A carbonate fracture classification detection method

Abstract

The application discloses a carbonate rock fracture classification detection method, and relates to the technical field of oil and gas reservoir exploration, and has the technical scheme as follows: in view of the problem that the complex characteristics of a carbonate fracture wave field affect fracture detection effect, attribute analysis is performed on original seismic data of a target area, the fracture is classified into a strong dissolution area and a weak and non-dissolution area, a tensor thinning attribute body is used as a fracture prediction result of the strong dissolution area, and a conventional coherent fracture detection result is used as a fracture prediction result of the weak and non-dissolution area. Through the classification detection technology, the influence of the string-bead anomaly is effectively eliminated, strong dissolution fracture detection results are formed, and the conventional medium and small scale fractures with phase axis disturbance and fault representation are reserved, so that the two are organically fused in space, and the carbonate rock fracture detection precision is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil and gas reservoir exploration, and more particularly to a carbonate rock fracture classification detection method. BACKGROUND

[0002] The fault-dissolved body oil and gas reservoir is a new type of oil and gas reservoir proposed in China in recent years, which is formed around a large strike-slip fault zone after the carbonate rock stratum is subjected to multi-stage tectonic deformation and karstification. Geological research finds that the fault-dissolved body oil and gas reservoirs produced on the fractures in different stress backgrounds are affected by differences in tectonic stress, direction, size, smoothing distance, dissolution degree, etc. in the plane and vertical direction, and have the characteristics of being distributed along the strike-slip fault in the plane and being developed in the "depth" along the fault zone in space.

[0003] For the fault-dissolved body oil and gas reservoir, the carbonate rock fracture is not only the main channel for oil and gas charging and adjustment, but also the favorable oil and gas storage space. Therefore, it is very important to identify the carbonate rock fracture, and how to accurately depict the existing state of the fracture underground is a difficult point in the key.

[0004] At present, there are mainly two kinds of post-stack prediction methods for carbonate rock fractures. One is to use gradient structure tensor attributes to identify carbonate rock fault-dissolved bodies in the journal articles "Research and Application of Carbonate Rock Fault-dissolved Body Depiction Technology in Tahe Oilfield", "Seismic Response Characteristics and Description Technology of Fault-dissolved Body Reservoir in Shunbei Oilfield" and "Spatial Carving and Quantitative Description Technology of Ultra-deep Fault-dissolved Body Reservoir in Shunbei Area". This kind of method has obvious amplification in space due to the large width of the tensor attribute, and it is difficult to identify the position of high-quality reservoirs on the seismic section. The other is to use correlation analysis to identify fractures in the journal articles "Review and Research Progress of Fault Identification Method" and "Identification of Fault-dissolved Body in Middle-lower Ordovician in Yubei Area of Tarim Basin". This kind of method can obtain good prediction effect in the plane, but it is greatly affected by the "bead" anomaly, and the attribute anomaly detected by the coherence analysis technology is mostly the "bead" strong anomaly boundary, which cannot represent the real fracture.

[0005] Based on the problem that the complex characteristics of carbonate fracture wave field affect the detection effect of the fracture, the present application aims to provide a carbonate rock fracture classification detection method to eliminate the fracture false image caused by the "bead" anomaly in the strong dissolution area and retain the conventional detection fracture in the weak dissolution area, so as to improve the detection accuracy of the carbonate rock fracture. SUMMARY

[0006] To address the shortcomings of existing technologies, this invention aims to provide a method for classifying and detecting carbonate rock fractures. This method involves analyzing the properties of the original seismic data volume of the target area to classify fractures into strongly dissolved zones and weakly or non-dissolved zones. Tensor-thinned property volumes are used as the predicted fracture results for strongly dissolved zones, while conventional coherent fracture detection results are used as the predicted fracture results for weakly or non-dissolved zones. This invention not only effectively eliminates the influence of beaded anomalies, resulting in strong dissolved fracture detection, but also preserves in-phase axis perturbations and conventional small-to-medium-scale fractures characterized by misalignment. The two are spatially integrated, effectively improving the accuracy of carbonate rock fracture detection.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for classifying and detecting fractures in carbonate rocks, comprising the following steps:

[0008] S1. Obtain the gradient structure tensor matrix T based on the three-dimensional seismic data of the sampling area.

[0009] S2. Thinning the gradient structure tensor to obtain the tensor thinning attribute volume F1. The specific steps are as follows:

[0010] S21. Obtain all eigenvalues ​​of the gradient structure tensor matrix T, and at a certain eigenvalue λ... i Based on this, all sampling points are scanned to obtain the set of sampling points G(x) centered on the feature value.

[0011] S22. Calculate the azimuth angle for each sampling point. and tilt angle θ k All gradient structure tensor values

[0012] for The total number of internal data, For a given value.

[0013] θ k ∈θ s θ s ={θ min ,θ min +d θ ,θ min +2d θ ,…,θ max}, θ min The minimum dip angle θ represents the distribution of faults in the sampling area. max The maximum dip angle of the fault distribution in the sampling area. θ num For θ s The total number of data within θ num Given value.

[0014] S23, calculate integral value c(x) of azimuth and dip of each sampling point, and the calculation formula of c(x) is as follows:

[0015]

[0016] S24, only the maximum gradient structure tensor value of each sampling point is reserved as m(x).

[0017]

[0018] S25, the thinning gradient structure tensor value h(x) of each sampling point is calculated to form a tensor thinning attribute body F1, and the calculation formula of h(x) is as follows:

[0019]

[0020] In the formula, Where N is the total number of all dip and azimuth combinations,

[0021] S3, according to the three-dimensional seismic data of the sampling area, coherent fracture detection is carried out to form a coherent attribute body F2.

[0022] S4, given a threshold value A, when the absolute value of λ i in S21 is greater than A, then the sampling area is a strong tensor development area, otherwise it is a weak or no dissolution area. The fracture prediction result F3 of the strong tensor development area is the tensor thinning attribute body F1, and the fracture prediction result F3 of the weak or no dissolution area is the coherent attribute body F2.

[0023] The application is further provided that when the maximum dip θ max of the fracture distribution of the sampling area cannot be measured, θ max is set to 85°.

[0024] The application is further provided that the specific steps of S1 are:

[0025] S11, determine the sampling points of the sampling area, and calculate the gradient vector g of each sampling point according to the three-dimensional seismic data of the sampling area:

[0026]

[0027] In the formula, x, y and z are respectively the line, trace direction and time direction of the two-way travel of the three-dimensional seismic body. g1, g2 and g3 are respectively the partial derivatives of g along the x, y and z directions, and the partial derivative steps used in the calculation process of g1, g2 and g3 are all given values.

[0028] S12, construct a gradient structure tensor matrix T:

[0029]

[0030] The application is further configured that the partial derivative step length used in the calculation of g1, g2 and g3 is the same.

[0031] The application is further configured that the partial derivative step length used in the calculation of g1, g2 and g3 is 300-600 meters.

[0032] The application is further configured that the gradient structure tensor matrix T is a real symmetric matrix, and the eigenvalue λ satisfies λ1≥λ2≥λ3>0, and λ2 is selected in S21. i The application is further configured that the gradient structure tensor matrix T is a real symmetric matrix, and the eigenvalue λ satisfies λ1≥λ2≥λ3>0, and λ2 is selected in S21.

[0033] The application is further configured that S22 calculates all gradient structure tensor values of each sampling point along the azimuth and the dip angle θ k by using the coherence cube technique.

[0034] The application is further configured that S22 specifically comprises the following steps for each sampling point:

[0035] S221, fixing the azimuth and fixing the dip angle θ k , and obtaining the three-dimensional seismic data of the sampling point.

[0036] S222, selecting adjacent J channels of data for the three-dimensional seismic data of the sampling point, collecting N data for each channel, and obtaining N×J sub-data d to form a seismic sub-cube matrix U1.

[0037]

[0038] S223, calculating the covariance matrix W1, and calculating all non-zero eigenvalues λ ji of W1.

[0039] S224, calculating the gradient structure tensor value of each sampling point at the azimuth and the dip angle θ k .

[0040]

[0041] In the formula, λ j1 represents the maximum eigenvalue of the covariance matrix W1.

[0042] S225, keeping the azimuth unchanged, changing the dip angle θ k , and obtaining the three-dimensional seismic data of the sampling point. Repeating S222-S224 to obtain the gradient structure tensor value of the sampling point at the azimuth and different dip angles θ kall gradient structure tensor values of

[0043] S226, changing the azimuth angle Obtain the three-dimensional seismic data of the sampling points. Repeat S222-S225 to obtain different azimuth angles different dip angles θ k all gradient structure tensor values of constitute

[0044] The application is further provided that: the threshold value in S4 is calibrated by drilling time, and the lower limit of the tensor attribute of the reservoir drilling is used as the threshold value of the abnormal body. In the drilling time calibration, the tensor attribute of the reservoir section is determined by the point-by-point arithmetic mean value, and the drilling time of the reservoir section is determined by the arithmetic mean value of the drilling time of the smooth section.

[0045] The application is further provided that: the coherent fracture detection of S3 is specifically that, the azimuth angle and the dip angle are fixed, the three-dimensional seismic data of all sampling points in the sampling area are obtained, and the following steps are performed for each sampling point:

[0046] S31, select adjacent Q channel data, M data per channel, QxM sub-data d, to form a seismic sub-body matrix U2.

[0047]

[0048] S32, obtain the covariance matrix W2, and all non-zero eigenvalues λ of W2 qi .

[0049] S33, calculate the coherence value C of each sampling point to form a coherent attribute body F2.

[0050]

[0051] In the formula, λ q1 represents the maximum eigenvalue of the covariance matrix W2.

[0052] The application is further provided that: the carbonate rock fracture classification detection method is applied in oil and gas exploration, and is used to determine the fracture position, number, fracture direction, dip angle, fracture surface shape and characteristics of the carbonate rock.

[0053] In summary, the present application has the following beneficial effects compared to the prior art: the present application aims to solve the problem of the influence of complex features of carbonate fracture wave field on fracture detection effect, performs attribute analysis on the original seismic data body of the target area, classifies the fractures into strong dissolution area and weak and no dissolution area, uses the tensor thinning attribute body as the fracture prediction result of the strong dissolution area, and uses the conventional coherent fracture detection result as the fracture prediction result of the weak and no dissolution area. Through this classification detection technology, not only the influence of the string bead anomaly is effectively eliminated to form the strong dissolution fracture detection result, but also the conventional small-scale fracture with phase axis disturbance and fault representation is retained, and the two are organically integrated in space, effectively improving the carbonate fracture detection precision. BRIEF DESCRIPTION OF DRAWINGS

[0054] Figure 1 is a carbonate fracture seismic profile of the target area.

[0055] Figure 2 is a flowchart of the embodiment.

[0056] Figure 3 is a fracture structure tensor profile of the target area.

[0057] Figure 4 is a fracture structure tensor thinning profile of the target area.

[0058] Figure 5 is a conventional coherent attribute profile of the target area.

[0059] Figure 6 is a conventional coherent plan of the target area.

[0060] Figure 7 is a fusion fracture profile of the target area.

[0061] Figure 8 is a fusion fracture plan of the target area. DETAILED DESCRIPTION

[0062] The technical solutions of the present application will be clearly described below in combination with the description of the drawings. Obviously, the described embodiments are not all the embodiments of the present application, and all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0063] EMBODIMENT

[0064] The present application is illustrated by taking the carbonate fracture detection of a certain area of Tahe Oilfield as an example, Figure 1 that is, the fracture seismic profile of the area.

[0065] For example, Figure 2As shown, it is a flowchart of the embodiment of the present application, a carbonate rock fracture classification detection method, applied in oil and gas exploration, used to determine the fracture position, number, direction, dip angle, shape and characteristics of the fracture surface of the carbonate rock, which comprises the following steps:

[0066] S1, as shown in the figure, Figure 3 According to the three-dimensional seismic data of the sampling area, the gradient structure tensor matrix T is obtained.

[0067] S11, determine the sampling points in the sampling area, calculate the gradient vector g at each sampling point according to the three-dimensional seismic data of the sampling area:

[0068]

[0069] In the formula, x, y, z are the line, trace direction and two-way travel time direction of the three-dimensional seismic body respectively. g1, g2, g3 are the partial derivatives of g along x, y, z directions respectively, and the partial derivative step length used in the calculation process of g1, g2, g3 is a given value. The partial derivative step length used in the calculation process of g1, g2, g3 is 300-600 meters, and in this embodiment, the partial derivative step length used in the calculation process of g1, g2, g3 is 600 meters.

[0070] S12, construct the gradient structure tensor matrix T:

[0071]

[0072] S2, as shown in the figure, Figure 4 Thin the gradient structure tensor to obtain the tensor thinning attribute body F1, the specific steps are:

[0073] S21, obtain all eigenvalues of the gradient structure tensor matrix T, and on the basis of a certain eigenvalue λ i , scan all sampling points to obtain a sampling point set G(x) centered on the eigenvalue. The gradient structure tensor matrix T is a real symmetric matrix, the eigenvalue λ satisfies λ1≥λ2≥λ3>0, λ1high value anomaly mainly reflects the horizontal event, and the horizontal event is the part with the most dramatic energy change; λ3 has low signal-to-noise ratio; and λ2 mainly reflects the chaotic reflection characteristics, and has ideal effect on the description of the fracture dissolution body, which coincides with the profile characteristics of the fracture dissolution body. Therefore, in this embodiment, λ i λ2 is selected.

[0074] S22, calculate all gradient structure tensor values of each sampling point along the azimuth k and dip angle θ

[0075] ​Total number of data within θ is given value.

[0076] θ k ∈θ s , θ s = {θ min , θ min + d θ , θ min + 2d θ , …, θ max}, θ min represents the minimum dip angle of the fault distribution in the sampling area, θ max is the maximum dip angle of the fault distribution in the sampling area, θ num is the total number of data within θ s , θ num is given value.

[0077] When the maximum dip angle θ max of the fault distribution in the sampling area cannot be measured, θ max is set to 85°.

[0078] S23, calculate the integral value c(x) of the structural layer azimuth and dip angle at each sampling point, and the calculation formula of c(x) is as follows:

[0079]

[0080] S24, for each sampling point, only the maximum gradient structural tensor value is retained as m(x).

[0081]

[0082] S25, calculate the thinning gradient structural tensor value h(x) of each sampling point, and form the tensor thinning attribute body F1, and the calculation formula of h(x) is as follows:

[0083]

[0084] In the formula, where N is the total number of all dip angle and azimuth combinations,

[0085] S3, as shown in Figure 5 , Figure 6 , according to the three-dimensional seismic data of the sampling area, coherent fault detection is carried out to form the coherent attribute body F2.

[0086] S4, as shown in Figure 7 , Figure 8 , a threshold value A is given, and when λ iIf the absolute value of the difference is greater than A, the sampling area is a strong tensor development area, otherwise it is a weak or non-dissolution area. The fracture prediction result F3 of the strong tensor development area is a tensor thinning attribute body F1, and the fracture prediction result F3 of the weak or non-dissolution area is a coherent attribute body F2. In this embodiment, the threshold value is calibrated by drilling time, and the lower limit of the tensor attribute encountered by the reservoir drilling is taken as the threshold value of the anomaly body. In the drilling time calibration, the tensor attribute of the reservoir section is determined by the point-by-point arithmetic mean value, and the drilling time of the reservoir section is determined by the arithmetic mean value of the drilling time of the smooth section.

[0087] For drilling targeting the Ordovician fracture-vug reservoir, emptying and loss will occur when drilling the fracture-vug development position, and generally no logging curve can be obtained. The drilling time curve of the sidetrack horizontal well can be used to calibrate the fracture lateral crushing range. In order to achieve the mutual combination between the reservoir and the attribute, the drilling time of multiple wells is calibrated. In this embodiment, the drilling time of 15 wells is calibrated, and the tensor attribute is extracted at a sampling rate of 0.125 m. Through the calibration of the tensor attribute and the drilling time curve, when the drilling time is less than 20 min / m, the abnormal value of the tensor attribute is greater than 9.

[0088] Specifically, S22 calculates all gradient structural tensor values of each sampling point along the azimuth and the dip angle θ k of the sampling point. Specifically, the following steps are performed for each sampling point:

[0089] S221, fix the azimuth and the dip angle θ k , and obtain the three-dimensional seismic data of the sampling point.

[0090] S222, for the three-dimensional seismic data of the sampling point, select adjacent J channel data, collect N data for each channel, and obtain N×J sub-data d to form a seismic sub-body matrix U1.

[0091]

[0092] S223, calculate the covariance matrix W1, and all non-zero eigenvalues λ ji of W1.

[0093] S224, calculate the gradient structural tensor value of each sampling point at the azimuth k and the dip angle θ

[0094]

[0095] In the formula, λ j1 represents the maximum eigenvalue of the covariance matrix W1.

[0096] S225, keep the azimuth unchanged, change the dip angle θ k , obtain the 3D seismic data of the sampling point. Repeat S222-S224 to obtain different azimuth angles unchanged, change the dip angle θ k of all gradient structure tensor values

[0097] S226, change the azimuth angle obtain the 3D seismic data of the sampling point. Repeat S222-S225 to obtain different azimuth angles unchanged, change the dip angle θ k of all gradient structure tensor values consisting of

[0098] Specifically, the coherent fracture detection of S3 specifically consists in fixing the azimuth angle and the dip angle, obtaining the 3D seismic data of all sampling points in the sampling area, and performing the following steps on each sampling point:

[0099] S31, select adjacent Q channel data, M data per channel, a total of QxM sub-data d, to form a seismic sub-body matrix U2.

[0100]

[0101] S32, calculate the covariance matrix W2, and all non-zero eigenvalues λ qi of W2 are calculated.

[0102] S33, calculate the coherence value C of each sampling point to form a coherence attribute body F2.

[0103]

[0104] where λ q1 represents the maximum eigenvalue of the covariance matrix W2.

[0105] In summary, the method aims to solve the problem of the influence of the complex characteristics of carbonate fracture wave field on fracture detection effect. The original seismic data body of the target area is analyzed by attribute analysis, and the fractures are classified into strong dissolution area and weak and no dissolution area. The tensor thinning attribute body is used as the fracture prediction result of the strong dissolution area, and the conventional coherent fracture detection result is used as the fracture prediction result of the weak and no dissolution area. Through this classification detection technology, not only the influence of the string bead anomaly is effectively eliminated to form the strong dissolution fracture detection result, but also the conventional small-scale fractures with phase axis disturbance and fault representation are retained. The two are organically integrated in space, effectively improving the carbonate fracture detection precision.

[0106] The above merely provides the preferred embodiments of the present application, and is not used to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the principles and technical scope of the present application shall fall into the scope of the present application.

Claims

1. A method for detecting and classifying carbonate rock fractures, characterized by, The method comprises the following steps: S1, obtaining a gradient structure tensor matrix T according to three-dimensional seismic data of a sampling area; S2, thinning the gradient structure tensor to obtain a tensor thinning attribute volume F1, and the specific steps are as follows: S21, obtain all eigenvalues of the gradient structure tensor matrix T, on the basis of a certain eigenvalue λ i , scan all sampling points to obtain a sampling point set centered on the eigenvalue ; S22, compute all gradient structure tensor values for each sample point along azimuth and dip ;​ , , ; for The total number of internal data, For a given value; , The minimum dip angle representing the fault distribution in the sampling area. The maximum dip angle of the fault distribution in the sampling area. ; for The total number of data within, Given value; S23, calculating an integral value c(x) of each sampling point along a structural layer azimuth and dip, and the calculation formula of c(x) is as follows: S24, for each sampling point, only keep the largest gradient structure tensor value as ; S25, calculating the thinning gradient structure tensor value of each sampling point , the tensor thinning attribute body F1 is constituted, The calculation formula is as follows: wherein wherein N is the total number of all combinations of dip and azimuth angles, ; S3, detecting a coherent fracture according to three-dimensional seismic data of the sampling area to form a coherent attribute volume F2; S4. Given a threshold value A, when λ in S21 i When the absolute value of is greater than A, the sampling area is a strong tensor development area; otherwise, it is a weak or non-dissolvable area. The fracture prediction result F3 of the strong tensor development area is the tensor thinning attribute body F1, and the fracture prediction result F3 of the weak or non-dissolvable area is the coherent attribute body F2.

2. The method for detecting and classifying carbonate rock fractures according to claim 1, characterized in that: When the maximum dip of the fault distribution in the sampled area is greater than 85° When the maximum dip of the fault distribution in the sampled area is greater than 85° When the maximum dip of the fault distribution in the sampled area is greater than 85° 3. The method according to claim 1 or 2, characterized in that, The specific steps of S1 are as follows: S11, determining sampling points in the sampling area, and calculating a gradient vector g of each sampling point according to three-dimensional seismic data of the sampling area; In the formula, x, y, z are respectively the line, trace direction and time direction of double travel of three-dimensional seismic body; respectively Partial derivative along x, y, z direction, The partial derivative step length used in the calculation process is a given value; S12, constructing the gradient structure tensor matrix T; 。 4. The method according to claim 3, characterized in that: The same partial derivative step size is used in the calculation process.

5. The method for detecting and classifying carbonate fractures according to claim 4, characterized in that: The partial derivative step length used in the calculation was 300-600 m.

6. The method according to claim 3, characterized in that: The gradient structure tensor matrix T is a real symmetric matrix, the eigenvalue λ satisfies λ1≥λ2≥λ3>0, λ i λ2is selected.

7. The method according to claim 1 or 2, characterized in that: S22 calculates all gradient structure tensor values along azimuth and dip for each sample point using the coherent structures technique .

8. The method according to claim 7, wherein, S22 specifically comprises the following steps for each sampling point: S221, fixed azimuth , fixed dip angle , acquiring the 3D seismic data of the sampling points; S222. For the 3D seismic data of the sampling points, select adjacent J traces, and collect N data points for each trace, for a total of N. Individual data points d form the seismic sub-matrix U1; S223, the covariance matrix W1 is calculated, = W1 and all non-zero eigenvalues λ of W1 are calculated ji ; S224, calculate the gradient structure tensor value at the azimuth and inclination of each sample point ; wherein denotes the largest eigenvalue of the covariance matrix denotes the largest eigenvalue of the covariance matrix S225, keep azimuth unchanged, change dip , obtain the sampling point three-dimensional seismic data; S222-S224 are repeated to obtain the azimuth angle below, different inclination angles of all gradient structure tensor values ; S226, change the azimuth , obtaining the sampling point three-dimensional seismic data; S222-S225 are repeated to obtain different azimuthal angles , different inclination angles of all gradient structure tensor values , constituting .

9. The method of claim 1, wherein: In S4, the threshold value is calibrated by drilling time, and a lower limit of the tensor attribute of a reservoir drilled is taken as the threshold value of the abnormal volume; in the calibration of the drilling time, the tensor attribute of the reservoir section is determined by a point-by-point arithmetic average value, and the drilling time of the reservoir section is determined by an arithmetic average value of the drilling time of a smooth section.

10. The method for classifying and detecting fractures in carbonate rocks according to claim 1, characterized in that, The coherent fracture detection of S3 specifically comprises the following steps for each sampling point: S31, selecting adjacent Q channel data, M data in each channel, a total of sub data d, to form a seismic sub-body matrix U2; ； S32. Calculate the covariance matrix , = and find All non-zero eigenvalues ​​λ qi ; S33, calculating a coherent value C of each sampling point to form the coherent attribute volume F2; wherein denotes the largest eigenvalue of the covariance matrix denotes the largest eigenvalue of the covariance matrix 11. A method for detecting and classifying carbonate rock fractures, the method comprising: The carbonate rock fracture classification and detection method according to any one of claims 1-10 is applied in oil and gas exploration, and is used to determine the fracture position, number, fracture direction, dip, fracture surface shape and characteristics of the carbonate rock. ​

Images