A sand body distribution prediction method, device and medium

Abstract

The present invention provides a sandbody distribution prediction method, device, and medium, relating to the field of seismic exploration technology. The method comprises: obtaining seismic response waves and logging data alongside a wellbore in a target area; using the seismic response waves and logging data alongside the wellbore, using a singular value decomposition algorithm, determining the correspondence between the seismic response wave components and different sedimentary facies in the target area; constructing a likelihood function describing the planar distribution probability of the sedimentary facies using the correspondence between the seismic response wave components and different sedimentary facies in the target area as prior information; determining the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response waves in the target area; grouping the seismic response waves corresponding to different sediments based on the planar distribution probabilities of different sedimentary facies to obtain multiple sample sets; and performing seismic wave inversion on each sample set to obtain the vertical thickness and planar distribution of the sandbody in the target area. The sandbody distribution prediction method provided by the present invention improves the lateral resolution of the sandbody.

Description

Technical Field

[0001] The present invention relates to the field of seismic exploration technology, and in particular to a sand body distribution prediction method, equipment and medium. Background Art

[0002] Delineating sandbody boundaries is crucial for oil and gas exploration. Existing sandbody distribution prediction methods suffer from low lateral resolution, resulting in fuzzy delineation of sandbody boundaries with varying thicknesses (i.e., varying sandbody types), making it difficult to accurately identify their distribution and morphology. This is particularly true when sandbody types vary rapidly across the sandbody, where fuzzy boundary delineation makes it impossible to predict sandbody distribution. Therefore, a sandbody distribution prediction method with high lateral resolution is urgently needed to accurately delineate sandbodies with lateral phase changes, enabling in-depth oil and gas exploration and development. Summary of the Invention

[0003] The present invention provides a sand body distribution prediction method, equipment and medium, which are used to solve the defect of low lateral resolution in the prior art and achieve high-precision characterization of sand bodies with lateral phase change.

[0004] The present invention provides a sand body distribution prediction method, comprising the following steps:

[0005] Obtain seismic response waves and logging data beside the wellbore in the target area;

[0006] Based on the wellbore side seismic response waves and the well logging data, a singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies;

[0007] Using the correspondence between the seismic response wave components and different sedimentary facies in the target area as prior information, a likelihood function is constructed to describe the probability of sedimentary facies plane distribution.

[0008] determining the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response wave of the target area;

[0009] Based on the planar distribution probability of different sedimentary phases, the seismic response waves corresponding to different sediments are grouped to obtain multiple sample sets;

[0010] Seismic wave inversion is performed based on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area.

[0011] A sand body distribution prediction method provided by the present invention further includes:

[0012] Obtaining original seismic waveform data of the target area;

[0013] Based on a matching pursuit algorithm, the coal seam reflection wave is filtered out from the original seismic waveform data to obtain the seismic response wave.

[0014] According to a sand body distribution prediction method provided by the present invention, based on a matching pursuit algorithm, coal seam reflection waves are filtered out from the original seismic waveform data, comprising:

[0015] Performing seismic forward modeling on the target area to obtain simulated reflection waves of the coal seams in the target area;

[0016] Based on a matching pursuit algorithm, the simulated reflection wave is used as a matching wavelet to perform spectral decomposition on the original seismic wave to obtain seismic wavelets at different frequencies; the seismic wavelets include a target seismic wavelet that best matches the simulated reflection wave;

[0017] Based on the seismic wavelets other than the target seismic wavelet, the seismic wave is reconstructed to obtain the seismic response wave.

[0018] According to a sand body distribution prediction method provided by the present invention, the well logging data includes: well logging sand body thickness data;

[0019] Based on the wellbore side seismic response waves and the well logging data, a singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies, including:

[0020] Based on the well bypass seismic response wave and the well logging sand body thickness data, a singular value decomposition algorithm is used to determine the correspondence between the well bypass seismic response wave components and different sedimentary facies;

[0021] The corresponding relationship between the seismic response wave components in the inter-well area and the different sedimentary facies is determined based on the corresponding relationship between the seismic response wave components in the well bypass and the different sedimentary facies.

[0022] According to a sand body distribution prediction method provided by the present invention, the likelihood function is as follows:

[0023]

[0024] Indicates the i Earthquake waves, Indicates the i The disturbance amount of the elastic parameters corresponding to the seismic wave, the elastic parameters include thickness, I is the prior information, w represents the seismic wavelet, and N represents the number of seismic waves.

[0025] According to a sand body distribution prediction method provided by the present invention, the seismic wave inversion is seismic wave inversion based on phase-controlled waveform.

[0026] According to a sand body distribution prediction method provided by the present invention, the seismic wave inversion includes:

[0027] Constructing a posterior probability distribution function based on the likelihood function and the prior probability distribution function;

[0028] Seismic wave inversion is performed based on the posterior probability distribution.

[0029] According to a sand body distribution prediction method provided by the present invention, the seismic wave inversion includes:

[0030] Selecting a lithology sensitivity curve based on a correlation analysis between the well logging data and lithology;

[0031] The lithologic sensitivity curve is used as prior information to invert the seismic wave.

[0032] The present invention also provides a sand body distribution prediction device, comprising the following modules:

[0033] The data acquisition module is used to obtain seismic response waves and logging data beside the wellbore in the target area.

[0034] The singular value decomposition module is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies based on the seismic response wave beside the wellbore and the well logging data by using a singular value decomposition algorithm.

[0035] The likelihood function construction module is used to construct a likelihood function for describing the probability of sedimentary facies planar distribution based on the correspondence between the seismic response wave components and different sedimentary facies in the target area as prior information.

[0036] The distribution probability determination module is used to determine the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response wave of the target area.

[0037] The sample grouping module is used to group the seismic response waves corresponding to different sediments based on the planar distribution probability of different sedimentary phases to obtain multiple sample sets.

[0038] The sand body distribution prediction module is used to perform seismic wave inversion based on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area.

[0039] The present invention also provides an electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, any of the above-described sand body distribution prediction methods is implemented.

[0040] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the method for predicting sand body distribution as described above is implemented.

[0041] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements any of the above-mentioned sand body distribution prediction methods.

[0042] The sandbody distribution prediction method, device, and medium provided by the present invention utilize seismic response waves and well logging data from well bypasses in the target area, employing a singular value decomposition algorithm to determine the correspondence between seismic response wave components and different sedimentary facies (different types of sandbody). Based on this correspondence, the planar distribution probabilities of different sedimentary facies are determined. The seismic response waves corresponding to the different sedimentary facies on the plane are then grouped, and inversion is performed separately for each group of seismic response waves, ultimately determining the vertical thickness and planar distribution of the sandbody in the target area. In other words, by determining the planar distribution probabilities of different sedimentary facies, grouping the seismic response waves corresponding to the different sedimentary facies, and then performing group inversion, the present invention achieves the ability to distinguish and invert each sedimentary facies horizontally, thereby ensuring lateral resolution. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0044] Figure 1 It is a flow chart of the sand body distribution prediction method provided by an embodiment of the present invention.

[0045] Figure 2-a It is a geological model diagram in the two-dimensional forward modeling provided by an embodiment of the present invention.

[0046] Figure 2-b It is a forward seismic profile diagram in the two-dimensional forward modeling provided by an embodiment of the present invention.

[0047] Figure 2-c A graph showing the amplitude and frequency at the maximum peak of a seismic waveform in a two-dimensional forward modeling according to an embodiment of the present invention.

[0048] Figure 3-a This is a 30 Hz time-domain forward seismic profile provided by an embodiment of the present invention.

[0049] Figure 3-b This is a 60 Hz time-domain forward seismic profile provided by an embodiment of the present invention.

[0050] Figure 3-c This is a 90 Hz time-domain forward seismic profile provided by an embodiment of the present invention.

[0051] Figure 3-dThis is a diagram of a 30 Hz time-domain forward theoretical model provided by an embodiment of the present invention.

[0052] Figure 3-e This is a diagram of a 60 Hz time-domain forward theoretical model provided by an embodiment of the present invention.

[0053] Figure 3-f This is a diagram of a 90 Hz time-domain forward theoretical model provided by an embodiment of the present invention.

[0054] Figure 4-a This is a through-well seismic profile diagram of Well A and Well B provided in an embodiment of the present invention.

[0055] Figure 4-b These are seismic waveform diagrams of Well A and Well B provided by an embodiment of the present invention.

[0056] Figure 5 3 is a sand body thickness distribution histogram provided by an embodiment of the present invention.

[0057] Figure 6 This is a sand body lithology interpretation diagram provided by an embodiment of the present invention.

[0058] Figure 7 This is a seismic phase-logging phase characteristic map of the Benxi Formation in the study area provided by an embodiment of the present invention.

[0059] Figure 8 This is a quality control chart of the waveform indicating the optimal cutoff frequency provided by an embodiment of the present invention.

[0060] Figure 9 This is a phase-controlled waveform indication inversion diagram provided by an embodiment of the present invention.

[0061] Figure 10 This is a plan view of sandstone distribution provided by an embodiment of the present invention.

[0062] Figure 11 It is a structural schematic diagram of a sand body distribution prediction device provided by an embodiment of the present invention.

[0063] Figure 12 It is a structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0064] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0065] The following combination Figures 1-10The sand body distribution prediction method provided by an embodiment of the present invention is described.

[0066] Figure 1 It is a schematic flow chart of the sand body distribution prediction method provided by the present invention, such as Figure 1 As shown, the method includes the following:

[0067] Step 101: Obtain seismic response waves and logging data beside the wellbore in the target area.

[0068] Step 102: Based on the wellbore side seismic response waves and the well logging data, a singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies.

[0069] Step 103: Using the correspondence between the seismic response wave components in the target area and different sedimentary facies as prior information, a likelihood function is constructed to describe the probability of planar distribution of sedimentary facies.

[0070] Step 104: Determine the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response wave of the target area.

[0071] Step 105: Based on the planar distribution probabilities of different sedimentary phases, the seismic response waves corresponding to different sediments are grouped to obtain multiple sample sets.

[0072] Step 106: Perform seismic wave inversion based on each sample set to obtain the sand body distribution in the target area, wherein the sand body distribution includes vertical thickness and planar distribution.

[0073] It should be noted that the seismic response waves in this embodiment are multiple seismic waves distributed on a plane. Grouping the seismic response waves corresponding to different sedimentary formations is done in a planar space, where the seismic response waves within a particular sedimentary facies are grouped together. Different sedimentary facies refer to sedimentary facies with different thicknesses.

[0074] The inversion is performed based on each group of seismic response waves separately because each group of seismic impact waves has commonalities because they are response waves from the same sedimentary phase. It is understandable that each sedimentary phase has its own adaptive inversion parameters. In this embodiment, each group of seismic response waves can be inverted using its own adaptive inversion parameters during inversion, thereby achieving an accurate depiction of the sedimentary phase. Compared with the prior art in which sedimentary phases are not classified and inverted, and the same inversion parameters are used to simultaneously invert the seismic response waves corresponding to different sediments, thereby resulting in the inversion effects of different sedimentary phases not being able to achieve their own optimal results (because the same inversion parameters cannot take into account the inversion accuracy of all sedimentary phases), the sand body distribution prediction method provided by the embodiment of the present invention has a higher resolution and a higher inversion accuracy.

[0075] Specifically, different sedimentary facies have their own optimal choices for inversion parameters such as the optimal number of samples, optimal cutoff frequency, and smoothing radius.

[0076] The optimal number of samples represents the degree of influence of the spatial variation of the seismic response wave on the sedimentary facies. The selection of the optimal number of samples needs to consider the influence of the sedimentary facies and the number of wells in the target area. Taking the well spacing of 300 m as an example, the optimal number of samples for tidal channel microfacies is set to 5, the optimal number of samples for sand flats is set to 6, and the optimal number of samples for swamp-mud flats is set to 7.

[0077] The optimal cutoff frequency needs to take into account the thickness of the sand body in the target area. The thickness distribution range of the strata in the target area is mainly between 2m and 8m, and the average velocity of the strata is 5504 m / s. When the resolution is calculated at 1 / 4 wavelength and 1 / 8 wavelength, the optimal cutoff frequency for tidal channel microfacies is about 80 Hz, the optimal cutoff frequency for sand flats is about 170 Hz, and the optimal cutoff frequency for swamp-mud flats is about 200 Hz.

[0078] The smoothing radius affects the smoothness of the inversion results. A larger smoothing radius results in smoother inversion results, but may lose some important geological details. A smaller smoothing radius retains more details but may increase the randomness of the inversion results. The smoothing radius for tidal channel microfacies is 8, for sand flats it is 7, and for swamp-mud flats it is 6. By fine-tuning these key parameters, the inversion effect can be significantly improved, making the inversion results more consistent with the actual geological situation and providing more accurate geological information for oil and gas exploration and development.

[0079] The embodiment of the present invention combines well logging data, a singular value decomposition algorithm, a likelihood function, etc. when determining the planar distribution of different sedimentary facies, which will be described in detail later.

[0080] The embodiments of the present invention, through the above method, can accurately characterize sand bodies in target areas with different sedimentary facies. This can also be achieved in scenarios where the target area has unequal lateral sedimentary facies thicknesses, or even rapid lateral phase transitions. For example, in Jiaxian District, located in the northeastern part of the Yishan Slope in the Ordos Basin, the main development strata are the Permian Shihezi Formation, Shanxi Formation, Taiyuan Formation, and Carboniferous Benxi Formation. Compared with the Lower Permian Shanxi Formation and the Middle Permian Lower Shihezi Formation, the Upper Carboniferous Benxi Formation has greater potential for tight gas exploration and development. However, the Upper Carboniferous Benxi Formation is a set of sandstone deposits located beneath the coal seam, primarily tidal channel deposits in a tidal flat environment. The sand bodies are characterized by rapid lateral phase transitions, thin sand bodies, and low permeability and density. Existing inversion methods cannot accurately characterize the lateral distribution of sand bodies. However, the sand body distribution prediction method provided by the embodiments of the present invention can accurately characterize the lateral distribution of sand bodies by identifying and dividing the lateral sedimentary facies and then performing inversion separately.

[0081] In some embodiments of the present invention, to improve the accuracy of the prediction of the planar distribution of sedimentary facies, the seismic waves used for inversion are preprocessed. In summary, the raw seismic wave data of the target area is preprocessed to remove interference, thereby obtaining the seismic response waves of the target area in step 101. It should be noted that the seismic response waves described in step 101 are actually the response waves remaining after removing interference from the raw seismic response wave data.

[0082] Coal seam reflection waves, as strong interference information in the original seismic waveform data, seriously affect the inversion results. For example, through analysis of the reservoir characteristics and logging phases of the target layer in the target area (Jiaxian District), the target layer sand body thickness (ΔZ) is generally 2m to 15m, and the coal seam thickness is 7m to 14m, averaging around 14m. The main frequency of the seismic data is only 32Hz, and the effective bandwidth is 5Hz to 92Hz. The upper part of the frequency-divided profile is a set of 14-meter coal seams with strong reflection phase axis characteristics. This strong reflection shields the reflection characteristics of the lower sandstone, resulting in the distribution of the lower sandstone cannot be identified.

[0083] Therefore, the preprocessing of raw seismic wave data in this embodiment primarily involves filtering out coal seam reflections from the raw seismic waveform data. This filtered-out seismic data is then used for subsequent inversion. This not only allows thin sand bodies to be identified, allowing for accurate sedimentary facies distribution, but also eliminates interference, making the data used for inversion more accurate, leading to more accurate inversion results.

[0084] To filter out coal seam reflections, we primarily use a matching pursuit algorithm, supplemented by seismic forward modeling. Specifically, we first use seismic forward modeling to theoretically simulate coal seam reflections. Then, using the theoretical simulation results as constraints for the matching pursuit algorithm, we perform spectral decomposition on the raw seismic wave data, ultimately removing the filtered coal seam reflections from the raw seismic wave data.

[0085] The filtering of coal seam reflected waves is introduced in detail below.

[0086] (1) Earthquake forward modeling

[0087] Seismic forward modeling is the process of obtaining the corresponding seismic response based on a known or designed geological model. Seismic forward modeling can analyze the relationship between the lithology, physical properties and seismic response of the geological body, and verify the reliability of the inversion method and results. In this embodiment, we design a geological model based on the actual geological conditions, and understand the seismic reflection response of lithology changes through forward modeling. The following is an explanation with a specific example. The target area (the target area is, for example, a research area in Jiaxian District) has a target layer, the Benxi Formation sandstone thickness mainly between 2m and 15m, and there is a set of 8# coal seams with stable thickness on the top. Design a geological model based on the actual geological conditions, such as Figure 2-a As shown in the figure, the model is based on mudstone as the background, with a 14 m thick coal seam overlying it and two sets of sandstone underneath, namely a thick rectangular sand body and a thin trapezoidal sand body. The parameters of each lithology are set according to the actual situation, among which the average layer velocity of sandstone is 2800 m / s, the average layer velocity of mudstone is 3300 m / s, the average layer velocity of coal rock is 2200 m / s, and the density of sandstone is 2260 kg / m 3 The density of mudstone is 2650 kg / m 3 , the density of coal rock is 1500 kg / m 3 The 2D forward geological model is convolved with the statistical wavelet (with a main frequency of about 32 Hz) to synthesize a 2D seismic model, such as Figure 2-b The wavelet used is a statistical wavelet extracted from the Jiaxian block, and the interference from adjacent areas is quantitatively represented by the thickness and RMS amplitude of the river channel sand body. Based on this geological model, the reflection waveform of the coal seam in the coal-bearing strata is obtained through forward modeling.

[0088] (2) Spectral decomposition based on matching pursuit algorithm

[0089] Matching pursuit, as a wavelet decomposition method, can decompose the signal into a linear combination of multiple optimally matched wavelets. The reflection waveforms of the coal seams and sandstones in the coal-bearing strata obtained through forward modeling in the previous step can be effectively removed from the original seismic data using the matching pursuit algorithm, as shown in Figure 3. The process is as follows: the reflection waveforms of the coal seams in the coal-bearing strata obtained through forward modeling are used as the initial matching wavelets; the matching pursuit algorithm (MP) is used to iteratively search for the optimal time-frequency atom until a high similarity coefficient is achieved with the forward modeling result. This optimal wavelet is then used as the strong reflection of the coal seam in that trace; the matching waveform subtraction method is used to remove the strong reflection of the coal seam and obtain the seismic responses of the coal seams and sandstones at different frequencies.

[0090] Specifically, this embodiment adopts seismic matching pursuit time-frequency processing technology and uses the Ricker wavelet matching pursuit algorithm to complex analyze the seismic signal to obtain three seismic parameters: instantaneous amplitude, instantaneous phase, and instantaneous frequency. The original seismic wave data is decomposed into multiple seismic wavelets at different frequencies. The seismic wavelet includes the target seismic wavelet obtained by iteration above and that best matches the forward simulated coal seam reflection wave. Finally, based on the seismic wavelets other than the target seismic wavelet, superposition and reconstruction are performed to obtain the seismic response wave described in this article. The specific parameters involved are expressed as follows:

[0091] Assume that Ricker wavelet is used in earthquake forward modeling, with a dominant frequency of f j The expression of the zero-phase Ricker wavelet in the time domain is:

[0092] (1)

[0093] Among them, f j is the main frequency of the earthquake, unit Hz, t is the earthquake reflection time, unit s, W R (t,f j ) is the Ricker wavelet in the time domain; j is the number of iterations.

[0094] The sparse expression of seismic signal is:

[0095] (2)

[0096] Where S(t) is the time domain band-limited seismic signal, R s （m） (t) is the residual after matching pursuit, m is the number of iterative atoms determined by the set threshold to judge the iteration termination condition, and the number of atoms is determined by the time delay t j (Unit: s), earthquake main frequency f j (Unit: Hz), correlation coefficient a j and phase φ j And so on 4 processing parameters to control.

[0097] Through a matching pursuit optimization algorithm, the matching wavelet W corresponding to the target waveform (the reflection waveform of the coal seam and sandstone in the coal-bearing stratum obtained by seismic forward modeling) can be calculated. R (t,f j ) and the wavelet amplitude A, thereby obtaining the matching wavelet W that represents the reflected waveform. R (the target seismic wavelet mentioned above), i.e.

[0098] (3)

[0099] From the original seismic data S(t0, f j ) minus WR , and obtain the new seismic record S (t,f j ) (that is, the seismic response wave described in this article), which can highlight the weak reflection characteristics.

[0100] (4)

[0101] Where λ is the separation coefficient. By identifying and separating strong reflections, the weak reflection information can be better restored and the reservoir prediction accuracy can be improved.

[0102] Through the seismic matching pursuit waveform subtraction technology under the constraints of seismic forward modeling, strong reflections are removed and weak signals of sandstones are enhanced, providing favorable data for the prediction of subsequent seismic sand bodies. The effect of seismic matching pursuit waveform subtraction technology in removing strong reflections is closely related to the accuracy of the strong reflection matching wavelet. The Ricker wavelet or Morlet wavelet used in matching pursuit cannot match the actual complex strong reflection waveform well. Therefore, the reflection waveforms of coal seams and sandstones obtained by seismic forward modeling can be used as the initial matching wavelet, and continuous iterations can be used to finally obtain the reconstruction of seismic signals at different frequencies. In a certain study area in Jiaxian District mentioned above, it can be seen that the seismic profile of the coal seam reflection is removed by matching pursuit controlled by forward modeling, see Figure 3-a to Figure 3-f The dominant frequency corresponding to the 12m sand body is around 60Hz, where the tuning response is strongest, reducing the interference of coal rock on the sandstone. Spectral decomposition preserves the low-frequency information in the seismic data while expanding the dominant frequency from 32Hz to 60Hz, resulting in higher resolution. While 30Hz matching pursuit can only vaguely distinguish some sand body reflections when removing coal seam reflections, it effectively reduces the interference of coal rock on the sandstone at frequencies above 60Hz. The lateral variation of the event axis is reasonable, and the waveform is natural and distortion-free, effectively highlighting the sand body reflections.

[0103] After the above preprocessing, the seismic wave data with the coal seam reflection wave filtered out can effectively highlight the sand body reflection, thereby improving the accuracy of the inversion basic data and increasing the vertical resolution.

[0104] The following is a detailed introduction to the combination of logging data, the application of the singular value decomposition algorithm, and the construction of the likelihood function involved in improving the lateral resolution.

[0105] (1) Combining well logging data with the seismic response waves near the wellbore, the singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components and different sedimentary facies.

[0106] First, an isochronous stratigraphic framework is established through seismic horizon interpretation, and the seismic response waveforms of the wellside channels of two adjacent wells are extracted. After top surface alignment and thickness correction, the seismic response waveforms are compared. Figure 4-a and Figure 4-b As can be seen from the figure, the seismic waveforms of the two wells are 93% similar. It can be concluded that the seismic waveforms of two wells with similar sedimentary environments are also similar.

[0107] Then, the seismic waveforms are dynamically clustered according to the singular value decomposition to establish the initial seismic wave impedance model.

[0108] Based on the above, after removing the strong reflection of the coal seam, the seismic response wave is obtained. For the seismic data in a given window, it can be represented by an M×N matrix A (where M is the number of seismic waves, that is, the number of seismic waves, and N is the number of sample points). According to the principle of singular value decomposition (SVD), it can be expressed as:

[0109] (5)

[0110] Where: U is an M×N matrix, whose column U i AA T Orthogonal eigenvectors of V i is an N×N matrix, whose columns V i A T The orthogonal eigenvectors of A; Σ is a singular value matrix whose elements are the eigenvalues ​​λ of matrix A i Where U and V are seismic waveform data and sand body thickness data, which are n×n and m×m orthogonal matrices respectively, and V T is the conjugate transpose of V, which is an n×m-order non-negative real diagonal matrix, representing the correlation between seismic waveform data and sand body thickness. The above formula can be expanded to:

[0111] (6)

[0112] Right now:

[0113] (7)

[0114] For seismic profiles, the first component of the decomposition is Represents the seismic waveform with the most commonality with sand body thickness, the second component Represents removing the first component The seismic waveform with the most commonality is shown below. Similarly, different components represent different sedimentary environments and geological characteristics, and the same component represents the seismic waveforms corresponding to different lithologic combinations.

[0115] After the singular value decomposition of matrix A is performed using formula (6), the main features of matrix A can be represented by the singular vectors corresponding to the first three non-zero singular values, that is, dynamic clustering analysis of seismic waveforms is realized, and the corresponding relationship between the seismic response wave components of the wellbore channel and different sedimentary facies is obtained.

[0116] By expanding the correspondence between the seismic response wave components of the well bypass and different sedimentary facies, we can obtain the correspondence between the seismic response wave components of the interwell area and different sedimentary facies. Thus, the correspondence between the seismic response wave components of each location in the target area and different sedimentary facies is obtained.

[0117] Specifically, by comparing the seismic response wave components of the well bypass with the lithologic, structural, and tectonic characteristics of known sedimentary facies, a template for the correspondence between seismic response waves and sedimentary facies can be established. This template is then applied to the interwell area. By analyzing the seismic data, the variation pattern of the seismic response wave components is identified and matched with the sedimentary facies characteristics, thereby obtaining the correspondence between the seismic response wave components in the interwell area and different sedimentary facies. Ultimately, through a comprehensive analysis of the seismic response wave components at the well locations and between the wells in the target area, we are able to construct a comprehensive correspondence network, achieving an accurate grasp of the correspondence between the seismic response wave components at each location in the target area and different sedimentary facies. This not only enhances our understanding of the underground geological structure, but also provides valuable geological basis for oil and gas exploration and development.

[0118] (2) Using the correspondence between the seismic response wave components at each location in the target area and different sedimentary facies as prior information, a likelihood function is constructed to describe the probability of the planar distribution of the sedimentary facies.

[0119] An initial model of seismic wave impedance is established based on the correspondence between the seismic response wave components at each location in the target area and different sedimentary facies, and it is used as prior information to complete the construction of the likelihood function.

[0120] The likelihood function is calculated using matched filtering between the initial impedance model and the seismic waveform.

[0121] (8)

[0122] Among them, L (θ|x) is the likelihood function, θ is the seismic waveform, and x is the elastic parameter sandstone model to be solved. For a convolution model:

[0123] (9)

[0124] Where w is the seismic wavelet and n is the noise. Assume that the noise n is Gaussian distributed:

[0125] (10)

[0126] Estimating the posterior probability distribution of the elastic parameter model x from known seismic data θ can be viewed as a Bayesian inversion problem. I is the prior information. Substituting (9) into (10), the likelihood function of the seismic data is:

[0127] (11)

[0128] Indicates the i Earthquake waves, Indicates the i The disturbance amount of the elastic parameters corresponding to the seismic wave, the elastic parameters include thickness, I is the prior information, w represents the seismic wavelet, and N represents the number of seismic waves.

[0129] Based on the likelihood function and the seismic response waves in the target area, the planar distribution probability of different sedimentary facies can be determined.

[0130] (3) Based on the planar distribution probability of different sedimentary phases, the seismic response waves corresponding to different sediments are grouped.

[0131] Each type of sedimentary facies has its own thickness range. Based on this thickness, we determine the sedimentary facies type for each region on the plane. The planar (lateral) distribution probability of different sedimentary facies is also the distribution probability of different sand body thickness ranges. Based on this distribution probability, the seismic waves corresponding to different sedimentary facies are grouped.

[0132] For example, a first sedimentary facies (tidal channel), a second sedimentary facies (sand flat), and a third sedimentary facies (swamp-mud flat) are distributed laterally. These first, second, and third sedimentary facies are divided according to their respective thickness intervals. Seismic response waves collected from the planar region corresponding to the first sedimentary facies constitute a first sample set, seismic response waves collected from the planar region corresponding to the second sedimentary facies constitute a second sample set, and seismic response waves collected from the planar region corresponding to the third sedimentary facies constitute a third sample set. Furthermore, the first, second, and third sample sets may each include well logging data for their corresponding regions.

[0133] (4) Perform seismic wave inversion based on each sample set.

[0134] In this example, during the inversion process, the lithologic sensitivity curve was selected and a histogram was established using the well logging data to select the sensitivity curve. A specific example is provided to illustrate this. The statistical analysis of the sandstone thickness of the Benxi Formation in the target area (a study area in Jiaxian District mentioned above) shows that the thickness of a single sand body in the Benxi Formation ranges from 2m to 15m, and 89% of the sand bodies range from 2m to 8m. Figure 5 To establish various lithologic interpretation diagrams of natural gamma ray curves, see Figure 6 . from Figure 5 and Figure 6 As can be seen from the figure, the natural gamma ray curve can well identify sandstone, mudstone and coal seams. Therefore, the natural gamma ray curve is added as a priori information in the inversion to better predict sandstone.

[0135] This embodiment adopts the seismic wave inversion technology based on phase-controlled waveform. After obtaining the above sample set, Figure 7 As shown in the figure, based on Bayesian theory, the posterior probability distribution is obtained using the sand body probability distribution and likelihood function.

[0136] First, through single-well sedimentary facies analysis, combined with logging curves and coring data, the logging characteristics corresponding to different sedimentary deposits are determined. Second, the lateral changes in the seismic waveform are used to effectively identify and accurately classify the seismic waveform shape, thereby clarifying the geological significance of the seismic facies (see Figure 2).

[0137] In Bayesian inversion, assuming that the elastic parameter model x to be solved is also Gaussian distributed, the prior distribution of the model can be calculated as:

[0138] (12)

[0139] The product of the data conditional probability distribution and the model prior probability distribution is used as the posterior probability distribution function of the model. Its specific expression is:

[0140] (13)

[0141] Where Δx is the disturbance of the model parameters, is the variance of the model parameter perturbation. Finally, the parameters are adjusted and the model perturbation is iterated continuously to obtain the waveform indication inversion results under the phase control condition, and finally the sand body distribution under the influence of the coal seam is predicted.

[0142] In addition, key parameters need to be selected and optimized during phase-controlled waveform inversion, including the optimal number of samples, optimal cutoff frequency, and smoothing radius. Taking the optimal cutoff frequency as an example, when the seismic frequency band remains unchanged, supplementing with reliable high-frequency information can improve seismic resolution. The more high-frequency information is supplemented, the greater the randomness of the seismic inversion results. Therefore, we need to statistically set the optimal cutoff frequency for the waveform inversion sample wells. By statistically analyzing the correlation coefficient of the longitudinal wave impedance frequency of the Benxi Formation in 46 wells in the target area (Jiaxian District), see Figure 8 When the frequency is 170 Hz, an inflection point appears. When the frequency is greater than the inflection point, the correlation coefficient tends to be stable. When the frequency is greater than 200 Hz, the correlation index decreases slowly, indicating that the inversion randomness becomes greater. Therefore, 200 Hz can be selected as the high cutoff frequency for inversion.

[0143] Inversion effect analysis is a key indicator for evaluating the feasibility and rationality of inversion methods, directly influencing their promotion. Using wells in the study area that were not included in the inversion as test wells, the consistency between their inversion effects and the sandstones actually drilled was calculated. This verification demonstrates that the coal-bearing stratum thin sandbody prediction method based on seismic frequency division and phase-controlled waveform indication inversion can effectively reduce the impact of coal seams on sandbody prediction. While remaining faithful to well point data, it also offers good predictability for interwell sandbodies.

[0144] Phase-controlled seismic waveform indication inversion better reflects the characteristics of rapid lateral phase change of tidal channel sand bodies on the plane, and the description of sand body boundaries is clearer, such as Figure 9 and Figure 10 As shown in the inversion plan, the tidal channel is distributed in a northeast-southwest direction, forming a strip. The inversion effect is clear and reliable.

[0145] This embodiment of the present invention uses a method for predicting thin sand bodies in coal-bearing strata based on seismic frequency division and phase-controlled waveform inversion to identify and characterize the distribution of sandstones within the Benxi Formation coal-bearing strata. Spectral decomposition of seismic data improves vertical resolution, while analysis of sedimentary and seismic facies optimizes key inversion parameters to enhance lateral resolution.

[0146] The following describes the sand distribution prediction device provided by the present invention. The sand distribution prediction device described below and the sand distribution prediction method described above can be referenced to each other. The sand distribution prediction device includes the following modules:

[0147] The data acquisition module 1101 is used to acquire seismic response waves and logging data beside the wellbore in the target area.

[0148] The singular value decomposition module 1102 is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies based on the seismic response wave beside the wellbore and the well logging data using a singular value decomposition algorithm.

[0149] The likelihood function construction module 1103 is used to construct a likelihood function for describing the probability of planar distribution of sedimentary facies using the correspondence between the seismic response wave components in the target area and different sedimentary facies as prior information.

[0150] The distribution probability determination module 1104 is used to determine the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response wave of the target area.

[0151] The sample grouping module 1105 is used to group the seismic response waves corresponding to different sediments based on the planar distribution probabilities of different sedimentary phases to obtain multiple sample sets.

[0152] The sand body distribution prediction module 1106 is used to perform seismic wave inversion based on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area.

[0153] The specific working of each module in the device can be referenced with the sand body distribution prediction method described above, and will not be described in detail here.

[0154] Figure 12 An example of a physical structure diagram of an electronic device is shown below. Figure 12 As shown, the electronic device may include: a processor 810 , a communication interface 820 , a memory 830 and a communication bus 840 , wherein the processor 810 , the communication interface 820 and the memory 830 communicate with each other via the communication bus 840 . Processor 810 can call logic instructions in memory 830 to execute a sand body distribution prediction method, which includes: obtaining seismic response waves and well logging data near a target area; determining the correspondence between seismic response wave components and different sedimentary facies in the target area using a singular value decomposition algorithm based on the seismic response waves and the well logging data; constructing a likelihood function for describing the planar distribution probability of the sedimentary facies using the correspondence between the seismic response wave components and the different sedimentary facies in the target area as prior information; determining the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response waves in the target area; grouping the seismic response waves corresponding to different sedimentary facies based on the planar distribution probability of different sedimentary facies to obtain multiple sample sets; and performing seismic wave inversion on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area. For a more specific method, please refer to the sand body distribution prediction method described above and will not be repeated here.

[0155] Furthermore, the logic instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0156] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the sand body distribution prediction method provided by the above-mentioned methods, the method comprising: obtaining seismic response waves and logging data beside the wellbore in the target area; based on the seismic response waves beside the wellbore and the logging data, using a singular value decomposition algorithm to determine the correspondence between the seismic response wave components in the target area and different sedimentary phases; using the correspondence between the seismic response wave components in the target area and different sedimentary phases as prior information, constructing a likelihood function for describing the planar distribution probability of the sedimentary phase; based on the likelihood function and the seismic response waves in the target area, determining the planar distribution probability of different sedimentary phases; based on the planar distribution probability of different sedimentary phases, grouping the seismic response waves corresponding to different sediments to obtain multiple sample sets; performing seismic wave inversion based on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area. As for a more specific method, reference can be made to the sand body distribution prediction method described above, which will not be repeated here.

[0157] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the sand body distribution prediction method provided by the above-mentioned methods, the method comprising: obtaining seismic response waves and logging data beside the wellbore in the target area; based on the seismic response waves and logging data beside the wellbore, using a singular value decomposition algorithm to determine the correspondence between the seismic response wave components in the target area and different sedimentary phases; using the correspondence between the seismic response wave components in the target area and different sedimentary phases as prior information, constructing a likelihood function for describing the planar distribution probability of the sedimentary phase; based on the likelihood function and the seismic response waves in the target area, determining the planar distribution probability of different sedimentary phases; based on the planar distribution probability of different sedimentary phases, grouping the seismic response waves corresponding to different sediments to obtain multiple sample sets; performing seismic wave inversion on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area. As for a more specific method, reference can be made to the sand body distribution prediction method described above, which will not be repeated here.

[0158] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0159] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0160] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A sand body distribution prediction method, characterized in that: include: Obtain seismic response waves and logging data beside the wellbore in the target area; Based on the wellbore side seismic response waves and the well logging data, a singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies; Using the correspondence between the seismic response wave components and different sedimentary facies in the target area as prior information, a likelihood function is constructed to describe the probability of sedimentary facies plane distribution. determining the planar distribution probability of different sedimentary facies based on the likelihood function and the seismic response wave of the target area; Based on the planar distribution probability of different sedimentary phases, the seismic response waves corresponding to different sediments are grouped to obtain multiple sample sets; Seismic wave inversion is performed based on each sample set to obtain the vertical thickness and planar distribution of the sand body in the target area.

2. The sand body distribution prediction method according to claim 1, wherein: Also includes: Obtaining original seismic waveform data of the target area; Based on a matching pursuit algorithm, the coal seam reflection wave is filtered out from the original seismic waveform data to obtain the seismic response wave.

3. The sand body distribution prediction method according to claim 2, wherein: Based on a matching pursuit algorithm, the coal seam reflection wave is filtered out from the original seismic waveform data, including: Performing seismic forward modeling on the target area to obtain simulated reflection waves of the coal seams in the target area; Based on a matching pursuit algorithm, the simulated reflection wave is used as a matching wavelet to perform spectral decomposition on the original seismic waveform data to obtain seismic wavelets at different frequencies; the seismic wavelets include a target seismic wavelet that best matches the simulated reflection wave; Based on the seismic wavelets other than the target seismic wavelet, the seismic wave is reconstructed to obtain the seismic response wave.

4. The sand body distribution prediction method according to claim 1, wherein: The well logging data includes: well logging sand body thickness data; Based on the wellbore side seismic response waves and the well logging data, a singular value decomposition algorithm is used to determine the correspondence between the seismic response wave components of the target area and different sedimentary facies, including: Based on the wellbore side seismic response wave and the well logging sand body thickness data, a singular value decomposition algorithm is used to determine the correspondence between the wellbore side seismic response wave components and different sedimentary facies; The corresponding relationship between the seismic response wave components and different sedimentary facies in the inter-well area is determined based on the corresponding relationship between the seismic response wave components beside the wellbore and different sedimentary facies.

5. The sand body distribution prediction method according to claim 1, wherein: The likelihood function is as follows: ； Indicates the i Earthquake waves, Indicates the i The disturbance amount of the elastic parameters corresponding to the seismic wave, the elastic parameters include thickness, I is the prior information, w represents the seismic wavelet, and N represents the number of seismic waves.

6. The sand body distribution prediction method according to claim 1, characterized in that: The seismic wave inversion is based on phase-controlled waveform.

7. The sand body distribution prediction method according to claim 6, characterized in that: The seismic wave inversion comprises: Constructing a posterior probability distribution function based on the likelihood function and the prior probability distribution function; Seismic wave inversion is performed based on the posterior probability distribution.

8. The sand body distribution prediction method according to claim 1, characterized in that: The seismic wave inversion comprises: Selecting a lithology sensitivity curve based on a correlation analysis between the well logging data and lithology; The lithologic sensitivity curve is used as prior information to invert the seismic wave.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the sand body distribution prediction method according to any one of claims 1 to 8 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the sand body distribution prediction method according to any one of claims 1 to 8 is implemented.

Images