated parameter values ​​as the parameter values ​​for the missing well section of the target well, and then proceed to step S26. S26. Use the extreme value normalization method to transform the parameter data into a new data volume.

4. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 3, characterized in that, Step S21 includes: Where Num0 is the required amount of data for the parameter; L is the total length from point A to point B in the horizontal well, in meters; D is the data interval for logging, in meters; and k is the parameter type, including rock mechanics parameter R, reservoir physical property parameter P, and geostress parameter S.

5. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 3, characterized in that, The extreme value normalization methods in step S26 include: in, This is the normalized value of the i-th data point of the k-th class of parameters; P is the i-th data of class k parameters; k For all data of class k parameters; Max(P) k Min(P) represents the maximum value of all data in class k parameters; k ) represents the minimum value of all data for parameter type k; k is the parameter type; i is the data index.

6. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 1, characterized in that, Step S3 includes the following steps: S31. Calculate the cumulative percentage of reservoir length covered by the well logging data. S32, Calculate the cumulative percentage of logging parameters S33. Arrange all the cumulative percentages of reservoir length and parameter percentages in ascending order, and plot the Lorentz curve Q on the Lorentz curve diagram with the cumulative percentage of reservoir length as the horizontal axis and the cumulative percentage of parameter percentages as the vertical axis. S34. In the Lorentz curve diagram, draw a straight line M passing through the two points A(0,0) and B(100,100).

7. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 6, characterized in that, Step S31 includes: Among them, L n L represents the length of the reservoir section corresponding to the logging parameters, in meters; L represents the total length of the target reservoir, in meters. is the cumulative percentage of reservoir interval length, and n is the total amount of logging parameter data.

8. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 6, characterized in that, Step S32 includes: Among them, K n This is the nth data point of parameter K; K represents different parameter types, including rock mechanics parameters, reservoir physical property parameters, and geostress parameters. The cumulative percentage of logging parameters; n is the total amount of data; k is the parameter type.

9. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 6, characterized in that, Step S4 includes: fitting a linear relationship to the Lorentz curve Q via programming, and determining the slope k of each fitted line. m This includes the following steps: S41. Let p be the total number of data in the Lorentz curve, and let n = m = i = 1, where n, m, and i are counts; S42. Points in the Lorentz curve are represented as follows: The data set is represented as {X}, where n = n + 1; S43. Determine if i equals 5. If not, go to step S42 and set i = i + 1; otherwise, proceed to the next step. S44. Perform a linear fit on the data set {X} and calculate the goodness of fit R. 2 ; S45. Add a new data point, i.e., n = n + 1, and incorporate the new data into a new dataset, i.e., {X} = {X}. n+1 }; S46. Perform linear fitting on the data set {X} again and calculate the goodness of fit R. N 2 ; S47. Determine the goodness of fit R. 2 and R N 2 relative difference ratio Is it less than 0.1? If yes, proceed to step S45; otherwise, proceed to the next step. S48. Let i = 1, calculate the slope k of the curve formed by the data set {X}. m ; S49. Determine if all data has been analyzed, i.e., determine if n equals p. If n = p, then summarize all slopes to form a slope data set {k}. m Otherwise, m = m + 1, and adjust to step S42.

10. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 9, characterized in that, In step S4, the analysis of reservoir heterogeneity using the Lorentz curve includes: If the slope k m The closer it is to 1, the better it indicates that k m The smaller the difference between all the corresponding data, the weaker the heterogeneity; Adjacent slope difference ratio The closer to 0, the better k m and k m+1 The smaller the differences between all the corresponding data, the weaker the heterogeneity.

11. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 6, characterized in that, The S5 step includes the following steps: S51. Calculate the area S1 of the triangle formed by points A(0,0), B(100,100), and C(100,0) in the Lorentz curve graph. S52. Calculate the area S2 formed by the curve Q and the straight line M; S53. Calculate the Gini coefficient using areas S1 and S2, where Gini = S2 / S1; S54. Use the Gini coefficient to determine the non-homogeneity parameter m in the Weibull distribution; S55. Using the scale of the 3D seismic data as the grid scale, assign the 3D seismic data values ​​to the corresponding grid as grid data values; S56. The grid for each seismic scale is further subdivided, with the subdivision scale based on the logging accuracy or the accuracy required for later simulations. The Weibull distribution is used for parameter assignment, including: using the values ​​in the seismic grid as the scale parameter X in the Weibull distribution, using the logging accuracy as the dimension of the Weibull distribution matrix, and using the heterogeneity parameter m as the shape parameter in the Weibull distribution to assign values ​​to the entire reservoir, thereby quantitatively evaluating the reservoir heterogeneity.

12. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 11, characterized in that, In step S53, the Gini coefficient ranges from 0 to 1.

0. The correspondence between the Gini coefficient and the degree of parameter heterogeneity includes: When 0 ≤ Gini coefficient ≤ 0.2, the degree of heterogeneity is extremely weak; When 0.2 < Gini coefficient ≤ 0.3, the degree of heterogeneity is relatively weak; When 0.3 < Gini coefficient ≤ 0.4, the degree of heterogeneity is moderate; When 0.4 < Gini coefficient ≤ 0.5, the degree of heterogeneity is relatively strong; When 0.5 < Gini coefficient ≤ 1.0, the degree of heterogeneity is extremely strong.

13. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 12, characterized in that, Step S54 includes: If any parameter has a Gini coefficient less than 0.2, then take m0 = 6 based on the smallest Gini coefficient; If all the Gini coefficients of the parameters are greater than 0.2, then take m0 = 1.5 based on the largest Gini coefficient; The non-homogeneous parameters of the remaining parameters are determined by the ratio of the Gini coefficient corresponding to m0 to the Gini coefficients of other parameters.

14. The method for quantitative evaluation and analysis of reservoir heterogeneity as described in claim 11, characterized in that, Step S56 includes: Establish the Weibull probability distribution function Where X is the parameter value corresponding to each seismic scale grid, x is the assigned parameter, and m is the non-homogeneous parameter; The programmatically implements the Weibull probability distribution function, performing a sampling operation for each calculation. Each sampling yields a set of grid scales that can be further subdivided according to the logging or simulation requirements. The sampled data is input into the rock mechanics simulator to obtain the rock stress-strain curve under the data conditions. It is then compared with the core stress-strain curve of the corresponding layer obtained in the experiment. If the characteristics of the rock stress-strain curve under the sampled data conditions are significantly different from those of the core stress-strain curve under the experimental conditions, then after multiple samplings, its mechanical properties are found to be consistent with those of the real core.

Images