Seismic OVT gather data fault detection method

Abstract

The invention provides a seismic OVT gather data fault detection method, which belongs to the technical field of petroleum and natural gas development, and comprises the following steps: obtaining seismic observation system information data; geological knowledge of the research area is obtained; carrying out azimuth sorting to obtain N narrow azimuth seismic trace sets; carrying out superposition processing on the narrow azimuth seismic gather to obtain azimuth modal data; carrying out seismic geometric attribute calculation on each piece of azimuth data to obtain geometric features; denoising the geometric features to obtain a fault model; and performing data fusion on the fault model, and outputting a result. After the original seismic OVT gather is subjected to azimuth decomposition, denoising is carried out by adopting different denoising parameters, noise obviously influenced by the azimuth is removed, and the signal-to-noise ratio level is improved; after the original seismic OVT gather is subjected to azimuth decomposition, the weight of the low signal-to-noise ratio azimuth modal data can be reduced by setting the prior weight of the azimuth, so that the signal-to-noise ratio feature is improved.

Description

Technical Field

[0001] The invention belongs to the technical field of oil and gas development, and in particular relates to a fault detection method for seismic OVT gather data. Background Art

[0002] Fault detection is one of the important links in the process of oil and gas exploration and development. Conventional fault detection methods generally only use post-stack seismic data, and realize fault detection by calculating attributes such as coherence or curvature of post-stack seismic data. The fault detection results are greatly affected by the quality of post-stack seismic data.

[0003] The spatial distribution characteristics of faults are controlled by the geostress characteristics and tectonic system of the study area, and conventional fault detection methods cannot use the geological information contained in the faults. The shortcomings of conventional fault methods often lead to low signal-to-noise ratio and resolution of fault detection results, which are difficult to meet the actual needs of fine exploration and development of oil reservoirs.

[0004] The Chinese invention patent application document CN113296152A discloses a fault detection method and device, wherein the method comprises: obtaining a fault sample data set; constructing an initial convolutional neural network model for fault detection based on the fault sample data set and a deep learning algorithm; eliminating the influence of specified features and noise in the initial convolutional neural network model based on a self-attention gate mechanism, and establishing a convolutional neural network model for fault detection; obtaining real-time seismic data; inputting the real-time seismic data into the convolutional neural network model to obtain a fault detection result. The method realizes fault detection of seismic data by establishing a convolutional neural network model. The patent adopts a neural network method to establish a mapping relationship between known fault samples and seismic data, thereby realizing fault detection. The premise of the method is that the fault sample itself is accurate, and the seismic data features and the fault sample are matched. In fact, the fault sample is the result calculated by manual interpretation or a certain fault detection algorithm under specified seismic data conditions, which results in data dependence and obvious uncertainty in the fault sample. Due to the uncertainty and data dependence of fault samples, using fault sample data sets to establish labels and then using deep learning algorithms to perform fault detection is logically a circular argument. It can be seen that this method is difficult to improve the accuracy of fault detection. Summary of the invention

[0005] In view of the problems existing in the above-mentioned prior art, the present invention provides a method for fault detection of seismic OVT gather data. The technical problem to be solved by the present invention is how to overcome the problems of low signal-to-noise ratio and resolution of fault detection results in the prior art, thereby improving the accuracy of fault detection.

[0006] In order to solve the above technical problems, the present invention provides a method for fault detection of seismic OVT gather data, comprising the following steps:

[0007] Step S1: using seismic OVT gather data to obtain seismic observation system information data;

[0008] Step S2: Obtain geological knowledge of the study area where the seismic OVT gather data are located;

[0009] Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers;

[0010] Step S4: stacking each narrow azimuth seismic gather to convert the seismic OVT gather data into N azimuth modal data;

[0011] Step S5: Calculate the seismic geometric attributes for each azimuth modal data to obtain the geometric features corresponding to each azimuth modal data;

[0012] Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth data;

[0013] Step S7: performing data fusion on the fault models corresponding to the N azimuth modal data, and using the result of data fusion as the output result of fault detection.

[0014] Furthermore, in step S1, the seismic observation system information data includes azimuth, coverage times and aspect ratio.

[0015] Furthermore, in step S2, the geological knowledge includes geostress information and structural information.

[0016] Furthermore, in step S2, the geostress information includes information on the directions of maximum and minimum horizontal principal stresses, and the structural information includes structural style characteristics of the target geological body.

[0017] Furthermore, in step S3, the requirement for azimuth sorting is that each narrow azimuth seismic gather has a geostress characteristic that is nearly parallel or nearly perpendicular to the direction of the maximum horizontal principal stress; or, each narrow azimuth seismic gather has a structural characteristic that is parallel or perpendicular to the structural strike.

[0018] Furthermore, in step S5, the seismic geometric attributes include coherence attributes, curvature attributes, dip attributes and ant body attributes.

[0019] Furthermore, in step S6, different denoising parameters are used to process geometric features corresponding to modal data in different orientations.

[0020] Furthermore, in step S7, the plane features and cross-sectional features of each fault model in step S6 at the target layer are extracted, and then data fusion is performed on the plane features and cross-sectional features.

[0021] Furthermore, in step S7, data fusion is performed on the fault models corresponding to the N azimuthal modal data according to the linear joint priori Bayesian fusion algorithm.

[0022] Furthermore, in step S7, the steps of performing data fusion on the fault model according to the linear joint priori Bayesian fusion algorithm are as follows:

[0023] Step H1: according to the plane features and profile features of the fault model of each azimuth modal data, combined with the geological knowledge in step S2, the prior weight of each azimuth is set, and the prior weight is set according to the signal-to-noise ratio difference of the plane and profile features of the fault model corresponding to the modal data of different azimuths; the weight of the fault model with a lower signal-to-noise ratio is set to a low value, and the weight of the fault model with a lower signal-to-noise ratio is set to a high value;

[0024] Step H2: normalize and sum the seismic geometric attributes corresponding to each fault model to obtain the likelihood function of each orientation;

[0025] Step H3: Calculate the posterior probability using the Bayesian formula and likelihood function.

[0026] Furthermore, in step H3, the calculation formula of the posterior probability is:

[0027]

[0028] Where: p(y|x (1) ,x (2) ,x (3) …x (N) ) is the posterior probability of the fault model data after fusion of N azimuthal modal data, x (1) , x (2) , x (3) ...x (N) represents the azimuthal modal data, y represents the fault model, n represents the count of the azimuthal modal data, ω n is the prior weight of the nth azimuth modal data, ρ(y|x (n) ) is the likelihood function of the nth orientation mode.

[0029] A fault detection method for seismic OVT gather data of the present invention has the following advantages: 1) after the original seismic OVT gather is azimuthally decomposed, by adopting different denoising parameters for the geometric features corresponding to each orientation, the noise significantly affected by the orientation can be effectively removed, thereby improving the signal-to-noise ratio level of the fault detection result after fusion; 2) after the original seismic OVT gather is azimuthally decomposed, by setting the prior weight of the orientation when fusion of decision data, the weight of the azimuth modal data with low signal-to-noise ratio can be reduced, thereby improving the signal-to-noise ratio characteristics of the fault detection result after fusion; 3) compared with the conventional method of only using post-stack data or only optimizing a certain azimuth modal data to carry out fault detection, the decision fusion of fault features corresponding to different orientations can effectively improve the resolution and signal-to-noise ratio of the fault detection result. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 The present invention discloses a flow chart of a fault detection method for seismic OVT gather data.

[0031] Figure 2 It is a schematic diagram of seismic OVT gather data of a seismic OVT gather data fault detection method of the present invention.

[0032] Figure 3 It is a schematic diagram of seismic profiles corresponding to multiple azimuth modal data of a fault detection method for seismic OVT gather data of the present invention, Figure 3 (a) is a schematic diagram of the seismic section with an azimuth of 42 degrees; Figure 3 (b) is a schematic diagram of the seismic section with an azimuth of 87 degrees; Figure 3 (c) is a schematic diagram of the seismic section with an azimuth of 132 degrees; Figure 3 (d) in the figure is a schematic diagram of the seismic profile with an azimuth of 177 degrees.

[0033] Figure 4 It is a plane diagram of geometric features corresponding to multiple azimuth modal data of a fault detection method for seismic OVT gather data of the present invention; Figure 4 (a) is a plan view of the geometric feature with an orientation of 42 degrees; Figure 4 (b) is a plan view of the geometric feature with an orientation of 87 degrees; Figure 4 (c) is a plan view of the geometric feature with an orientation of 132 degrees; Figure 4 (d) in the figure is a plan view of the geometric feature with an orientation of 177 degrees.

[0034] Figure 5 It is a comparison diagram of fault detection results of a seismic OVT gather data fault detection method based on the present invention and a conventional method; Figure 5 (a) is a schematic diagram of plane characteristics obtained by using a fault detection method of seismic OVT gather data of the present invention; Figure 5(b) is a schematic diagram of the plane features obtained by tomographic detection using conventional methods. DETAILED DESCRIPTION

[0035] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0036] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0038] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected or indirectly connected through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0039] In order to better understand the purpose, structure and function of the present invention, a seismic OVT gather data fault detection method of the present invention is further described in detail below in conjunction with the accompanying drawings.

[0040] Embodiment 1:

[0041] Figure 1 The present invention shows a fault detection method for seismic OVT gather data, comprising the following steps:

[0042] Step S1: using seismic OVT gather data to obtain seismic observation system information data;

[0043] Step S2: Obtain geological knowledge of the study area where the OVT gather data are located;

[0044] Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers;

[0045] Step S4: stacking each narrow azimuth seismic gather in step S3 to convert the seismic OVT gather data into N azimuth modal data;

[0046] Step S5: Calculate the seismic geometric attributes for each azimuth modal data to obtain the geometric features corresponding to each azimuth modal data;

[0047] Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth modal data;

[0048] Step S7: performing data fusion on the fault models corresponding to the N azimuth modal data, and using the result of data fusion as the output result of fault detection.

[0049] Embodiment 2:

[0050] Figure 1 The present invention shows a fault detection method for seismic OVT gather data, comprising the following steps:

[0051] Step S1: using seismic OVT gather data to obtain seismic observation system information data;

[0052] Step S2: Obtain geological knowledge of the study area where the OVT gather data are located;

[0053] Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers;

[0054] Step S4: stacking each narrow azimuth seismic gather in step S3 to convert the seismic OVT gather data into N azimuth modal data;

[0055] Step S5: Calculate the seismic geometric attributes for each azimuth modal data to obtain the geometric features corresponding to each azimuth modal data;

[0056] Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth modal data;

[0057] Step S7: performing data fusion on the fault models corresponding to the N azimuth modal data, and using the result of data fusion as the output result of fault detection.

[0058] The difference between this embodiment and the first embodiment is that:

[0059] In step S1, the seismic observation system information data includes azimuth, coverage times and aspect ratio.

[0060] In step S2, the geological knowledge includes ground stress information and structural information.

[0061] In step S3, the requirement for azimuth sorting is that each narrow azimuth seismic gather has a geostress feature that is nearly parallel or nearly perpendicular to the direction of the maximum horizontal principal stress; or, each narrow azimuth seismic gather has a structural feature that is parallel or perpendicular to the structural strike.

[0062] In step S5, the seismic geometric attributes include coherence attributes, curvature attributes, dip attributes and ant body attributes.

[0063] In step S6, the denoising parameters used in the denoising process are related to the orientation. The denoising process is used to filter out coherent noise that is highly correlated with the orientation. Since the extension direction of the noise varies significantly in different orientations, different denoising parameters need to be selected for different orientations. Practice shows that it is necessary to eliminate the noise in the vertical direction of the azimuth angle corresponding to each orientation.

[0064] In step S7, the plane features and cross-sectional features of each fault model in step S6 at the target layer are extracted, and then data fusion is performed on the plane features and cross-sectional features.

[0065] In the step S7, data fusion is performed on the fault models corresponding to the N azimuthal modal data according to the linear joint priori Bayesian fusion algorithm.

[0066] Embodiment 3:

[0067] Figure 1 The present invention shows a fault detection method for seismic OVT gather data, comprising the following steps:

[0068] Step S1: using seismic OVT gather data to obtain seismic observation system information data;

[0069] Step S2: Obtain geological knowledge of the study area where the OVT gather data are located;

[0070] Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers;

[0071] Step S4: stacking each narrow azimuth seismic gather in step S3 to convert the seismic OVT gather data into N azimuth modal data;

[0072] Step S5: Calculate the seismic geometric attributes for each azimuth modal data to obtain the geometric features corresponding to each azimuth modal data;

[0073] Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth modal data;

[0074] Step S7: performing data fusion on the fault models corresponding to the N azimuth modal data, and using the result of data fusion as the output result of fault detection.

[0075] The present embodiment is different from the above embodiments in that:

[0076] In step S7, the steps of performing data fusion on the fault model according to the linear joint priori Bayesian fusion algorithm are as follows:

[0077] Step H1: according to the plane characteristics and cross-sectional characteristics of the fault model of each azimuth data, combined with the geological knowledge in step S2, setting the prior weight of each azimuth;

[0078] Step H2: normalize and sum the seismic geometric attributes corresponding to each fault model to obtain the likelihood function of each orientation;

[0079] Step H3: Using the Bayesian formula, calculate the posterior probability, and use the posterior probability result after fusing the N fault model data as the output result of fault detection. The calculation formula of the posterior probability is:

[0080]

[0081] Where: p(y|x (1) ,x (2) ,x (3) …x (N) ) is the posterior probability of the fault model data after fusion of N azimuthal modal data, x (1) , x (2) , x (3) ...x (N) represents the azimuthal modal data, y represents the fault model, n represents the count of the azimuthal modal data, ω n is the prior weight of the nth azimuth modal data, ρ(y|x (n) ) is the likelihood function of the nth orientation mode.

[0082] Embodiment 4:

[0083] Figure 1 The present invention shows a fault detection method for seismic OVT gather data, comprising the following steps:

[0084] Step S1: Figure 2 As shown, using seismic OVT gather data, the seismic observation system information data is obtained;

[0085] Step S2: Obtain geological knowledge of the study area where the seismic OVT gather data are located;

[0086] Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers;

[0087] Step S4: stack each narrow azimuth seismic gather in step S3 to convert the seismic OVT gather data into N azimuth modal data, such as Figure 3 As shown, azimuth data of 42 degrees, 87 degrees, 132 degrees, and 177 degrees were obtained;

[0088] Step S5: Calculate the seismic geometric attributes for each azimuth data to obtain the geometric features corresponding to each azimuth data, such as Figure 4 As shown, geometric features with orientations of 42 degrees, 87 degrees, 132 degrees, and 177 degrees were obtained;

[0089] Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth modal data;

[0090] Step S7: Perform data fusion on the fault models corresponding to the N azimuth data, and use the result of data fusion as the output result of fault detection, such as Figure 5 As shown, the planar features of the fault detection results obtained using the present method and the planar features of the fault detection results obtained using conventional technology, and the planar features obtained using the present method have higher resolution and signal-to-noise ratio.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A fault detection method for seismic OVT gather data, characterized in that: The following steps are involved: Step S1: using seismic OVT gather data to obtain seismic observation system information data; Step S2: Obtain geological knowledge of the study area where the seismic OVT gather data are located; Step S3: performing azimuth sorting according to the seismic observation system information data and the geological knowledge to obtain N narrow azimuth seismic gathers; Step S4: stacking each narrow azimuth seismic gather to convert the seismic OVT gather data into N azimuth modal data; Step S5: Calculate the seismic geometric attributes for each azimuth modal data to obtain the geometric features corresponding to each azimuth modal data; Step S6: performing denoising processing on the geometric features to obtain a fault model corresponding to each azimuth data; Step S7: performing data fusion on the fault models corresponding to the N azimuth modal data, and using the result of data fusion as the output result of fault detection.

2. The seismic OVT gather data fault detection method according to claim 1, characterized in that: In step S1, the seismic observation system information data includes azimuth, coverage times and aspect ratio.

3. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S2, the geological knowledge includes ground stress information and structural information.

4. The fault detection method of seismic OVT gather data according to claim 3, characterized in that: In step S2, the geostress information includes the maximum and minimum horizontal principal stress direction information, and the structural information includes the structural style characteristics of the target geological body.

5. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S3, the requirement for azimuth sorting is that each narrow azimuth seismic gather has a geostress feature that is nearly parallel or nearly perpendicular to the direction of the maximum horizontal principal stress; or, each narrow azimuth seismic gather has a structural feature that is parallel or perpendicular to the structural strike.

6. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S5, the seismic geometric attributes include coherence attributes, curvature attributes, dip attributes and ant body attributes.

7. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S6, different denoising parameters are used to process the geometric features corresponding to modal data in different orientations.

8. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S7, the plane features and cross-sectional features of each fault model in step S6 at the target layer are extracted, and then data fusion is performed on the plane features and cross-sectional features.

9. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In the step S7, data fusion is performed on the fault models corresponding to the N azimuthal modal data according to the linear joint priori Bayesian fusion algorithm.

10. The fault detection method of seismic OVT gather data according to claim 1, characterized in that: In step S7, the steps of performing data fusion on the fault model according to the linear joint priori Bayesian fusion algorithm are as follows: Step H1: according to the plane characteristics and cross-sectional characteristics of the fault model of each azimuthal modal data, combined with the geological knowledge in step S2, setting the prior weight of each azimuthal modal; Step H2: normalize and sum the seismic geometric attributes corresponding to each fault model to obtain the likelihood function of each azimuth mode; Step H3: Calculate the posterior probability using the Bayesian formula and likelihood function.

11. The fault detection method of seismic OVT gather data according to claim 10, characterized in that: In step H3, the calculation formula of the posterior probability is: Where: p(y|x (1) ,x (2) ,x (3) …x (N) ) is the posterior probability of the fault model data after fusion of N azimuthal modal data, x (1) , x (2) , x (3) ...x (N) represents the azimuthal modal data, y represents the fault model, n represents the count of the azimuthal modal data, μ n is the prior weight of the nth azimuth modal data, ρ(y|x (n) ) is the likelihood function of the nth orientation mode.

Images