Fracture prediction method for horizontal target body section of deep carbonate rock horizontal well

Abstract

The invention discloses a crack prediction method for a horizontal target body section of a deep carbonate rock horizontal well, and the method comprises the steps: obtaining high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data based on a harmonic theory high-frequency recovery technology, and carrying out the weighted reconstruction of dip angle guiding filtering according to signal spectrum features; the method comprises the following steps: decomposing a frequency spectrum into low-frequency, intermediate-frequency and high-frequency bands by using a least square method, selecting single-frequency bodies capable of respectively representing the low-frequency, intermediate-frequency and high-frequency spectrums, carrying out dip angle scanning by using a discrete dip angle scanning maximum correlation method, carrying out dominant orientation ant tracking, and highlighting a micro-crack conforming to a structural activity background of a target stratum in a geological historical period; the method comprises the following steps of: selecting a multi-scale fracture fusion data body, filtering out false fractures which do not conform to the data body, and outputting the high-precision multi-scale fracture fusion data body so as to provide a multi-scale fracture result with fracture trend, development density and spatial distribution condition which are matched with production and while-drilling tracking drilling well leakage and gas logging abnormity, and guide drilling engineering risk evaluation.

Description

Technical Field

[0001] The invention belongs to the technical field of petroleum geological exploration and drilling well control, and in particular relates to a method for predicting fractures in a horizontal target section of a deep carbonate rock horizontal well. Background Art

[0002] China's deep ancient carbonate formations are rich in natural gas resources, but due to the complex conditions of deep burial, thin reservoir thickness, strong heterogeneity, high degree of densification, and superposition and transformation of multiple tectonic movements, natural gas exploration and development are difficult. In many basins in China, deep ancient carbonate formations are generally well developed due to the multi-stage activation of deep basement strike-slip faults and associated structures. During drilling, large amounts of drilling fluid are often lost or even malignant well leakage problems occur, resulting in complex underground engineering conditions and challenges in leakage prevention and plugging. There is an urgent need to improve the accuracy of fracture prediction in deep carbonate formations.

[0003] Many oilfield scientific and technological workers have proposed a variety of geophysical methods to predict fractures, but the prediction results are less than 50% consistent with the actual drilling encounters, and there is still a lack of relatively accurate methods for high-precision fracture prediction of ancient deep carbonate formations. Therefore, high-precision multi-scale fracture prediction in the horizontal target section of deep horizontal wells has become a bottleneck for drilling risk assessment in deep ancient carbonate formations. For example, the carbonate formations of the Upper Sinian Dengying Formation in the central Sichuan Basin are particularly developed due to the activation of multi-stage basin tectonic activities of basement faults and associated structures. They are mainly high-angle fractures and staggered fractures, which are highly concealed and difficult to identify using conventional seismic prediction methods.

[0004] Many oilfield scientific and technological workers have predicted fractures in the carbonate formations of the Dengying Formation of the Upper Sinian System based on seismic geometric attributes (coherence technology, curvature technology), using cluster analysis methods to integrate multiple seismic attributes, and pre-stack data based on azimuthal anisotropy in the OVT (Offset Vector Tiles) domain. However, the prediction results only match about 50% with the actual drilling leakage and gas logging abnormal sections of horizontal wells, which cannot effectively guide the risk assessment of drilling projects and the countermeasures for leakage.

[0005] In summary, the fracture results obtained using the existing 3D seismic data and the commonly used geophysical attribute fracture prediction methods are not consistent with the drilling encounters during drilling tracking, especially in identifying small-scale fractures. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes a fracture prediction method for the horizontal target section of a deep carbonate horizontal well, the fracture results have a high consistency rate with the drilling tracking, and are particularly effective in identifying small-scale fractures.

[0007] To achieve the above object, the technical solution adopted by the present invention is: a method for predicting fractures in a horizontal target section of a deep carbonate horizontal well, comprising the steps of: S10, based on the high-frequency recovery technology of harmonic theory, the original 3D prestack depth migration seismic data is used to obtain high-fidelity and high-resolution 3D prestack depth migration seismic data; nonlinear dip-guided filtering is selectively performed on different frequency bands of the seismic data, and then weighted reconstruction is performed to highlight the contribution rate of effective signals in different frequency bands; S20, according to the actual drilling tracking requirements for multi-scale fractures, and the different sensitivities of seismic data bodies of different frequencies to fractures of different scales, the signal spectrum characteristics of the dip-guided filtering weighted reconstruction data are decomposed into low frequency, medium frequency and high frequency using the least squares method, and low frequency fracture detection, medium frequency fracture detection and high frequency fracture detection are performed respectively; S30, selecting single frequency bodies that can represent low-frequency, medium-frequency and high-frequency spectra respectively, performing inclination scanning by discrete inclination scanning maximum correlation method, and performing ant tracking in advantageous directions, wherein low-frequency data bodies are conducive to identifying large-scale cracks, medium-frequency data bodies are conducive to identifying medium-scale cracks, and high-frequency data bodies are conducive to identifying small-scale cracks; S40, filters out false fractures whose low-frequency, medium-frequency and high-frequency single-frequency body ant tracking results are inconsistent with the tectonic activity background of the target layer in the geological history period, and highlights the micro-fractures that match the production and drilling tracking drilling situation; integrates the low-frequency, medium-frequency and high-frequency single-frequency body fracture information, extracts fractures of various scales, and finally fuses them into a high-precision multi-scale fracture data body.

[0008] Furthermore, based on the high-frequency recovery technology of harmonic theory, the original three-dimensional pre-stack depth migration seismic data is decomposed into various fundamental frequency signals by wavelet transform using seismic signals within the effective frequency band. The harmonics and subharmonics of each fundamental wave are calculated according to the harmonic criterion, and then added back to the original wavelet coefficients. The inverse wavelet transform is performed to realize the recovery of the missing high and low frequencies of the signal, so as to obtain high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data.

[0009] Furthermore, multi-band nonlinear dip-guided filtering is performed on the high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data, the data is expanded into multi-band frequency division, and different parameters are selected for each band data to perform nonlinear dip-guided filtering, and then weighted reconstruction is performed to highlight the contribution rate of effective signals in different frequency bands.

[0010] Furthermore, based on the results of the dip-guided filtering weighted reconstruction data, short-time Fourier transform, generalized S transform or wave transform methods are used to perform time-frequency analysis to generate time-frequency data volumes or single-frequency data volumes, highlighting the characteristics of different scales of cracks hidden in different frequency bands in conventional seismic data; using the least squares spectral decomposition technology, the frequency domain of the dip-guided filtering weighted reconstruction data volume is decomposed into three frequency bands: high frequency, medium frequency and low frequency.

[0011] Furthermore, if the crack detection results retained after filtering are consistent with the production and actual drilling crack exposure, the frequency-divided crack detection data is directly output; otherwise, the spectrum decomposition link is returned to further optimize the frequency-divided frequency band, and the operation is repeated in sequence until the false cracks are filtered out and the retained crack results match the production and drilling and tracking drilling conditions, and the frequency-divided crack detection data is output.

[0012] Furthermore, whether the initial prediction results match the production and drilling and tracking of well leakage, if so, frequency division multi-scale fracture fusion is performed to obtain a high-precision multi-scale fracture data body of fracture direction, development density and spatial distribution that matches the production and drilling and tracking of well leakage and gas logging anomalies; if not, spectrum decomposition is re-performed, and low-frequency, medium-frequency and high-frequency frequency domain bands are optimized, and the above operations are repeated until the fracture prediction results match the drilling and tracking of well leakage, and then high-precision multi-scale fracture data is output.

[0013] The beneficial effects of adopting this technical solution are: The present invention provides a multi-scale fracture prediction result with fracture direction, development density and spatial distribution matching production and drilling-while-drilling tracking of lost circulation and gas logging anomalies, thereby improving the fracture prediction accuracy of the carbonate rock formation of the Upper Sinian Dengying Formation, effectively guiding the efficient implementation of drilling engineering operations, and taking timely measures to deal with lost circulation.

[0014] Through the recent positive guidance wells and just completed drilling, the consistency rate of multi-scale fracture prediction results was statistically analyzed. Fractures were predicted 18 times in 6 wells, of which 15 were encountered in drilling, with a consistency rate of up to 83.3%. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a schematic flow chart of a fracture prediction method for a horizontal target section of a deep carbonate horizontal well according to the present invention; Figure 2 It is a high-frequency recovery principle and technical diagram based on harmonic theory of an embodiment of the present invention; Figure 2 (a) represents harmonics and sub-waves, Figure 2 (b) represents harmonics and frequency band extension, Figure 2 (c) represents the calculation principle of harmonic high-frequency and low-frequency expansion, Figure 2 (d) Harmonic high frequency restoration technology process; Figure 3 It is a comparison diagram of the frequency spectrum of the prestack depth migration seismic data volume before and after the harmonic high frequency recovery processing dip-guided filtering weighted reconstruction according to an embodiment of the present invention; Figure 4 It is a comparison diagram of the spectrum decomposition and frequency division scanning of prestack depth migration seismic data before and after the harmonic high frequency recovery processing dip-guided filtering weighted reconstruction according to an embodiment of the present invention; Figure 5 This is a cross-sectional diagram of multi-scale crack identification and fusion in three frequency bands: low frequency, medium frequency and high frequency according to an embodiment of the present invention; Figure 6 It is a multi-scale crack identification and fusion plan view of three frequency bands of low frequency, medium frequency and high frequency in an embodiment of the present invention; Figure 7 This is a cross-sectional diagram of the effectiveness of a well for risk assessment verification of a tracking-while-drilling project according to an embodiment of the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described below with reference to the accompanying drawings.

[0017] In this embodiment, see Figure 1 As shown, the present invention proposes a method for predicting fractures in a horizontal target section of a deep carbonate horizontal well, comprising the steps of: S10, based on the high-frequency recovery technology of harmonic theory, the original 3D prestack depth migration seismic data is used to obtain high-fidelity and high-resolution 3D prestack depth migration seismic data; nonlinear dip-guided filtering is selectively performed on different frequency bands of the seismic data, and then weighted reconstruction is performed to highlight the contribution rate of effective signals in different frequency bands; S20, according to the actual drilling tracking requirements for multi-scale fractures, and the different sensitivities of seismic data bodies of different frequencies to fractures of different scales, the signal spectrum characteristics of the dip-guided filtering weighted reconstruction data are decomposed into low frequency, medium frequency and high frequency using the least squares method, and low frequency fracture detection, medium frequency fracture detection and high frequency fracture detection are performed respectively; S30, selecting single frequency bodies that can represent low-frequency, medium-frequency and high-frequency spectra respectively, performing inclination scanning by discrete inclination scanning maximum correlation method, and performing ant tracking in advantageous directions, wherein low-frequency data bodies are conducive to identifying large-scale cracks, medium-frequency data bodies are conducive to identifying medium-scale cracks, and high-frequency data bodies are conducive to identifying small-scale cracks; S40, filters out false fractures whose low-frequency, medium-frequency and high-frequency single-frequency body ant tracking results are inconsistent with the tectonic activity background of the target layer in the geological history period, and highlights the micro-fractures that match the production and drilling tracking drilling situation; integrates the low-frequency, medium-frequency and high-frequency single-frequency body fracture information, extracts fractures of various scales, and finally fuses them into a high-precision multi-scale fracture data body.

[0018] As an optimization scheme for the above-mentioned embodiment, the original three-dimensional pre-stack depth migration seismic data is decomposed into various fundamental frequency signals by wavelet transform based on the harmonic theory high-frequency recovery technology and utilizing the seismic signals within the effective frequency band. The harmonics and subharmonics of each fundamental wave are calculated according to the harmonic criterion, and are added back to the original wavelet coefficients. The inverse wavelet transform is performed to realize the recovery of the missing high and low frequencies of the signal, thereby obtaining high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data.

[0019] Specifically, Figure 2 The high frequency recovery principle and technical flow chart based on harmonic theory of the embodiment of the present invention are shown as follows: Figure 2 As shown, Figure 2 a shows five sub-waves with different frequency components, all of which are composed of the superposition of the fundamental frequency cosine signal and its higher harmonic cosine signal; from sub-wave 1 to 5, the higher harmonic components increase successively, which eventually leads to the superimposed synthesized sub-waves becoming sharper and sharper, that is, the frequency band becomes wider and wider, and the resolution becomes higher and higher, such as Figure 2 As shown in b; high-frequency expansion = calculation of the harmonics of the original signal (the frequency is an integer multiple of the fundamental frequency signal), low-frequency expansion = calculation of the subharmonics of the original signal (the frequency is a fractional multiple of the fundamental frequency signal), such as Figure 2 c; the specific technical implementation process of high-frequency recovery processing of harmonic theory, as shown in Figure 2 As shown in d.

[0020] As an optimization scheme for the above embodiment, multi-band nonlinear dip-guided filtering is performed on the high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data, the data is expanded into multi-band frequency division, and different parameters are selected for each band data to perform nonlinear dip-guided filtering, and then weighted reconstruction is performed to highlight the contribution rate of effective signals in different frequency bands.

[0021] Specifically, the frequency-band dip-guided filtering mainly utilizes the wavelet packet decomposition method to perform multi-band frequency division expansion on the full-band high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data volume obtained by harmonic high-frequency recovery processing, and selects different parameters for each frequency band data to perform nonlinear dip-guided filtering. The filtered seismic data of each frequency band are weighted reconstructed to highlight the contribution rate of effective signals in different frequency bands.

[0022] As an optimization scheme for the above-mentioned embodiment, according to the results of the dip-guided filtering weighted reconstruction data, short-time Fourier transform, generalized S transform or wave transform method are used to perform time-frequency analysis to generate a time-frequency data body or a single-frequency data body, highlighting the different scales of crack characteristics hidden in different frequency bands in conventional seismic data; using the least squares spectral decomposition technology, the frequency domain of the dip-guided filtering weighted reconstruction data body is decomposed into three frequency bands: high frequency, medium frequency and low frequency.

[0023] Figure 3 : is a comparison diagram of the prestack depth migration seismic data volume spectrum before and after the harmonic high frequency recovery processing dip-guided filtering weighted reconstruction according to an embodiment of the present invention, such as Figure 3 As shown in the figure, after the harmonic high-frequency recovery processing and dip-guided filtering weighted reconstruction, the spectrum bandwidth of the 3D prestack depth migration seismic data is 5-50Hz. Compared with the spectrum characteristics of the original 3D prestack depth migration seismic data, the low-frequency signal remains unchanged, and the high frequency is widened by about 20Hz (the high-frequency information is real and effective). Figure 4 1 is a comparison diagram of the frequency spectrum decomposition and frequency division scanning of prestack depth migration seismic data before and after the harmonic high frequency recovery processing dip-guided filtering weighted reconstruction according to an embodiment of the present invention, such as Figure 4 As shown in the figure, the least squares spectrum decomposition technology is used to decompose the dip frequency domain of the "double-high" three-dimensional pre-stack depth migration seismic data into three frequency bands: low frequency (0-20Hz), medium frequency (20-40Hz), and high frequency (40-60Hz).

[0024] Based on the above-mentioned spectrum decomposition results, single frequency bodies that can represent low-frequency (20Hz), medium-frequency (35Hz), and high-frequency (50Hz) spectra are selected, and the discrete scanning maximum correlation method is used to perform inclination scanning. Based on the ant tracking technology, ant tracking in the dominant direction is carried out; false fractures that do not match the tectonic activity background of the target layer in the geological history period are filtered out, and micro-cracks that are consistent with the tectonic activity background of the target layer in the geological history period are highlighted.

[0025] As an optimization scheme for the above embodiment, if the crack detection results retained after filtering are consistent with the production and actual drilling crack exposure conditions, the frequency-divided crack detection data is directly output; otherwise, the spectrum decomposition link is returned to further optimize the frequency-divided frequency bands, and the operation is repeated in sequence until the false cracks are filtered out and the retained crack results match the production and drilling tracking and exposure conditions, and the frequency-divided crack detection data is output.

[0026] Specifically, whether the initial prediction results match the production and drilling and tracking of well leakage, if so, frequency division multi-scale fracture fusion is performed to obtain a high-precision multi-scale fracture data body of fracture direction, development density and spatial distribution that matches the production and drilling and tracking of well leakage and gas logging anomalies; if not, spectrum decomposition is re-performed, and low-frequency, medium-frequency and high-frequency frequency domain bands are optimized, and the above operations are repeated until the fracture prediction results match the drilling and tracking of well leakage, and then high-precision multi-scale fracture data is output.

[0027] Fusion of frequency-division multi-scale crack prediction results: Figure 5 : is a multi-scale crack identification and fusion profile diagram of low frequency, medium frequency and high frequency bands in an embodiment of the present invention, such as Figure 5As shown in the figure, the crack distribution, scale and density profiles detected by three frequency bands, low frequency, medium frequency and high frequency, are optimally fused to obtain the multi-scale crack fusion profile results, so as to realize the accurate prediction of the crack body.

[0028] Figure 6 is a multi-scale crack identification and fusion plan view of low frequency, medium frequency and high frequency bands in an embodiment of the present invention, such as Figure 6 As shown in the figure, the crack distribution, scale and density planes detected by three frequency bands, low frequency, medium frequency and high frequency, are optimally fused to obtain the multi-scale crack fusion plane results, so as to realize the accurate prediction of the crack body.

[0029] Matching rate statistics and drilling tracking verification: Table 1 shows the statistical coincidence rate of the high-precision multi-scale fracture prediction of the Upper Sinian Dengying Formation in the recent pilot wells being drilled and the wells just completed with the embodiments of the present invention. As shown in Table 1, fractures were predicted 18 times in 6 wells, of which 15 were encountered during drilling, and the actual coincidence rate was 83.3%.

[0030] Table 1 Figure 7 : is a cross-sectional diagram of the risk assessment verification well of the tracking-while-drilling engineering according to an embodiment of the present invention, such as Figure 7 As shown in the figure, the GSC and GSD drilling projects verify that the fracture prediction results of the method of the present invention are reliable and the effect is more significant.

[0031] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for predicting fractures in a horizontal target section of a deep carbonate horizontal well, characterized in that: Includes steps: S10, obtaining high-fidelity and high-resolution three-dimensional prestack depth migration seismic data based on the harmonic theory high-frequency recovery technology; Selectively perform nonlinear dip-guided filtering on different frequency bands of seismic data, and then perform weighted reconstruction to highlight the contribution rate of effective signals in different frequency bands; S20, according to the actual drilling tracking requirements for multi-scale fractures, and the different sensitivities of seismic data bodies of different frequencies to fractures of different scales, the signal spectrum characteristics of the dip-guided filtering weighted reconstruction data are decomposed into low frequency, medium frequency and high frequency using the least squares method, and low frequency fracture detection, medium frequency fracture detection and high frequency fracture detection are performed respectively; S30, selecting single frequency bodies that can represent low-frequency, medium-frequency and high-frequency spectra respectively, performing inclination scanning by discrete inclination scanning maximum correlation method, and performing ant tracking in advantageous directions, wherein low-frequency data bodies are conducive to identifying large-scale cracks, medium-frequency data bodies are conducive to identifying medium-scale cracks, and high-frequency data bodies are conducive to identifying small-scale cracks; S40, filters out false fractures whose low-frequency, medium-frequency and high-frequency single-frequency body ant tracking results are inconsistent with the tectonic activity background of the target layer in the geological history period, and highlights the micro-fractures that match the production and drilling tracking drilling situation; integrates the low-frequency, medium-frequency and high-frequency single-frequency body fracture information, extracts fractures of various scales, and finally fuses them into a high-precision multi-scale fracture data body.

2. The method for predicting fractures in a horizontal target section of a deep carbonate horizontal well according to claim 1, characterized in that: The original three-dimensional pre-stack depth migration seismic data is decomposed into fundamental frequency signals by wavelet transform based on the high-frequency recovery technology of harmonic theory and using the seismic signals in the effective frequency band. The harmonics and subharmonics of each fundamental wave are calculated according to the harmonic criterion, added to the original wavelet coefficients, and the inverse wavelet transform is performed to realize the high and low frequency recovery of the missing signals, so as to obtain high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data.

3. The method for predicting fractures in a horizontal target section of a deep carbonate horizontal well according to claim 2, characterized in that: Multi-band nonlinear dip-guided filtering is performed on the high-fidelity and high-resolution three-dimensional pre-stack depth migration seismic data. The data is expanded into multi-band frequencies, and different parameters are selected for each band data to perform nonlinear dip-guided filtering, which is then weighted and reconstructed to highlight the contribution rate of effective signals in different frequency bands.

4. The method for predicting fractures in a horizontal target section of a deep carbonate horizontal well according to claim 3, characterized in that: According to the results of the dip-guided filtering weighted reconstruction data, short-time Fourier transform, generalized S transform or wave transform methods are used for time-frequency analysis to generate time-frequency data volumes or single-frequency data volumes, highlighting the characteristics of different scales of cracks hidden in different frequency bands in conventional seismic data; using the least squares spectral decomposition technology, the frequency domain of the dip-guided filtering weighted reconstruction data volume is decomposed into three frequency bands: high frequency, medium frequency and low frequency.

5. The method for predicting fractures in a horizontal target section of a deep carbonate horizontal well according to claim 1, characterized in that: If the fracture detection results retained after filtering are consistent with the production and actual drilling cracks, the frequency-divided fracture detection data is directly output; otherwise, the spectrum decomposition link is returned to further optimize the frequency-divided frequency bands, and the operation is repeated in sequence until the fracture results retained after the false fractures are filtered out match the production and drilling and tracking conditions, and the frequency-divided fracture detection data is output.

6. The method for predicting fractures in a horizontal target section of a deep carbonate horizontal well according to claim 5, characterized in that: Whether the initial prediction results match the production and the well leakage encountered by drilling tracking, if they match, the frequency division multi-scale fracture fusion is performed to obtain the high-precision multi-scale fracture data volume of the fracture trend, development density and spatial distribution that matches the production and the well leakage encountered by drilling tracking and gas logging anomalies; if they do not match, the spectrum decomposition is re-performed, and the low-frequency, medium-frequency and high-frequency frequency domain bands are optimized. The above operations are repeated until the fracture prediction results match the well leakage encountered by drilling tracking, and then the high-precision multi-scale fracture data is output.

Images