A method and apparatus for predicting fractured sweet spot reservoirs based on 3D seismic data

By employing a method based on 3D seismic data and utilizing dip-guided coherent enhancement denoising, fault identification, morphological dilatation, and conditional random field fusion techniques, the problem of insufficient accuracy in predicting fracture-type sweet spot reservoirs was solved. This enabled a fine characterization of the complex relationship between faults, fractures, and porosity, as well as a comprehensive interpretation of reservoir characteristics.

CN119758453BActive Publication Date: 2025-10-28BEIJING ZHONGHENG LIHUA PETROLEUM TECH RES INST
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202510247694.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-03-04
Publication Date
2025-10-28
Estimated Expiration
2045-03-04

AI Technical Summary

Technical Problem

Existing technologies are insufficient to fully reflect the complex relationship between fractures, faults, and porosity, resulting in inaccurate predictions of fracture-type sweet spot reservoirs, especially in areas with highly developed fracture systems, where it is difficult to reflect the true distribution and connectivity of fractures.

Method used

Using a method based on 3D seismic data, this study employs dip-guided coherence enhancement denoising, fault identification, fracture volume data generation, morphological dilatation operation, and conditional random field attribute fusion, combined with seismic waveform indicator inversion technology, to finely characterize fracture zones and porosity data volumes, thereby generating fracture-type sweet spot reservoir data volumes.

Benefits of technology

It enables detailed characterization of fractured sweet spot reservoirs, allowing for a more comprehensive interpretation of reservoir geological features, identification of "sweet spot" areas within the reservoir, and improved risk assessment capabilities for exploration and development.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119758453B_ABST
    Figure CN119758453B_ABST
Patent Text Reader

Abstract

This invention provides a method and apparatus for predicting fracture-type sweet spot reservoirs based on 3D seismic data, belonging to the field of oil and gas geophysical exploration technology. The method involves denoising 3D seismic data to obtain denoised 3D seismic data, and then extracting fault data from the denoised 3D seismic data. Next, fracture volume data is further extracted from the fault data, and morphological dilation is performed on the fracture volume data to generate fracture zone data. Then, 3D seismic porosity parameter simulation is performed on the 3D seismic data to obtain a porosity data volume. Finally, conditional random field attribute fusion is performed on the fracture zone data and the porosity data volume to obtain a fracture-type sweet spot reservoir data volume. This data volume is then visualized to obtain a fracture-type sweet spot reservoir image. This invention, by fusing fracture zone and porosity data, can more comprehensively interpret the geological characteristics of the reservoir and helps to identify "sweet spot" areas within the reservoir.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of oil and gas geophysical exploration technology, and in particular to a method and apparatus for predicting fracture-type sweet spot reservoirs based on three-dimensional seismic data. Background Technology

[0002] In tight oil and gas reservoirs, the matrix reservoir often has low porosity and permeability, resulting in low exploration and development efficiency. Fractured sweet spots refer to reservoir areas with superior storage and permeability formed on the basis of the matrix reservoir through the modification of faults and fractures. Accurate characterization of fractured sweet spots is of great significance for tight oil and gas reservoirs.

[0003] Current research has focused on predicting fractured sweet spot reservoirs, resulting in methods such as fracture detection based on coherence volume and gradient structure tensor algorithms, fracture prediction techniques driven by seismic rock physics, and shale oil reservoir sweet spot prediction based on well logging data. These methods have addressed the problem of predicting fractured sweet spot reservoirs to some extent.

[0004] However, the prediction of fractured sweet spot reservoirs still has certain shortcomings: On the one hand, traditional fracture prediction techniques often only provide linear fault and fracture information, limiting the understanding of the complexity of fracture networks. This is especially true in areas with highly developed fracture systems, such as fracture zones, where linear information struggles to reflect the true distribution and connectivity of fractures. On the other hand, commonly used methods for fusing fractures and matrix reservoirs are mainly weighted fusion methods. These methods may be too simplistic and fail to fully reflect the complex relationships between fractures, faults, and porosity. They cannot demonstrate the connectivity of the fracture network or the complexity of the spatial configuration of fractures and porosity.

[0005] Therefore, there is an urgent need for a method for predicting fractured sweet spot reservoirs based on 3D seismic data. Summary of the Invention

[0006] In view of this, the present invention provides a method and apparatus for predicting fractured sweet spot reservoirs based on three-dimensional seismic data, in order to solve the problem that sweet spot reservoir prediction methods are unable to fully reflect the complex relationship between fractures, faults and porosity, and are not accurate enough in characterizing the fault zone structure.

[0007] The technical solution adopted in this invention is:

[0008] In a first aspect, the present invention provides a method for predicting fractured sweet spot reservoirs based on three-dimensional seismic data, comprising:

[0009] The dip-guided coherent enhancement denoising method was used to denoise the 3D seismic data, resulting in denoised 3D seismic data.

[0010] The intrinsic coherence volume technique is used to extract key feature values ​​from the denoised 3D seismic data, and fault identification is performed on the strata based on the key feature values ​​to obtain fault data.

[0011] Approximate derivative calculations are performed on the fault data, and the corresponding fracture body data are obtained based on the approximate derivative calculation results;

[0012] Morphological dilation is performed on the fracture body data to generate fracture zone data;

[0013] The three-dimensional seismic porosity parameters were simulated using the seismic waveform indication inversion technique to obtain the porosity data volume;

[0014] Conditional random field attribute fusion was performed on the fracture zone data and porosity data to obtain the fracture-type sweet spot reservoir data.

[0015] The data volume of the fractured sweet spot reservoir is input into a preset 3D visualization model for data visualization conversion to obtain the image of the fractured sweet spot reservoir.

[0016] Furthermore, the morphological dilation operation performed on the fracture body data to generate fracture zone data includes:

[0017] Based on the fracture image in the fracture image data, define the input fracture image. and structural elements ; Indicates the input image The pixel values ​​in the image represent the sampling points of the fracture surface. and These represent the input images respectively. The x and y coordinates of the middle pixel; Represents structural element Pixel values ​​in and Representing structural elements respectively The x and y coordinates of the middle pixel;

[0018] The structural element is designed to be circular, and the direction of its outward expansion is taken as the perpendicular direction to the fracture surface. This is used to analyze the fracture surface image. Perform expansion and padding for each pixel coordinate. The pixel value at that location is used to calculate the result of the morphological dilation operation. :

[0019]

[0020] in, Represents structural element Pixel coordinates in;

[0021] Repeatedly examine the image of the fractured body. Perform dilation operation until the dilated image meets the preset image standard, and then obtain the corresponding fracture zone data.

[0022] Furthermore, the method of using seismic waveform indication inversion technology to simulate three-dimensional seismic porosity parameters from three-dimensional seismic data to obtain a porosity data volume includes:

[0023] Using the singular value decomposition method, efficient dynamic clustering analysis of well-side seismic trace waveforms is performed on 3D seismic data to establish the mapping relationship between seismic waveform structure and porosity curve structure, and to generate a set of porosity curve sample sets representing different seismic phase types.

[0024] By analyzing the distribution of porosity curve sample sets for different seismic facies types, corresponding Bayesian inversion frameworks were established. Under each Bayesian framework, the common part of the sample set was selected as the initial model, and iterative simulation was performed to obtain high-resolution porosity data volumes.

[0025] Furthermore, the conditional random field attribute fusion of the fracture zone data and porosity data volume to obtain the fracture-type sweet spot reservoir data volume includes:

[0026] Define a Conditional Random Field (CRF) model. The CRF model takes the geological features of the fracture zone and the reservoir parameters of the porosity data volume as input to predict the reservoir characteristics of each pixel or sampling point.

[0027] Potential functions are constructed by selecting reservoir characteristic-related features from fracture zone data and porosity data volumes to quantify the values ​​of individual variables or the relationships between variables;

[0028] The parameters of the potential function are learned by using the maximum likelihood estimation algorithm. Using the learned parameters, the conditional random field model is used to perform parameter fusion inference and prediction on the input fracture zone data and porosity data volume, and generate the corresponding fracture-type sweet spot reservoir data volume.

[0029] Secondly, the present invention provides a device for predicting fractured sweet spot reservoirs based on three-dimensional seismic data, comprising:

[0030] The data denoising module is used to denoise 3D seismic data using the dip-guided coherent enhancement denoising method to obtain denoised 3D seismic data.

[0031] The fault identification module is used to extract key feature values ​​from the denoised 3D seismic data using intrinsic coherence volume technology, and to identify faults in the strata based on the key feature values ​​to obtain fault data.

[0032] The fracture body extraction module is used to perform approximate derivative calculation on the fault data and obtain the corresponding fracture body data based on the approximate derivative calculation result;

[0033] The fracture zone generation module is used to perform morphological dilation operations on the fracture body data to generate fracture zone data.

[0034] The seismic inversion module is used to simulate the three-dimensional seismic porosity parameters of three-dimensional seismic data using seismic waveform indication inversion technology, and obtain the porosity data volume.

[0035] The attribute fusion module is used to perform conditional random field attribute fusion on fracture zone data and porosity data volume to obtain fracture-type sweet spot reservoir data volume;

[0036] The reservoir prediction module is used to input the data volume of fractured sweet spot reservoir into a preset 3D visualization model for data visualization conversion, and obtain the image of fractured sweet spot reservoir.

[0037] In summary, the beneficial effects of the present invention are as follows:

[0038] This invention provides a method for predicting fractured sweet spot reservoirs based on 3D seismic data. The method first denoises the 3D seismic data to obtain denoised 3D seismic data, and then extracts fault data from the denoised 3D seismic data. Next, it further extracts fracture body data from the fault data and performs morphological dilation on the fracture body data to generate fracture zone data. By applying intrinsic coherence and gradient analysis techniques, it can more accurately identify and extract faults and fractures. Furthermore, it uses morphological dilation to refine the properties of the fracture bodies, going beyond providing only linear fault and fracture information to finely characterize the complex structure of fault zones. Then, it performs 3D seismic porosity parameter simulation on the 3D seismic data to obtain porosity data volumes. Finally, it fuses the fracture zone data and porosity data volumes using conditional random field properties, which can more comprehensively interpret the geological characteristics of the reservoir, including the development degree of faults and fractures and their coupling relationship with high-porosity reservoirs. This helps to identify "sweet spot" areas in the reservoir and can be used to better assess the risks of exploration and development. Attached Figure Description

[0039] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, and these are all within the protection scope of the present invention.

[0040] Figure 1 This is a schematic diagram of the process for predicting fracture-type sweet spot reservoirs based on three-dimensional seismic data according to the present invention.

[0041] Figure 2 This is a schematic diagram of the seismic profile before the noise reduction process of this invention;

[0042] Figure 3 This is a schematic diagram of the seismic profile after noise reduction processing according to the present invention;

[0043] Figure 4 This is a schematic diagram of the overlay profile of fault data and seismic data of the present invention;

[0044] Figure 5 This is a schematic diagram of the fault data profile of the present invention;

[0045] Figure 6 This is a schematic cross-sectional view of the fracture body data of the present invention;

[0046] Figure 7 This is a schematic cross-sectional view of the fracture band data of the present invention;

[0047] Figure 8 This is a schematic diagram of the cross-sectional view of the fracture body data superimposed with fracture zone data of the present invention;

[0048] Figure 9 This is a schematic diagram of the porosity profile of the present invention;

[0049] Figure 10 This is a schematic diagram of the cross-sectional data of the fractured sweet spot reservoir of the present invention;

[0050] Figure 11 This is a schematic diagram of the cross-sectional data of the superimposed fracture body of the fractured sweet spot reservoir data of the present invention;

[0051] Figure 12 This is a schematic diagram illustrating the three-dimensional visualization effect of the fractured sweet spot reservoir of the present invention;

[0052] Figure 13 This is a functional block diagram of the fracture-type sweet spot reservoir prediction device based on three-dimensional seismic data of the present invention. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Unless otherwise specified, the present invention and the various features in the embodiments can be combined with each other, all of which are within the protection scope of the present invention.

[0054] Example 1: Refer to Figure 1 As shown, Figure 1This is a flowchart of a quantitative characterization method for seismic waveforms according to the present invention. (Refer to...) Figure 1 As shown, the present invention provides a method for predicting fractured sweet spot reservoirs based on three-dimensional seismic data, comprising:

[0055] S1: The dip-guided coherent enhancement denoising method is used to denoise the 3D seismic data to obtain the denoised 3D seismic data.

[0056] S2: Using intrinsic coherence volume technology, key feature values ​​are extracted from the denoised 3D seismic data to identify faults in the strata based on the key feature values, and fault data is obtained.

[0057] S3: Perform approximate derivative calculation on the fault data, and obtain the corresponding fracture body data based on the approximate derivative calculation result;

[0058] S4: Perform morphological dilation on the fracture body data to generate fracture zone data;

[0059] S5: Using seismic waveform indication inversion technology, three-dimensional seismic porosity parameters are simulated from three-dimensional seismic data to obtain porosity data volume;

[0060] S6: Perform conditional random field attribute fusion on the fracture zone data and porosity data to obtain the fracture-type sweet spot reservoir data;

[0061] S7: Input the data volume of the fractured sweet spot reservoir into the preset 3D visualization model for data visualization conversion to obtain the image of the fractured sweet spot reservoir.

[0062] Specifically, in the embodiments of the present invention, Figure 1 and Figure 2 This is a schematic diagram of seismic profiles before and after denoising of 3D seismic data using dip-guided coherent enhancement denoising. Denoising is a crucial step in seismic data processing because random noise can mask or obscure seismic reflection information, especially for subtle geological structures such as faults. Intrinsic coherence techniques improve the quality of seismic signals through specific algorithms, and dip-guided coherent enhancement denoising is an effective technique in this regard.

[0063] This method first uses plane wave destruction filtering to estimate the dip angle of the seismic body. Dip angle information reflects the geometric characteristics of the strata and helps identify the energy direction of seismic reflections. By calculating the similarity between seismic traces, coherence properties are generated, which highlight the continuity and consistency in the seismic signal, thereby enhancing the effective signal. Combined with the dip angle information, directional denoising is performed on the seismic signal. This method does not simply suppress noise in the horizontal direction, but rather along the dip and strike of the seismic reflection interface, more closely reflecting the true geological situation. During denoising, the dip-guided filtering method better protects important geological information such as faults in the seismic signal, avoiding the loss of these crucial details during denoising. Through the above processing, the signal-to-noise ratio of the seismic data is improved, making geological structures such as faults more clearly visible, thus providing a more reliable data foundation for subsequent geological interpretation and reservoir prediction.

[0064] Specifically, intrinsic coherence is an advanced technique for interpreting seismic data. It characterizes the lateral heterogeneity of strata by analyzing the similarity of signals from adjacent seismic traces within a seismic data volume, thereby identifying faults within the strata. The core of this technique lies in calculating the coherence value of the seismic data volume to highlight discontinuities in the seismic data, such as faults and fractures.

[0065] When processing seismic data, intrinsic coherence technology defines three-dimensional seismic data as a matrix and extracts key feature values ​​from seismic waveforms through mathematical methods to characterize the properties of the strata (as shown in Formula 1 below).

[0066] (1)

[0067] In the formula, Indicates the intrinsic coherence value, This refers to a specific trace in the earthquake data. This represents the total number of traces in the seismic data. This indicates the number of sampling points in the seismic data. and Indicates the first and second in the earthquake data 1 eigenvalue, For the first The j-th sampling point.

[0068] Fault data obtained through intrinsic coherence calculations, such as Figure 4 and Figure 5 As shown, where Figure 4 This is a profile overlaid with fault and seismic data. Figure 5 This is a fault data profile.

[0069] Furthermore, based on the intrinsic coherence extraction of fault information, further extraction of fracture body properties is a key technology in oil and gas exploration. Fracture bodies typically refer to a region of a certain width surrounding a fault line, affected by fault activity. This region may contain fracture networks, porosity variations, and fluid flow channels, which are crucial for oil and gas accumulation and flow.

[0070] Gradient analysis algorithms are an important technique in seismic data processing, particularly in fault and fracture detection. The finite difference method, a commonly used method for calculating gradients, is a numerical method based on Taylor series expansion, used to approximate the derivative of a continuous function at discrete data points. For seismic data, this means that the derivative can be approximated using the values ​​of data points without an analytical functional form. Commonly used methods include forward differencing, backward differencing, and central differencing.

[0071] The forward difference method calculates the derivative of a function at a point and uses the difference between the function value at that point and the value at the next point. The backward difference method, conversely, uses the value of the function at the previous point to calculate the derivative. The central difference method combines the advantages of both forward and backward difference methods, using the values ​​of the function at two points before and after a given point to approximate the derivative (Equation 2). This method is more stable and accurate than the previous two methods. Therefore, this embodiment of the invention uses the central difference method to calculate the approximate derivative of fault data.

[0072] (2)

[0073] In the formula, This indicates the location of a sampling point in the seismic data; This represents the distance between two adjacent sampling points, i.e., the step size; Representative represents function At point The derivative at point, and They represent exist The values ​​at the next sampling point and the previous sampling point.

[0074] Finally, based on the approximate derivative calculation results, the final result is as follows: Figure 6 The data profile of the fracture body is shown.

[0075] Furthermore, in this embodiment of the invention, morphological dilation is an image processing technique based on structuring elements, used to expand and fill binary images. During the process of expanding fault lines to form fault zones, morphological dilation can be used to extend the influence of fault points to the surrounding area. Therefore, the morphological dilation operation performed on fault body data to generate fault zone data in this embodiment of the invention includes the following process:

[0076] The input fracture surface is an approximate binary image, let it be... The structural element is The result of the expansion operation is .

[0077] First, based on the fracture image in the fracture body data, define the input fracture image. and structural elements ; Indicates the input image The pixel values ​​in the image represent the sampling points of the fracture surface. and These represent the input images respectively. The x and y coordinates of the middle pixel; Represents structural element Pixel values ​​in and Representing structural elements respectively The x and y coordinates of the middle pixel;

[0078] The structural element is designed to be circular, and the direction of its outward expansion is taken as the perpendicular direction to the fracture surface. This is used to analyze the fracture surface image. Perform expansion and padding for each pixel coordinate. The pixel value at that location is used to calculate the result of the morphological dilation operation. (Formula 3):

[0079] (3)

[0080] in, Represents structural element Pixel coordinates in;

[0081] Repeatedly examine the image of the fractured body. Perform dilation operation until the dilated image meets the preset image standard, and then obtain the corresponding fracture zone data.

[0082] Considering that the formation of the fracture zone involves uniform outward expansion, the structural element is designed to be circular in shape, and its outward expansion direction is perpendicular to the fracture body. This ensures that the direction of the fracture zone formed after expansion is consistent with the direction of the fracture body.

[0083] In a fracture zone, the fault strata are larger structural elements, resulting in a more pronounced expansion effect, while cracks, being smaller structural elements, produce a more subtle expansion effect. The dilation operation is repeated on the entire fracture zone image, and the results are continuously observed until they meet the geologist's acceptable criteria. At this point, dilation is stopped, yielding the final predicted fracture zone result. Figure 7 and Figure 8The data profile of the fracture zone is shown. Figure 7 This is a data profile of the fracture zone. Figure 8 The fracture body data is superimposed with the fracture zone data profile.

[0084] Furthermore, in three-dimensional seismic reservoir prediction, the seismic waveform indication simulation method is a commonly used technique. By analyzing the correspondence between seismic waveforms and porosity curves, it can directly simulate the porosity volume. This method has higher accuracy and reliability compared to other reservoir inversion methods. In this embodiment of the invention, the seismic waveform indication inversion technique is used to simulate three-dimensional seismic porosity parameters from three-dimensional seismic data to obtain the porosity data volume, specifically including the following process:

[0085] Using the singular value decomposition method, efficient dynamic clustering analysis of well-side seismic trace waveforms is performed on 3D seismic data to establish a mapping relationship between the seismic waveform structure and the porosity curve structure, generating a set of porosity curve samples representing different seismic phase types.

[0086] By analyzing the distribution of porosity curve sample sets for different seismic facies types, corresponding Bayesian inversion frameworks were established. Within each Bayesian framework, the common components of the sample sets were selected as the initial model, and iterative simulations were performed to obtain high-resolution porosity data volumes. The porosity profiles within the porosity data volumes are shown below. Figure 9 As shown.

[0087] Furthermore, attribute fusion technology is a method that integrates information from different data sources to improve prediction accuracy. In seismic reservoir prediction, Conditional Random Fields (CRF) is a commonly used attribute fusion technique. It can combine the geological characteristics of fracture zones with reservoir parameters of porosity data to generate a comprehensive data volume reflecting reservoir characteristics. In this embodiment of the invention, conditional random field attribute fusion is performed on fracture zone data and porosity data to obtain fracture-type sweet spot reservoir data, specifically including the following process:

[0088] Define a Conditional Random Field (CRF) model. The CRF model takes the geological features of the fracture zone and the reservoir parameters of the porosity data volume as input to predict the reservoir characteristics of each pixel or sampling point.

[0089] Potential functions are constructed by selecting reservoir-related features from fracture zone and porosity data volumes to quantify the values ​​of individual variables or the relationships between variables. Potential functions typically include two types: nodal potentials, which describe the distribution of individual variables, and edge potentials, which describe the relationships between adjacent variables. Selecting edge potentials during the fusion of fracture zone and porosity data volumes better handles the matching relationships between them.

[0090] The parameters of the potential function are learned using the maximum likelihood estimation algorithm. Using these learned parameters, a conditional random field model is used to perform parameter fusion inference and prediction on the input fracture zone and porosity data volumes, generating corresponding fracture-type sweet spot reservoir data volumes. Finally, the desired results are obtained. Figure 10 and Figure 11 The data profile of the fractured sweet spot reservoir shown is as follows, Figure 10 This is a data profile of a fractured sweet spot reservoir. Figure 11 This is a cross-section of fractured body data superimposed on fractured sweet spot reservoir data.

[0091] Specifically, in this embodiment of the invention, the three-dimensional visualization model integrates different geological and geophysical properties (such as faults, fractures, and porous reservoirs) into a single three-dimensional model, transforming complex subsurface structures into intuitive images. This helps geologists identify reservoir geometries, such as layered structures, fracture networks, or irregular sweet spot regions. It also helps demonstrate the spatial relationships between reservoirs and fault networks, fracture networks, and other geological structures, which is crucial for understanding fluid flow paths within the reservoir. (Refer to...) Figure 12 The diagram shown illustrates the 3D visualization of a fractured sweet spot reservoir. The intuitive 3D model can guide the formulation of development strategies, such as determining the optimal drilling location, assessing potential oil and gas production, and predicting long-term development effects.

[0092] The method of this invention has the following technical effects:

[0093] Firstly, by applying intrinsic coherence and gradient analysis techniques, faults and fractures can be identified and extracted more accurately, and morphological dilatation operations can be used to refine the properties of the fracture body. Compared with traditional prediction techniques, this technique is no longer limited to providing linear fault and fracture information, but can finely characterize the complex structure of fault zones. This detailed analysis method is closer to the actual occurrence state of underground faults.

[0094] Secondly, the core advantage of using seismic waveform indication inversion technology to conduct three-dimensional seismic porosity parameter simulation is that it can obtain high-resolution inversion results and provide more accurate information on the pore structure of underground reservoirs.

[0095] Third, conditional random field attribute fusion technology can integrate information from different data sources, providing a unified framework for processing and fusing multi-attribute data. By fusing fault zone and porosity data, the geological characteristics of reservoirs can be interpreted more comprehensively, including the degree of development of faults and fractures and their coupling relationship with high-porosity reservoirs. This helps to identify "sweet spots" in reservoirs and can be used to better assess the risks of exploration and development.

[0096] Example 2: Refer to Figure 13 As shown, this embodiment of the invention also provides a device for predicting fractured sweet spot reservoirs based on three-dimensional seismic data, comprising:

[0097] The data denoising module is used to denoise 3D seismic data using the dip-guided coherent enhancement denoising method to obtain denoised 3D seismic data.

[0098] The fault identification module is used to extract key feature values ​​from the denoised 3D seismic data using intrinsic coherence volume technology, and to identify faults in the strata based on the key feature values ​​to obtain fault data.

[0099] The fracture body extraction module is used to perform approximate derivative calculation on the fault data and obtain the corresponding fracture body data based on the approximate derivative calculation result;

[0100] The fracture zone generation module is used to perform morphological dilation operations on the fracture body data to generate fracture zone data.

[0101] The seismic inversion module is used to simulate the three-dimensional seismic porosity parameters of three-dimensional seismic data using seismic waveform indication inversion technology, and obtain the porosity data volume.

[0102] The attribute fusion module is used to perform conditional random field attribute fusion on fracture zone data and porosity data volume to obtain fracture-type sweet spot reservoir data volume;

[0103] The reservoir prediction module is used to input the data volume of fractured sweet spot reservoir into a preset 3D visualization model for data visualization conversion, and obtain the image of fractured sweet spot reservoir.

[0104] The prediction device in this embodiment of the invention first denoises the 3D seismic data to obtain denoised 3D seismic data, and then extracts fault data from the denoised 3D seismic data. Next, it further extracts fracture body data from the fault data and performs morphological dilation on the fracture body data to generate fracture zone data. By applying intrinsic coherence and gradient analysis techniques, it can more accurately identify and extract faults and fractures. Furthermore, it uses morphological dilation to refine the properties of fracture bodies, going beyond simply providing linear fault and fracture information, and can finely characterize the complex structure of fault zones. Then, it performs 3D seismic porosity parameter simulation on the 3D seismic data to obtain a porosity data volume. Finally, it performs conditional random field attribute fusion on the fracture zone data and the porosity data volume, which can more comprehensively interpret the geological characteristics of the reservoir, including the degree of fault and fracture development and the coupling relationship with high-porosity reservoirs. This helps to identify "sweet spots" in the reservoir and can be used to better assess the risks of exploration and development.

[0105] Specifically, this embodiment of the invention also provides a computer storage medium storing computer program instructions, which, when executed by a processor, implement the seismic waveform quantitative characterization method as described in Embodiment 1.

[0106] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting fractured sweet spot reservoirs based on 3D seismic data, characterized in that, include: The dip-guided coherent enhancement denoising method was used to denoise the 3D seismic data, resulting in denoised 3D seismic data. The intrinsic coherence volume technique is used to extract key feature values ​​from the denoised 3D seismic data, and fault identification is performed on the strata based on the key feature values ​​to obtain fault data. Approximate derivative calculations are performed on the fault data, and the corresponding fracture body data are obtained based on the approximate derivative calculation results; Morphological dilation is performed on the fracture body data to generate fracture zone data, including: Based on the fracture image in the fracture image data, define the input fracture image. and structural elements ; Indicates the input image The pixel values ​​in the image represent the sampling points of the fracture surface. and These represent the input images respectively. The x and y coordinates of the middle pixel; Represents structural element The pixel values ​​in and Representing structural elements respectively The x and y coordinates of the middle pixel; The structural element is designed to be circular, and the direction of its outward expansion is taken as the perpendicular direction to the fracture surface. This is used to analyze the fracture surface image. Perform expansion and padding for each pixel coordinate. The pixel value at that location is used to calculate the result of the morphological dilation operation. : in, Represents structural element Pixel coordinates in; Repeatedly examine the image of the fractured body. Perform dilation operation until the dilated image meets the preset image standard, and then obtain the corresponding fracture zone data; The three-dimensional seismic porosity parameters were simulated using the seismic waveform indication inversion technique to obtain the porosity data volume; Conditional random field attribute fusion was performed on the fracture zone data and porosity data to obtain the fracture-type sweet spot reservoir data. The data volume of the fractured sweet spot reservoir is input into a preset 3D visualization model for data visualization conversion to obtain the image of the fractured sweet spot reservoir.

2. The method for predicting fractured sweet spot reservoirs according to claim 1, characterized in that, The method of using seismic waveform indication inversion technology to simulate three-dimensional seismic porosity parameters from three-dimensional seismic data to obtain a porosity data volume includes: Using the singular value decomposition method, efficient dynamic clustering analysis of well-side seismic trace waveforms is performed on 3D seismic data to establish the mapping relationship between seismic waveform structure and porosity curve structure, and to generate a set of porosity curve sample sets representing different seismic phase types. By analyzing the distribution of porosity curve sample sets for different seismic facies types, corresponding Bayesian inversion frameworks were established. Under each Bayesian framework, the common part of the sample set was selected as the initial model, and iterative simulation was performed to obtain high-resolution porosity data volumes.

3. The method for predicting fractured sweet spot reservoirs according to claim 1, characterized in that, The conditional random field attribute fusion of the fracture zone data and porosity data volume yields a fracture-type sweet spot reservoir data volume, including: Define a Conditional Random Field (CRF) model. The CRF model takes the geological features of the fracture zone and the reservoir parameters of the porosity data volume as input to predict the reservoir characteristics of each pixel or sampling point. Potential functions are constructed by selecting reservoir characteristic-related features from fracture zone data and porosity data volumes to quantify the values ​​of individual variables or the relationships between variables; The parameters of the potential function are learned by using the maximum likelihood estimation algorithm. Using the learned parameters, the conditional random field model is used to perform parameter fusion inference and prediction on the input fracture zone data and porosity data volume, and generate the corresponding fracture-type sweet spot reservoir data volume.

4. A device for predicting fracture-type sweet spot reservoirs based on three-dimensional seismic data, characterized in that, include: The data denoising module is used to denoise 3D seismic data using the dip-guided coherent enhancement denoising method to obtain denoised 3D seismic data. The fault identification module is used to extract key feature values ​​from the denoised 3D seismic data using intrinsic coherence volume technology, and to identify faults in the strata based on the key feature values ​​to obtain fault data. The fracture body extraction module is used to perform approximate derivative calculations on the fault data, and obtain the corresponding fracture body data based on the approximate derivative calculation results, including: Based on the fracture image in the fracture image data, define the input fracture image. and structural elements ; Indicates the input image The pixel values ​​in the image represent the sampling points of the fracture surface. and These represent the input images respectively. The x and y coordinates of the middle pixel; Represents structural element The pixel values ​​in and Representing structural elements respectively The x and y coordinates of the middle pixel; The structural element is designed to be circular, and the direction of its outward expansion is taken as the perpendicular direction to the fracture surface. This is used to analyze the fracture surface image. Perform expansion and padding for each pixel coordinate. The pixel value at that location is used to calculate the result of the morphological dilation operation. : in, Represents structural element Pixel coordinates in; Repeatedly examine the image of the fractured body. Perform dilation operation until the dilated image meets the preset image standard, and then obtain the corresponding fracture zone data; The fracture zone generation module is used to perform morphological dilation operations on the fracture body data to generate fracture zone data. The seismic inversion module is used to simulate the three-dimensional seismic porosity parameters of three-dimensional seismic data using seismic waveform indication inversion technology, and obtain the porosity data volume. The attribute fusion module is used to perform conditional random field attribute fusion on fracture zone data and porosity data volume to obtain fracture-type sweet spot reservoir data volume; The reservoir prediction module is used to input the data volume of fractured sweet spot reservoirs into a preset 3D visualization model for data visualization conversion, and obtain images of fractured sweet spot reservoirs.

Citation Information

Patent Citations

  • Morphological filtering-based electrical imaging reservoir fracture and cave body quantitative characterization method and system

    CN106443802A

  • Sand body configuration boundary intelligent extraction method based on seismic data

    CN119375953A