Shale reservoir fracture prediction method and fracture characterization and description tool

Abstract

The application provides a shale reservoir fracture prediction method and a fracture characterization and description tool, and the method comprises the following steps: decomposing post-stack seismic data based on a multi-level multi-azimuth method to obtain N seismic decomposition data of different levels; performing spectrum analysis and direction optimization on the seismic decomposition data of each level, selecting level x seismic decomposition data with the highest frequency information and at least one level seismic decomposition data with more original seismic effective information, and performing superposition combination to obtain high-resolution seismic data; performing ant body attribute calculation based on a variance body to obtain ant body seismic attributes, performing plane slice extraction on a target layer, and obtaining a fracture prediction result. The fracture prediction result can reach more than 60% in terms of the composite rate of actually drilled faults and well leakage, and the shale reservoir fracture prediction accuracy is significantly improved. The application has important guiding significance for actual exploration and development production.

Description

Technical Field

[0001] This invention relates to the fields of exploration geophysics and oil and gas development technology, particularly to the field of shale reservoir fracture characterization and description, specifically to shale reservoir fracture prediction methods and fracture characterization and description tools. Background Technology

[0002] Fractures not only store oil and gas but also serve as crucial pathways for their migration, significantly impacting reservoir oil and gas production. Accurate prediction of fracture spatial distribution is vital for studying fractured oil and gas reservoirs, improving exploration success rates, and enhancing development efficiency. This is particularly true for self-generating and self-storing shale reservoirs, where fracture development plays a crucial role in hydrocarbon gas accumulation, lateral and vertical migration, and shale oil and gas development. Therefore, predicting fracture spatial distribution is a key factor in shale gas preservation and development planning.

[0003] Currently, fracture prediction in shale reservoirs is still in its early stages, primarily drawing upon methods used for fracture prediction in carbonate and clastic reservoirs. Indirect methods for fracture prediction can be broadly categorized into three types: geological methods, well logging methods, and seismic methods. Geological methods are the most direct and simplest, enabling quantitative prediction, but they cannot meet the high-precision requirements of oil and gas field development and are not widely applicable in reservoir prediction. Well logging methods exhibit anomalous responses to fractures and can qualitatively identify them, but they are severely affected by environmental factors, frequently exhibit multiple interpretations, and are not suitable for large-scale production. Seismic methods allow for continuous observation in both vertical and horizontal directions and are relatively inexpensive, showing significant effectiveness in reservoir fracture prediction. Seismic methods are the primary method used in reservoir prediction. Seismic methods are further divided into pre-stack fracture prediction and post-stack fracture prediction techniques. Pre-stack fracture prediction techniques are costly, mainly used to identify high-angle fracture development zones, and require high-quality seismic data, being affected by factors such as high coverage frequency, uniform distribution, and wide azimuth. For fracture prediction in shale reservoirs, the most widely used method is the post-stack fracture prediction technique of the seismic method. This technique mainly uses seismic attributes such as coherence volume attributes, curvature attributes, variance volume attributes, ant volume attributes, and variance ant volume attributes to predict fracture development areas.

[0004] The primary criterion for fracture prediction using post-stack seismic attributes is the discontinuity of the in-phase axis in the post-stack seismic data, which reflects the seismic response characteristics at fracture development sites. However, limited by the resolution of seismic data, seismic attributes are often insufficient for accurately identifying medium- to small-order fractures. Furthermore, the type of seismic attribute and the parameter settings for each attribute significantly impact the fracture prediction results. In summary, current post-stack seismic attribute prediction methods suffer from low accuracy, high ambiguity, and limited prediction reliability. In practical production, the fracture prediction results using seismic attributes rarely achieve a good match with the actual fractures and faults encountered during drilling.

[0005] Therefore, developing a high-precision and reliable method for predicting fractures in shale reservoirs is of great guiding significance for actual exploration, development, and production. Summary of the Invention

[0006] The purpose of this invention is to address at least one of the aforementioned deficiencies in the prior art. For example, one objective of this invention is to provide a tool for characterizing and describing fractures in shale reservoirs; another objective is to improve the reliability of fracture prediction in shale reservoirs.

[0007] To achieve the above objectives, the present invention provides a method for predicting fractures in shale reservoirs, comprising the following steps:

[0008] The post-stack seismic data is decomposed based on a multi-level, multi-directional method to obtain N different levels of seismic decomposition data.

[0009] Spectral analysis and directional optimization were performed on the seismic decomposition data of each level to select the level x seismic decomposition data with the most high-frequency information and at least one level seismic decomposition data that retained more effective original seismic information.

[0010] By overlaying and combining the seismic decomposition data of level x and at least one level, high-resolution seismic data is obtained.

[0011] We perform ant-body attribute calculation based on variance volume on high-resolution seismic data to obtain ant-body seismic attributes;

[0012] Planar slices of the target layer are extracted from the seismic attributes of the ant body to obtain crack prediction results.

[0013] Alternatively, the crack prediction result can be a layer plane slice.

[0014] Optionally, the method further includes the steps of: constraining the calculation of the ant body attributes based on the variance volume using information on actual drilled faults and well leakage, and continuously adjusting the ant body attribute parameters to make the consistency rate between the layer plane slice and the actual drilled faults and well leakage greater than a preset value, thereby obtaining the shale reservoir fracture prediction result.

[0015] Alternatively, the method may further include: prior to constraining, collecting information on well leakage and encountered faults in the horizontal well at the target formation and projecting it onto the formation plane slice.

[0016] Alternatively, the method further includes the steps of: evaluating the post-stack seismic data before the step of decomposing the post-stack seismic data, and preprocessing the post-stack seismic data if the evaluation is poor.

[0017] Alternatively, the evaluation may include determining the signal-to-noise ratio, and the preprocessing may include at least one of denoising, flattening, and amplitude equalization.

[0018] Alternatively, N = 7, the order x is order 1, and the at least one order is order 2.

[0019] Alternatively, retaining more original effective earthquake information means retaining the most original effective earthquake information, or retaining the first and second most original effective earthquake information.

[0020] Alternatively, the preset value may be 50% or higher.

[0021] Another aspect of the present invention provides a computer device comprising: at least one processor and a memory storing program instructions, wherein the program instructions are configured to be executed by the at least one processor.

[0022] Alternatively, the program instructions may include instructions for performing the methods described above.

[0023] In another aspect, the present invention provides a computer-readable storage medium having stored thereon computer program instructions which, when executed by a processor, implement the above-described method.

[0024] Compared with the prior art, the beneficial effects of the present invention include at least one of the following:

[0025] (1) The accuracy of post-stack seismic attribute fracture prediction in shale reservoirs has been significantly improved.

[0026] (2) For seismic data with poor signal-to-noise ratio, the signal-to-noise ratio can be improved and crack prediction can be performed.

[0027] (3) The correlation between the predicted fractures in shale reservoirs and the actual encountered faults and well leakage has been significantly improved. Attached Figure Description

[0028] The above and other objects and / or features of the present invention will become clearer from the following description taken in conjunction with the accompanying drawings, in which:

[0029] Figure 1 A flowchart of a shale reservoir fracture prediction method in exemplary embodiment 1 is shown.

[0030] Figure 2 The seismic profile of the "multi-level, multi-azimuth" decomposition of seismic data in Example 1 is shown.

[0031] Figure 3 A stratigraphic planar slice of the study area in Example 1, predicted by conventional methods, is shown.

[0032] Figure 4A planar slice of the fracture horizon predicted according to the method of the present invention is shown in the study area of ​​Example 1.

[0033] Figure 5 The diagram showing the correlation between predicted fractures and encountered well leakage and faults in the study area according to the present invention in Example 1 is illustrated.

[0034] Figure 6A The diagram shows the matching between the fracture prediction profile obtained by conventional methods and the actual well leakage points.

[0035] Figure 6B The diagram shows the matching between the fracture prediction profile and the actual well leakage point in the drilling process according to the method of the present invention. Detailed Implementation

[0036] The present invention will be described in detail below with reference to exemplary embodiments and accompanying drawings. However, the following embodiments are only for the purpose of helping to understand the technology of the present invention and should not be construed as further limiting the scope of protection of the present invention.

[0037] Exemplary Example 1

[0038] This exemplary embodiment provides a method for predicting fractures in shale reservoirs, the operation flow of which is as follows: Figure 1 As shown, the following steps may be included:

[0039] S01: Based on the multi-level multi-azimuth method, the post-stack seismic data is decomposed to obtain N different levels of seismic decomposition data.

[0040] In this embodiment, the post-stack seismic data is evaluated before the decomposition step. If the post-stack seismic data fails the quality evaluation, it indicates low resolution (e.g., seismic frequency below 20Hz, such as 19Hz, 15Hz, or 10Hz) and / or low signal-to-noise ratio (SNR). A higher SNR generally indicates better seismic data, typically above 10:1. Data below 10:1 requires reprocessing, i.e., preprocessing. Preprocessing involves starting with pre-stack gathers and performing denoising, flattening, and amplitude equalization. The portion of the pre-stack gathers with high SNR and relatively amplitude-preserving characteristics (10° to 28°) is extracted as the raw post-stack seismic data for subsequent processing.

[0041] S02: Perform spectral analysis and directionality optimization on the seismic decomposition data of each level, and select the seismic decomposition data of the level with the most high-frequency information and at least one level of seismic decomposition data that retains more effective original seismic information.

[0042] In this embodiment, the core of the "multi-level, multi-azimuth" data decomposition method is to decompose the seismic signal in the frequency domain into multi-level data with different levels, directions, and containing different geological information. Combined with graded filter processing, the directional information of geological bodies in different levels of information is extracted. Different levels of information correspond to different resolutions and levels. Then, the level of data with better resolution is selected for recombination processing, achieving a significant frequency-enhancing effect that improves the resolution of seismic data. The graded filter can highlight the directional characteristics of faults and cracks in seismic data, while also improving filtering accuracy and reducing computational load. The functional expression of the graded filter is as follows:

[0043]

[0044] In the formula: f θ (x,y) refers to the function of the graded filter in the θ direction;

[0045] k j (θ) refers to the interpolation function in the θ direction;

[0046] The basis functions in the θ direction;

[0047] θ refers to the rotation angle;

[0048] j refers to the number of basis filters.

[0049] The specific processing procedure of the graded filter is to take each seismic data as input, perform convolution operation on the input seismic data with a set of three basis filters in different directions, then multiply the directional filtered seismic data by the corresponding interpolation function, and finally add the parts together to obtain the final filtered data.

[0050] Pre-stack gathers from 10° to 28° with high signal-to-noise ratio and relatively amplitude preservation are decomposed using a "multi-level, multi-azimuth" approach, resulting in, for example, seven different levels of seismic data. Spectral analysis is then performed on these data; the purpose of this spectral analysis is to verify the effectiveness of each level, although other levels may be selected in other applications. The dominant frequency of seismic data decreases sequentially from level 1 to level 7. The seismic profiles and post-stack results from level 2 are very similar, highlighting detailed information and partially enhancing the discontinuities in the original data's phase axes. For example, the seismic decomposition data from level 1 (with the most high-frequency information) and level 2 (with the most preserved effective information from the original seismic data) are preferred for subsequent analysis and processing.

[0051] S03: The seismic data of level x and the at least one level seismic decomposition data are superimposed and combined to obtain high-resolution seismic data.

[0052] In this embodiment, retaining more original effective earthquake information means retaining the most original effective earthquake information, or retaining the first and second most original effective earthquake information. For example, data of different orders can be selected and superimposed, such as superimposing order 1, order 2 and order 3, order 1 and order 2, or order 2 and order 3, etc.

[0053] The selection of seismic decomposition data is primarily based on choosing the order with the most high-frequency information to improve data resolution, and then selecting 1-2 orders close to the original seismic data for reconstruction. The reconstructed data retains the main effective information of the original data while improving resolution. In this embodiment, order 1 and order 2 seismic decomposition data are selected and overlaid to obtain high-frequency seismic reconstruction data of order 1+2. A profile comparison between the reconstructed data of order 1+2 and the original seismic data shows that the reconstructed data of order 1+2 retains the phase axis morphology of the original seismic data while significantly improving resolution. Therefore, the reconstructed data of order 1+2 is used as the high-resolution post-stack seismic data for the following fracture prediction operation.

[0054] S04: Perform ant-body attribute calculation based on variance volume on high-resolution seismic data to obtain ant-body seismic attributes.

[0055] This invention utilizes variance ant body fusion attribute technology. This technology first requires smoothing high-resolution 1+2 level seismic reconstruction data to eliminate discontinuities in phase axes caused by stratigraphic tilt and noise, and to highlight fracture and crack features. Next, the variance value σ of the structural interpretation seismic data needs to be calculated. 2 The formula is as follows:

[0056]

[0057] In the formula: i refers to the number of seismic traces, which is a real number;

[0058] j refers to the time for variance calculation;

[0059] I refers to the stratigraphic level and number of fault lines used when calculating variance;

[0060] L refers to the time window length during variance calculation;

[0061] w refers to the trigonometric weighting function of the seismic variance volume, with a range of [0, 1].

[0062] x refers to the average amplitude value over the variance calculation period.

[0063] The structural interpretation of the seismic variance value can be obtained from the above formula. Ant tracking technology is then used on the seismic variance data volume. The advantage of ant tracking technology is that it can transform the seismic variance value anomaly into fault boundaries that extend longitudinally and match medium to small-order fractures.

[0064] S05: Extract planar slices of the target layer from the seismic attributes of the ant body to obtain crack prediction results.

[0065] After performing the variance-ant body attribute fusion technique, the variance-ant body attribute data volume needs to be sliced ​​into planar slices of the target layer. Because the ant body technique not only identifies fault and fracture information but may also amplify noise, a preliminary assessment of the accuracy of fracture prediction in the target layer is required on the planar slice image.

[0066] During shale gas well drilling, lost circulation events are often associated with fractures and natural fissures. Information about medium to small-sized faults encountered during drilling directly confirms the existence of medium to small-order fractures in the formation. Therefore, to accurately evaluate fracture prediction in the target formation of this work area, it is necessary to collect drilling reports from the drilling teams in this area to identify information on lost circulation and encountered faults during actual drilling, thus verifying the actual effectiveness of fracture prediction. For more accurate verification, wellbore calibration can be performed on horizontal wells, and further verification can be conducted on seismic profiles. By continuously adjusting the variance ant body attribute parameters to update the fracture prediction results, the actual encountered faults and lost circulation information can be better matched. Finally, the ant body attribute parameters with the best fracture prediction effect for the target formation in this work area are selected, and a fracture distribution with a good match between the fracture density of the shale target formation and the actual encountered fractures and faults is obtained. When the consistency rate between the fracture prediction results and the actual encountered faults and lost circulation reaches more than 60%, the fracture prediction results are considered relatively reliable.

[0067] Exemplary Example 2

[0068] This exemplary embodiment provides a computer device including: at least one processor and a memory storing program instructions, wherein the program instructions are configured to be executed by the at least one processor, the program instructions including instructions for performing the shale reservoir fracture prediction method described in Exemplary Embodiment 1.

[0069] Exemplary Example 3

[0070] This exemplary embodiment provides a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, implement the method for predicting shale reservoir fractures in Exemplary Embodiment 1 above.

[0071] To better understand the above exemplary embodiment 1 of the present invention, it will be further described below with reference to specific examples.

[0072] Example 1

[0073] Within the study area, using the method of Exemplary Implementation 1, after performing data decomposition and theoretical frequency upscaling on the raw seismic data, seven different levels of information were obtained, and the seismic profiles are as follows: Figure 2 As shown, from Figure 2 It can be seen that level 1 has a high dominant frequency, significantly improving resolution and highlighting small-scale details in the seismic data. Level 2 improves resolution while retaining the main information of the original seismic data. Level 3 offers limited improvement in profile quality, and levels 4-7 have low resolution. Therefore, levels 3-7 are discarded in the analysis, and only the better levels 1 and 2 are retained for data recombination. After recombination, the high-spectral-optimized levels are obtained, and fracture prediction layer planar slices are generated using variance ant body fusion attribute technology, as shown in the figure. Figure 4 As shown, the study area is a planar slice of the predicted fracture horizon according to the method of the present invention; Figure 3 This study presents a stratigraphic planar slice map of fracture prediction using conventional methods in the research area, specifically a stratigraphic planar slice map of fracture prediction directly using the variance-ant-body fusion attribute technique based on the original post-stack seismic data. Figure 4 It can be clearly seen that the crack prediction results of the method of the present invention are much denser. The method of the present invention significantly improves the predicted crack density.

[0074] Figure 6A The image shows the matching between the fracture prediction profile and the actual well leakage point in the drilled well using conventional methods. The predicted fracture location on the profile does not match the well leakage point location. Figure 6B The matching between the fracture prediction profile and the actual well leakage point of the method of this invention is shown. The fracture prediction results are denser, and the predicted fracture locations match the well leakage point locations. The profile matching rate of the method of this invention is much higher than that of conventional methods. The matching rate of conventional methods is about 48%, while the matching rate of the updated method of this invention can reach 72.3%, which is about 25% higher than that of conventional methods, and the fracture density is significantly increased. This proves that the fracture content of the shale reservoir fracture prediction method based on data decomposition theory of this invention is richer.

[0075] The method of this invention significantly improves the accuracy of fracture prediction in the work area. The key is to constrain and verify the stratigraphic planar slices predicted by variance-ant body fusion attribute technology based on well leakage and fault information in actual drilling. By continuously modifying the variance-ant body fusion attribute parameters, the accuracy rate of fracture prediction can exceed 70%. Figure 5 The planar fracture prediction results of the method of the present invention are shown. The points on the well trajectory represent the locations of well leakage and faults encountered during drilling. The high consistency between the predicted fracture results and the actual locations of well leakage and faults in the planar plane indicates the effectiveness of the method. Figure 6A and Figure 6B A comparison of cross-sectional views between conventional crack prediction methods and the method of this invention shows that the method of this invention is superior to conventional methods.

[0076] Although the invention has been described above in conjunction with exemplary embodiments, those skilled in the art will understand that various modifications and changes can be made to the exemplary embodiments of the invention without departing from the spirit and scope defined by the claims.

Claims

1. A method of predicting fractures in a shale reservoir, the method comprising: The prediction method comprises the following steps: decomposing the post-stack seismic data based on a multi-order multi-azimuth method to obtain N different order seismic decomposition data; performing spectral analysis and direction optimization on the seismic decomposition data of each order, selecting order x seismic decomposition data with the most high-frequency information and at least one order seismic decomposition data with the most original seismic effective information; superimposing and combining the order x and the at least one order seismic decomposition data to obtain high-resolution seismic data; performing variance body-based ant body attribute calculation on the high-resolution seismic data to obtain ant body seismic attributes; performing plane slice extraction of the target horizon on the ant body seismic attributes to obtain a fracture prediction result; The fracture prediction result is a horizon plane slice. The method further comprises the step of: constraining the variance body-based ant body attribute calculation by using actual drilled faults and well leakage information, and continuously adjusting the ant body attribute parameters so that the coincidence rate of the horizon plane slice and the actual drilled faults and well leakage is greater than a preset value, to obtain a shale reservoir fracture prediction result. The method further comprises: before the step of constraining, collecting well leakage and drilled fault information of the target horizon horizontal well and projecting the information onto the horizon plane slice. The most original seismic effective information or the first and second most original seismic effective information is retained.

2. The method of claim 1, wherein, The method further comprises the step of: before the step of decomposing the post-stack seismic data, evaluating the post-stack seismic data, and in the case of poor evaluation, pre-processing the post-stack seismic data.

3. The method of claim 2, wherein, The evaluation comprises determining the signal-to-noise ratio, and the pre-processing comprises at least one of denoising, flattening and amplitude equalization.

4. The method of claim 1, wherein, N=7, order x is order 1, and the at least one order is order 2.

5. The method of claim 1, wherein, The preset value is more than 50%.

6. A computer device, comprising: It comprises: at least one processor, a memory storing program instructions, wherein the program instructions are configured to be executed by the at least one processor, and the program instructions comprise instructions for executing the method according to any one of claims 1-5.

7. A computer-readable storage medium having stored thereon computer program instructions, wherein, The computer program instructions are executed by the processor to realize the method of any one of claims 1-5.

Images