Method for carrying out well leakage early warning by utilizing earthquake prediction result

Abstract

The invention relates to the technical field of geophysical exploration, in particular to a method for performing well leakage early warning by utilizing an earthquake prediction result, which comprises the steps of basic data acquisition, regional geological condition analysis, earthquake data preprocessing, small fracture and fracture-cavity earthquake fine description and shadow relief display of a description result. According to the method, the quality of basic data is improved by introducing anisotropic diffusion and a three-dimensional multi-stage median joint filtering technology, multi-attribute fusion based on adjustable parameters is carried out on pre-stack multi-azimuth frequency attenuation gradient crack prediction and post-stack structural coherent tensor, and finally, a multi-attribute fusion result is spatially displayed by utilizing a shadow relief technology. The precision of small fracture and fracture-cavity body depicting is improved, the spatial forms of fractures and fracture-cavity bodies are displayed more visually, and well position design, well drilling design and well drilling trajectory optimization can be effectively guided.

Description

Technical Field

[0001] The present invention relates to the field of geophysical exploration technology, and in particular to a method for early warning of well leakage using earthquake prediction results. Background Art

[0002] With the continuous advancement of oil and gas exploration and development, seismic exploration technology is playing an increasingly important role in drilling projects. Lost circulation is a common occurrence during oil and gas drilling, posing a serious threat to drilling operations during this period. Traditional methods for warning lost circulation primarily rely on real-time drilling monitoring, comprehensive logging, drilling fluid performance monitoring, and formation pressure assessment. However, these methods often suffer from low lost circulation warning capabilities and incomplete analysis.

[0003] The existing technology has at least the following problems: At present, the well leakage warning technology based on seismic data mainly uses methods such as post-stack coherence, curvature or structural tensor attributes to characterize the fracture body, which is often limited by the quality of seismic data and the lack of anisotropic characteristics of post-stack data, making it difficult to achieve high-precision characterization of small faults and fracture bodies. Existing methods often do not involve the pre-processing link of seismic data and do not consider the anisotropic characteristics of pre-stack. The present invention improves the basic data quality by introducing anisotropic diffusion and three-dimensional multi-level median joint filtering technology, and combines pre-stack multi-directional frequency attenuation gradient fracture prediction and post-stack structural coherence tensor with multi-attribute fusion based on adjustable parameters. Finally, the multi-attribute fusion results are spatially displayed using shadow relief technology, thereby improving the accuracy of small fault and fracture body characterization and guiding well location design, drilling design and drilling trajectory optimization.

[0004] It can be seen that there is still room for improvement in the existing technology, so it is necessary to propose more reasonable technical solutions to solve the technical problems existing in the existing technology. Summary of the Invention

[0005] In order to overcome at least one of the defects mentioned above, the present invention proposes a method for well leakage warning using seismic prediction results, aiming to use the seismic data prediction results to make precise predictions of small-scale fracture zones or fracture-cavity bodies in the exploration and development area, to guide well location design, drilling design and drilling trajectory optimization, which has important practical significance and application value.

[0006] In order to achieve the above-mentioned purpose, the early warning method disclosed in the present invention can adopt the following technical solutions: A method for early warning of well leakage using earthquake prediction results, comprising: Basic data acquisition: collect and organize post-stack depth domain seismic data and pre-stack time domain data; Regional geological analysis: Based on regional drilling and lost circulation data, regional lost circulation analysis is completed; based on the results of 3D seismic interpretation in the depth domain, the fault type, scale, and distribution characteristics within the work area are determined; Seismic data preprocessing: Anisotropic diffusion and three-dimensional multi-level median filtering are performed on post-stack depth domain seismic data to improve the signal-to-noise ratio of basic data and enhance the response characteristics of faults and fractures; Fine seismic characterization of small faults and fractures: First, structural tensor coherence attribute calculations are performed based on pre-processed post-stack depth-domain seismic data to characterize the basic morphology of small faults and fractures in the region. Pre-stack multi-azimuth frequency attenuation gradients are then performed using pre-stack seismic data to obtain fracture density results, characterizing the anisotropy differences caused by fracture zones in the region. Finally, principal components analysis (PCA) calculations with adjustable parameters are performed on the fracture density inversion results and coherent structural tensor data to characterize the morphology of small faults and karst fractures. Shadow relief display of characterization results: The attribute values of the seismic fine characterization results of small faults and karst fractures and caves are converted into shadow reliefs to represent the changes in uplift and depression, and to highlight the size and scale of small faults and karst fractures and caves.

[0007] The above-mentioned public early warning method uses the prediction results of seismic data to make precise predictions of small-scale fracture zones or fracture-cavity bodies in the exploration and development area. It can discover the degree of development of fracture zones and fracture-cavity bodies, and facilitate early judgment of the possibility of well leakage based on changes in the degree of development, thereby realizing early exploration and discovery of safety hazards and reducing mining risks.

[0008] Furthermore, in the seismic data preprocessing step, anisotropic diffusion filtering calculation is performed according to the following method:

[0009] In the above formula, For earthquake data, is the diffusion time, For divergence calculation, is the diffusion tensor.

[0010] Furthermore, in the seismic data preprocessing step, three-dimensional multi-level median filtering is performed as follows:

[0011] In the above formula, is the amplitude of a certain point in the 3D seismic data, for The amplitude of the 26 adjacent points in the surrounding space, is the filter output value, To take the middle value, A i (s) is the value range after median processing.

[0012] Furthermore, in the step of fine seismic characterization of small faults and fractures, the calculation of the structural tensor coherence attribute body includes: calculating the directional derivative vector of each point of the three-dimensional seismic data and constructing a structural tensor, and obtaining the structural tensor coherence attribute body by calculating the eigenvalues and eigenvectors of its structural tensor matrix.

[0013] Furthermore, in the step of fine seismic characterization of small faults and fractures, the front multi-azimuth frequency attenuation gradient is determined according to the following method:

[0014] The position of the maximum energy point in the instantaneous amplitude spectrum is ( , ), the minimum point position is ( , ), is the frequency attenuation gradient, E is the amplitude energy, , are the frequency and amplitude energy of the maximum energy point, , are the frequency and amplitude energy of the minimum energy point respectively.

[0015] Furthermore, in the step of fine seismic characterization of small faults and fractures, the attribute fusion calculation under adjustable parameters includes: Create a matrix of two attribute bodies ,in It mainly refers to the attribute value of the i-th attribute and the j-th sample point, a is the number of attribute bodies, b is the number of sample points for each attribute, and is standardized to eliminate the influence of different dimensions; Calculate the correlation coefficients of different attribute matrices after standardization; By finding the non-negative characteristic roots of the correlation matrix R, we can obtain the eigenvector matrix E and calculate the principal component F=HD ;in, D The matrix created for the attribute body; Set the threshold factor k. If the ratio of the variance of the first i principal components to the total variance is greater than or equal to the threshold factor k, the first i principal components can be used to represent the original attribute sample. The principal components are fused, and the threshold factor k is adjusted based on the comprehensive evaluation of the fusion effect and geological conditions.

[0016] Furthermore, the correlation coefficients of the standardized attribute matrices are calculated as follows:

[0017] In the above formula, is the correlation coefficient, and R is the correlation coefficient matrix. The matrix created for the attribute body, is the attribute value of the sample point, The number of attribute bodies.

[0018] Furthermore, the principal components are fused according to the following method: The first i principal components are rearranged into a new eigenvector matrix W to obtain a new attribute matrix , here The matrix created for the attribute body.

[0019] Furthermore, the threshold factor k is adjusted as follows: Attribute Body If the spatial coincidence rate with the statistical number of lost circulation is greater than 80%, no adjustment is required. If it is less than 80%, increase the k factor, which is generally increased based on experience.

[0020] Compared with the prior art, some of the beneficial effects of the technical solution disclosed in the present invention include: The present invention improves the basic data quality by introducing anisotropic diffusion and three-dimensional multi-level median joint filtering technology, and carries out multi-attribute fusion based on adjustable parameters for pre-stack multi-directional frequency attenuation gradient fracture prediction and post-stack structural coherence tensor. Finally, the multi-attribute fusion results are spatially displayed using shadow relief technology, which improves the accuracy of small fault and fracture-cavity characterization and more intuitively displays the spatial morphology of faults and fracture-cavity bodies, which can effectively guide well location design, drilling design and drilling trajectory optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only represent some embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 The figure is a flow chart of a method for using earthquake prediction results to provide early warning of well leakage.

[0023] Figure 2 This is the result of fine seismic characterization of small faults and fractures and caves provided in the embodiments of the present invention.

[0024] Figure 3 This is a shadow relief display of the seismic fine characterization results of small faults and fracture-cavity bodies provided in an embodiment of the present invention.

[0025] Figure 4 Schematic diagram of azimuthal anisotropy of different attributes in the same CMP gather. DETAILED DESCRIPTION

[0026] This embodiment will be further explained below with reference to the accompanying drawings and specific examples.

[0027] In view of the fact that the prediction and early warning of lost circulation in the prior art is not accurate enough and the effect is not good, the following embodiments are optimized to overcome the defects in the prior art.

[0028] Example like Figure 1 As shown, the embodiment provides a method for using earthquake prediction results to provide early warning of well leakage, which is as follows: Step 1: Basic data acquisition: Collect and organize post-stack depth domain seismic data and pre-stack time domain data.

[0029] Step 2: Regional geological analysis: Based on regional drilling and lost circulation data, complete regional lost circulation analysis; based on the results of 3D seismic interpretation in the depth domain, further determine the type, scale, and distribution characteristics of the faults within the work area.

[0030] Step 3: Seismic data preprocessing for fault and fracture-cavity prediction. Anisotropic diffusion and three-dimensional multi-level median filtering are performed on the post-stack depth-domain seismic data to improve the signal-to-noise ratio of the basic data and enhance the response characteristics of faults and fractures. To further ensure the anisotropic characteristics of the pre-stack time-domain data, no preprocessing is performed on the pre-stack time-domain data.

[0031] Step 4: Fine seismic characterization of small faults and fractures. First, the structural tensor coherent attribute body is calculated based on the pre-processed post-stack depth domain seismic data to characterize the basic morphology of small faults and fractures in the region; then, the pre-stack multi-directional frequency attenuation gradient is performed using the pre-stack seismic data to obtain the fracture density results, and finely characterize the anisotropy differences caused by the fracture belts in the region; finally, the fracture density result inversion results are combined with the coherent structural tensor data to perform attribute fusion calculations under adjustable parameters, further combining the characteristics of the two, and finally achieving high-precision characterization of small faults and karst fractures. The specific results can be seen in Figure 2 The darker the color, the more developed the small faults and fractures are; the lighter the color, the less developed they are.

[0032] Step 5: Shadow relief display of the characterization results. The attribute values of the fine seismic characterization results of small faults and karst fractures and caves are converted into shadow relief to represent the changes in uplift and depression, and to highlight the size and scale of small faults and karst fractures and caves. The specific results can be seen in Figure 3 , its convex and concave changes can better reflect its scale and size, and need to be combined with Figure 2 , comprehensively identify the development degree, extension scale and size of its small faults and fracture-cavities.

[0033] The combined filtering of anisotropic diffusion and three-dimensional multi-level median method in step three refers to combining the anisotropic diffusion filtering method with alpha filtering. Alpha filtering is performed based on the anisotropic diffusion filtered data. Anisotropic diffusion filtering can enhance the discontinuity characteristics in faults, fractures, or seismic data, and three-dimensional multi-level median filtering can effectively improve the data signal-to-noise ratio. The combination of the two can enhance the discontinuity characteristics in seismic data while improving the signal-to-noise ratio of seismic data. The calculation formula for anisotropic diffusion filtering is as follows:

[0034] In the above formula, For earthquake data, is the diffusion time, For divergence calculation, is the diffusion tensor. The calculation formula of three-dimensional multi-level median filtering is as follows:

[0035] In the above formula, is the amplitude of a certain point in the 3D seismic data, for The amplitude of the 26 adjacent points in the surrounding space, is the filter output value, To take the middle value.

[0036] The calculation of the structural tensor coherence attribute volume in step 4 primarily involves two steps: first, calculating the directional derivative vector of each point in the 3D seismic data and constructing a structural tensor; and finally, obtaining the structural tensor coherence attribute volume by calculating the eigenvalues and eigenvectors of its structural tensor matrix. This method effectively characterizes structural features in seismic data and is generally used as a basic morphological identification method for faults and fracture-cavity bodies.

[0037] The multi-azimuth frequency attenuation gradient in step 4 refers to extracting the frequency attenuation gradient from multiple azimuth gathers of pre-stack data, and finally obtaining the fracture density inversion result through ellipse fitting. The frequency attenuation gradient expression is as follows:

[0038] The position of the maximum energy point in the instantaneous amplitude spectrum is ( , ), the minimum point position is ( , ). The frequency attenuation gradient mainly describes how quickly the energy in the high-frequency band of seismic data decays as the frequency increases. If there are factors such as cracks, the absolute value of the attenuation gradient will increase. The ellipse azimuth fitting mainly uses the reflection coefficient formula of the seismic signal amplitude changing with the azimuth as a theoretical guide (Ruger, 1996), which usually describes the azimuthal anisotropy of different attributes in the same CMP (Common Middle Point) data set, that is, the reflection information received by the same point underground at different azimuths is different. We arrange the information of the same reflection point in different azimuths in a circular manner according to different azimuths (0-360 degrees). The distance from the center point to the ring is the attribute value of the same reflection point in different azimuths. When the attribute values of the same reflection point in different azimuths are the same, it is isotropic (no difference in different azimuths), and there are often no cracks. When there are differences in the attribute values of the same reflection point in different azimuths, the circular arrangement is an ellipse, and the reflection point has azimuthal anisotropy. The major axis of the ellipse represents the direction of the crack, and the ratio of the major axis to the minor axis of the ellipse represents the crack density, such as Figure 4 shown.

[0039] The attribute fusion calculation under the adjustable parameters in step 4 is to fuse the structural tensor coherence volume with the fracture density inversion result, which mainly includes the following steps: (1) Create a matrix of two attribute bodies ,in It mainly refers to the attribute value of the i-th attribute and the j-th sample point, a is the number of attribute bodies, b is the number of sample points for each attribute, and is standardized to eliminate the influence of different dimensions.

[0040] (2) Calculate the correlation coefficient of different attribute matrices after standardization, and the calculation formula is:

[0041] In the above formula, is the correlation coefficient, and R is the correlation coefficient matrix. The matrix created for the attribute body, is the attribute value of the sample point, The number of attribute bodies.

[0042] (3) Obtain the eigenvector matrix by finding the non-negative characteristic roots of the correlation matrix R H , and calculate the principal component F= HD.

[0043] (4) Artificially set the threshold factor k. Assuming that the proportion of the variance of the first i principal components to the total variance is greater than or equal to the threshold factor k, the first i principal components can be used to represent the original attribute sample.

[0044] (5) The principal components are fused and the threshold factor k is adjusted based on the comprehensive evaluation of the fusion effect and geological conditions to achieve the best fusion.

[0045] The shadow relief display in step 5 is a technology that integrates computer graphics, optical principles and data visualization to convert two-dimensional plane data (such as elevation, grayscale or attribute values) into a relief image with depth perception. Its core principle is to reconstruct the spatial form by using the relationship between light and dark, so that the observer can perceive the convexity and concavity of the surface through the visual shadow changes, such as Figure 2 、 Figure 3 shown.

[0046] The above are the implementation methods listed in this embodiment, but this embodiment is not limited to the above optional implementation methods. Those skilled in the art can arbitrarily combine the above methods to obtain other various implementation methods. Anyone can derive other various implementation methods based on the inspiration of this embodiment. The above specific implementation methods should not be understood as limiting the scope of protection of this embodiment. The scope of protection of this embodiment should be based on the definition in the claims.

Claims

1. A method for early warning of well leakage using earthquake prediction results, characterized in that: include: Basic data acquisition: collect and organize post-stack depth domain seismic data and pre-stack time domain data; Regional geological analysis: Based on regional drilling and lost circulation data, regional lost circulation analysis is completed; based on the results of 3D seismic interpretation in the depth domain, the fault type, scale, and distribution characteristics within the work area are determined; Seismic data preprocessing: Anisotropic diffusion and three-dimensional multi-level median method combined filtering are performed on post-stack depth domain seismic data; Fine seismic characterization of small faults and fractures: First, structural tensor coherence attribute calculations are performed based on pre-processed post-stack depth-domain seismic data to characterize the basic morphology of small faults and fractures in the region. Pre-stack multi-azimuth frequency attenuation gradients are then performed using pre-stack seismic data to obtain fracture density results, characterizing the anisotropy differences caused by fracture zones in the region. Finally, principal component analysis calculations with adjustable parameters are performed on the fracture density inversion results and coherent structural tensor data to achieve morphological characterization of small faults and karst fractures. Shadow relief display of characterization results: The attribute values of the seismic fine characterization results of small faults and karst fractures and caves are converted into shadow reliefs to represent the changes in uplift and depression, and to highlight the size and scale of small faults and karst fractures and caves.

2. The method for using earthquake prediction results to provide early warning of lost circulation according to claim 1, characterized in that: In the seismic data preprocessing step, anisotropic diffusion filtering calculation is performed according to the following method: In the above formula, For earthquake data, is the diffusion time, For divergence calculation, is the diffusion tensor.

3. The method for using earthquake prediction results to provide early warning of lost circulation according to claim 1, characterized in that: In the seismic data preprocessing step, three-dimensional multi-level median filtering is calculated according to the following method: In the above formula, is the amplitude of a certain point in the 3D seismic data, for The amplitude of the 26 adjacent points in the surrounding space, is the filter output value, To take the middle value, A i (s) is the value range after median processing.

4. The method for using earthquake prediction results to provide early warning of lost circulation according to claim 1, characterized in that: In the step of fine seismic characterization of small faults and fractures, the calculation of the structural tensor coherent attribute body includes: calculating the directional derivative vector of each point of the three-dimensional seismic data and constructing the structural tensor, and obtaining the structural tensor coherent attribute body by calculating the eigenvalues and eigenvectors of its structural tensor matrix.

5. The method for using earthquake prediction results to provide early warning of well leakage according to claim 1, characterized in that: In the step of fine seismic characterization of small faults and fracture-cavity bodies, the front multi-azimuth frequency attenuation gradient is determined according to the following method: The position of the maximum amplitude energy point in the instantaneous amplitude spectrum is ( , ), the minimum point position is ( , ), is the frequency attenuation gradient, E is the amplitude energy, 、 are the frequency and amplitude energy of the maximum energy point, 、 are the frequency and amplitude energy of the minimum energy point respectively.

6. The method for using earthquake prediction results to provide early warning of well leakage according to claim 1, characterized in that: In the step of fine seismic characterization of small faults and fractures and caves, the attribute fusion calculation under adjustable parameters includes: Create a matrix of two attribute bodies ,in It mainly refers to the attribute value of the i-th attribute and the j-th sample point, a is the number of attribute bodies, b is the number of sample points for each attribute, and is standardized to eliminate the influence of different dimensions; Calculate the correlation coefficients of different attribute matrices after standardization; By finding the non-negative characteristic roots of the correlation matrix R, we can obtain the eigenvector matrix H , and calculate the principal component F=HD ;in, D The matrix created for the attribute body; Set the threshold factor k. If the ratio of the variance of the first i principal components to the total variance is greater than or equal to the threshold factor k, the first i principal components can be used to represent the original attribute sample. The first i principal components are selected and fused, and the threshold factor k is adjusted based on the comprehensive evaluation of the fusion effect and geological conditions.

7. The method for using earthquake prediction results to provide early warning of well leakage according to claim 6, characterized in that: The correlation coefficients of different attribute matrices after standardization are calculated as follows: In the above formula, is the correlation coefficient, and R is the correlation coefficient matrix. The matrix created for the attribute body, is the attribute value of the sample point, The number of attribute bodies.

8. The method for using earthquake prediction results to provide early warning of lost circulation according to claim 6, characterized in that: The first i principal components are fused as follows: The first i principal components are rearranged into a new eigenvector matrix W to obtain a new attribute matrix , here The matrix created for the attribute body.

9. The method for using earthquake prediction results to provide early warning of lost circulation according to claim 6, characterized in that: Adjust the threshold factor k as follows: Attribute Body If the spatial coincidence rate with the statistical number of lost circulation is greater than 80%, no adjustment is required; if it is less than 80%, the k factor should be increased.

Images