Fluvial facies thin sand body prediction method based on seismic bipolar transformation

Abstract

The invention relates to the technical field of petroleum geophysical prospecting, and provides a fluvial facies thin sand body prediction method based on seismic bipolar transformation. Comprising the steps of preprocessing a three-dimensional seismic data body; performing seismic phase transformation on the preprocessed three-dimensional seismic data volume; estimating self-adaptive segmented time-varying wavelets of the three-dimensional seismic data volume after phase transformation; calculating volume class regular reflection coefficients of seismic data of different phases; carrying out convolution on the quasi-regular reflection coefficient and the time-varying wavelet to obtain a single-phase seismic data volume; reconstructing different phase convolution results to generate wave crests and wave troughs, and obtaining a bipolar seismic data volume; tracking and describing a thin-layer sand body by using an earthquake bipolar data body; according to the method, the corresponding relation between the sand body and the seismic response is improved, and the correlation between the synthetic seismic record of the target interval and the seismic trace beside the well reaches 0.9 or above.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of petroleum exploration technology, and provides a river facies thin sand body prediction method based on seismic bipolar transformation. BACKGROUND

[0002] The river facies has the following characteristics:

[0003] First, the rock has a clear upward-fining unit structure, and the bottom surface of the unit sand body often has a bottom scouring phenomenon;

[0004] Second, the sand body is obviously lenticular in output, and the lithofacies changes rapidly;

[0005] Third, the mineral maturity is relatively low, and the gravel components mainly include various foreign gravels;

[0006] Fourth, it has a unidirectional flow water directional structure, such as cross-bedding, and gravels are arranged in a shingle shape;

[0007] Fifth, cross-bedding is particularly developed, especially the “epsilon” type layering formed by the high-curvature river meander bar accretion and the “pi” type layering developed by the wandering river channel.

[0008] Thin sand bodies are easily formed in river facies, which refer to the sand bodies with a thickness of less than 8 m and a wide lateral distribution range, which are separated by stable mudstone layers in the interbedded sandstone and mudstone profile developed in a sedimentary basin, and the sandstone ratio is generally between 0.1 and 0.6. The lithology mainly includes siltstone, fine siltstone, fine sandstone and medium sandstone.

[0009] Among them, the thin layer problem has always been a hot research field in geophysics, and there are generally four aspects in summarizing its theory, technology and application cases:

[0010] First, the tuning effect of the thin layer is used to realize the identification and thickness calculation of the thin layer by using the tuning amplitude (MARFURT, Chen Jisong, etc.);

[0011] Second, the frequency spectrum is decomposed based on the time-frequency characteristics such as short-time Fourier transform and wavelet transform to realize qualitative and quantitative description;

[0012] Third, the high-frequency information of the drilling data is used to carry out prestack and poststack impedance inversion, geostatistics and waveform indication inversion to improve the vertical identification of the thin layer for prediction;

[0013] Fourth, the lateral resolution of the seismic data is used to realize the boundary description of the thin layer through seismic sedimentology, stratal slicing and multi-attribute description.

[0014] Seismic reflection is the comprehensive superposition of multiple reflection interfaces in space, and the diversity of sand body combination patterns will further complicate the seismic reflection characteristics, so the prediction effect is often poor.

[0015] The upper section of Guantao Formation in Chengbei 26 wellblock has a burial depth of 1200-2000m. The lithology is mainly gray fine sandstone, siltstone and interbedded brown and gray mudstone. The main oil-bearing layer is the upper section of Guantao Formation in Neogene System, which has many thin layers vertically, large thickness variation horizontally and belongs to a set of fluvial facies sandstone reservoir.

[0016] The fluvial facies reservoir generally presents high amplitude, discontinuous, parallel or sub-parallel dense reflection structure on seismic profile. The meandering river sand body in the upper section of Guantao Formation in this area has frequent channel rechanneling and fast lateral migration, which generally shows short axis, medium-strong amplitude anomaly on seismic profile. The distribution and variation of the same phase axis reflect the geological accumulation process of fluvial deposition and the difference and spatial variation of channel sand body deposition, which is the basis for describing the single channel sand body.

[0017] The sand body thickness in the upper section of Guantao Formation in Chengbei 26 wellblock is 5-8m, the single sand body thickness is thin, the frequency band of seismic data in the target layer is narrow (8-42Hz), the main frequency is low (37Hz), the ability of seismic data to distinguish thin sand body is poor, the prediction accuracy of reservoir by seismic amplitude attribute is about 60%, which cannot meet the requirements of oilfield development plan adjustment. At present, Chengbei 26 wellblock has entered the development adjustment period, and multiple rounds of research have been carried out for reservoir prediction. The prediction accuracy of sand body is high when the reservoir thickness is large and the seismic response is good. When the reservoir thickness is small and the seismic response is poor, the seismic attribute cannot reflect the change of sand body, and the appropriate frequency range of logging information in the process of logging constrained inversion is difficult to determine, and the inversion effect of thin sand layer is also poor.

[0018] The patent application No. CN201610891188.2 discloses a method for predicting fluvial facies low sand rate reservoir, and discloses the following technical features:

[0019] Step 1, fine calibration of synthetic seismic record is carried out to determine the depth range of fluvial facies low sand rate reservoir; Step 2, multiple attributes are extracted to jointly analyze the macroscopic distribution characteristics of channel body; Step 3, rock and electricity analysis is carried out to optimize the sensitive curve of reservoir indicator; Step 4, inversion is carried out by using the sensitive curve to predict the vertical distribution of reservoir.

[0020] The above patent can finely determine the distribution of fluvial facies low sand rate reservoir and reveal potential target area, so as to better guide oil and gas exploration and provide guidance and reference for prediction of other sand body reservoirs.

[0021] However, the above patent cannot accurately predict thin sand layer in fluvial facies sand body, and the description is not fine enough.

[0022] Therefore, we have invented a method for predicting fluvial facies thin sand body based on seismic bipolar transformation, which solves the above technical problems. SUMMARY

[0023] The present application aims at solving the defects of the prior art and proposes a river facies thin sand body prediction method based on seismic bipolar transformation.

[0024] The river facies thin sand body prediction system based on seismic bipolar transformation and biological heuristic calculation and seismic data phase transformation theory improves the seismic data apparent resolution, makes the single sand body have corresponding seismic events, improves the matching relationship between the thin sand layer and the seismic response, and improves the thin reservoir prediction accuracy.

[0025] In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme:

[0026] A river facies thin sand body prediction method based on seismic bipolar transformation comprises the following steps:

[0027] S1, three-dimensional seismic data body pretreatment;

[0028] S2, seismic phase transformation on the pretreated three-dimensional seismic data body;

[0029] S3, estimation of adaptive segmented time-varying wavelets of the three-dimensional seismic data body after phase transformation;

[0030] S4, calculation of different phase seismic data body class positive reflection coefficients;

[0031] S5, single phase seismic data body obtained by class positive reflection coefficient and time-varying wavelet convolution;

[0032] S6, reconstruction of the peak and the valley of the different phase convolution results to obtain the bipolar seismic data body;

[0033] S7, thin layer sand body description by using the seismic bipolar data body tracking.

[0034] Further, the pretreatment of step S1 improves the seismic resolution pretreatment by spectrum decomposition high frequency gain signal reconstruction, comprising:

[0035] S11, seismic data spectrum decomposition is carried out, and gain calculation is carried out on the signal with low high frequency part energy;

[0036] S12, reconstruction is carried out on the low, medium and high frequency signals to achieve the purpose of widening the frequency band.

[0037] Further, the seismic phase transformation of step S2 comprises 90°, 180° and 270° seismic phase transformation by Fourier transform.

[0038] Further, the time-varying wavelet estimation in step S3 comprises:

[0039] S31, data preparation: select three-bit seismic data volume after phase transformation;

[0040] S32, time-frequency analysis: through short-time Fourier transform or wavelet transform, the frequency characteristics of the seismic signal changing with time are obtained;

[0041] S33, wavelet extraction: according to the results of time-frequency analysis, the main frequency of each time point is identified, and the corresponding time-varying wavelet is extracted by using the main frequency information combined with the time domain waveform of the seismic signal.

[0042] Further, the reflection coefficient calculation in step S4 includes:

[0043]

[0044] In the formula:

[0045] R-reflection coefficient; ρ1v1-wave impedance of medium 1, g / s.cm 2 10 4 ; ρ2v2-wave impedance of medium 2, g / s.cm 2 10 4 .

[0046] Further, the three-dimensional seismic data volume is a standard segy format file.

[0047] Further, the adaptive segmented time-varying wavelet estimation includes 0°, 90°, 180°, 270° adaptive segmented time-varying wavelet estimation.

[0048] Further, the different phase seismic data class positive reflection coefficient calculation includes the calculation of 0°, 90°, 180°, 270° phase seismic data class positive reflection coefficient.

[0049] Further, the seismic data volume class positive reflection coefficient adopts a biological heuristic calculation class positive reflection coefficient processing technology to optimize two or more algorithms for fusion calculation according to different geological conditions.

[0050] Further, a river facies thin sand body prediction method based on seismic bipolar transformation includes:

[0051] Input the three-dimensional seismic data volume of the quasi-segy format file;

[0052] The original three-dimensional seismic data is preprocessed by spectral decomposition high-frequency gain signal reconstruction to improve the resolution of the seismic data, and the thin layer sand body resolution capability of the seismic data is preliminarily improved;

[0053] The seismic data after resolution improvement preprocessing is subjected to 90°, 180°, 270° seismic phase transformation;

[0054] 0°, 90°, 180°, 270° adaptive segmented time-varying wavelet estimation;

[0055] 0°, 90°, 180°, 270° different phase seismic data type reflection coefficient calculation;

[0056] Reflection coefficient type and time-varying wavelet convolution get single phase high resolution seismic;

[0057] Different phase convolution results are reconstructed according to a specific combination to generate wave crest, wave trough, and get bipolar view high resolution seismic data;

[0058] The thin sand body is tracked and described by using the seismic bipolar data body.

[0059] The present application has the following beneficial effects:

[0060] The corresponding relationship between the sand body and the seismic response is improved, the correlation between the synthetic seismogram of the target layer and the well seismic trace reaches more than 0.9; according to the sand layer division contrast, reservoir prediction and each sand layer thickness data obtained by comprehensive interpretation of well logging data, the sandstone thickness contour map of each sand layer is prepared under the guidance of the fluvial facies model, so that the sand body distribution characteristics of each sand layer can be known. BRIEF DESCRIPTION OF DRAWINGS

[0061] Figure 1 It is a seismic bipolar transformation thin sand layer and seismic response corresponding relationship change schematic diagram, which reflects the corresponding relationship between the different phase seismic bipolar transformation thin sand layer and the seismic response, and each set of thin layer sand body has a seismic wave crest or wave trough in-phase axis corresponding relationship, thereby improving the corresponding relationship between the thin layer sand body and the seismic response;

[0062] Figure 2 It is a seismic bipolar transformation technology flow chart based on biological heuristic calculation, which clearly reflects the technology and process in the high resolution seismic data processing process of the system;

[0063] Figure 3 It is a spectrum decomposition high frequency gain signal reconstruction resolution improvement preprocessing process and spectrum change diagram, through processing the high frequency energy in the earthquake, the high frequency energy is supplemented, the seismic attenuation is reduced, and the purpose of preliminary improving the resolution of the seismic data is achieved;

[0064] Figure 4 It is a spectrum decomposition high frequency gain signal reconstruction frequency expansion processing before and after comparison chart of Chengbei 26 well area, after processing, weak reflection appears between thin layers, and the ability of the seismic data to distinguish thin sand layers is effectively improved;

[0065] Figure 5The figure is a bio-inspired computing class positive reflection coefficient processing flowchart of the present application.

[0066] Figure 6 The figure is a bio-inspired computing class positive reflection coefficient comparison chart of the actual seismic data of Chengbei 26 well block of the present application, the class positive reflection coefficient waveform is narrowed, the pinch-off point is clear, the lateral variation and the original seismic have consistency.

[0067] Figure 7 The figure is a seismic bipolar transformation high resolution processing flowchart of the actual seismic data of Chengbei 26 well block of the present application, which reflects the change of different phase seismic data in the process of seismic bipolar transformation of the actual seismic data and the final reconstruction result.

[0068] Figure 8 The figure is a comparison chart of the original seismic and the bipolar transformation seismic of the present application, it can be seen that the seismic events are obviously increased, the waveform is narrowed, and the seismic apparent resolution is improved by 1 times.

[0069] Figure 9 The figure is a seismic bipolar transformation processing seismic maximum amplitude chart of the upper segment 121 sand group of the Chengbei 26 well block of the present application.

[0070] Figure 10 The figure is a sand body thickness distribution chart of the upper segment 121 sand group of the Chengbei 26 well block of the present application, the sand body prediction result has strong regularity, the sand body distribution is consistent with the regional sedimentation understanding, the predicted sand body thickness and the actual drilled well are highly consistent, the interwell sand body change is reasonable, and the requirements of development plan adjustment can be met. DETAILED DESCRIPTION

[0071] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments.

[0072] Embodiment one:

[0073] A river facies thin sand body prediction method based on seismic bipolar transformation, comprising:

[0074] Three-dimensional seismic data body pretreatment, the pretreatment improves the seismic resolution by spectral decomposition high frequency gain signal reconstruction; the three-dimensional seismic data body is a standard segy format file; the spectral decomposition high frequency gain signal reconstruction comprises: first, seismic data spectrum decomposition is carried out, the signal with low energy in high frequency part is calculated according to the effective algorithm, and finally the low, medium and high frequency signals are reconstructed to achieve the purpose of widening the frequency band.

[0075] The preprocessed three-dimensional seismic data volume is subjected to seismic phase transformation; the seismic phase transformation includes 90o, 180o, 270o seismic phase transformation.

[0076] Adaptive segmented time-varying wavelet estimation of the phase-transformed three-dimensional seismic data volume; the adaptive segmented time-varying wavelet estimation includes 0o, 90o, 180o, 270o adaptive segmented time-varying wavelet estimation.

[0077] Different phase seismic data volume pseudo-reflectivity coefficient calculation; the different phase seismic data volume pseudo-reflectivity coefficient calculation includes 0o, 90o, 180o, 270o phase seismic data volume pseudo-reflectivity coefficient calculation; the seismic data volume pseudo-reflectivity coefficient uses a bio-inspired pseudo-reflectivity coefficient calculation processing technology to optimally fuse two or more algorithms according to different geological conditions.

[0078] Pseudo-reflectivity coefficient and time-varying wavelet convolution to obtain single-phase seismic data volume;

[0079] Different phase convolution results are reconstructed to generate wave peaks and wave troughs to obtain bipolar seismic data volume;

[0080] The seismic bipolar data volume is used to track and describe thin sand bodies.

[0081] Embodiment two:

[0082] A river facies thin sand body prediction method based on seismic bipolar transformation, comprising:

[0083] Inputting a three-dimensional seismic data volume in quasi-segy format;

[0084] The original three-dimensional seismic data is preprocessed by spectral decomposition high-frequency gain signal reconstruction to improve seismic data resolution and initially improve the thin sand body resolution capability of the seismic data;

[0085] The resolution-improved preprocessed seismic data is subjected to 90o, 180o, 270o seismic phase transformation;

[0086] 0o, 90o, 180o, 270o adaptive segmented time-varying wavelet estimation;

[0087] 0o, 90o, 180o, 270o different phase seismic data pseudo-reflectivity coefficient calculation;

[0088] Pseudo-reflectivity coefficient and time-varying wavelet convolution to obtain single-phase high-resolution seismic;

[0089] Different phase convolution results are reconstructed to generate wave peaks and wave troughs according to a specific combination method to obtain bipolar apparent high-resolution seismic data;

[0090] The seismic bipolar data volume is used to track and describe thin sand bodies

[0091] Embodiment three:

[0092] A seismic bipolar transform Chengdao oilfield fluvial thin sand body prediction method, comprising:

[0093] In step 101, input the three-dimensional seismic data body, that is, the standard segy format file. The flow enters step 102.

[0094] In step 102, for Chengbei 26 well three-dimensional seismic data body, the spectral decomposition high frequency gain signal reconstruction preprocessing technology preliminarily improves the seismic resolution flow enters step 103.

[0095] In step 103, using Chengbei 26 well block to improve the resolution of seismic data, 90°, 180°, 270° seismic phase transform, the flow enters step 104.

[0096] In step 104, 0°, 90°, 180°, 270° adaptive segmented time-varying wavelet estimation is carried out, and the flow enters step 105.

[0097] In step 105, 0°, 90°, 180°, 270° different phase seismic data biological heuristic algorithm type positive reflection coefficient calculation is carried out, and the flow enters step 106.

[0098] In step 106, the type positive reflection coefficient is multiplied with the time-varying wavelet to obtain single phase high resolution seismic, and the flow enters step 107.

[0099] In step 107, different phase convolution results are reconstructed to generate wave peak and wave trough according to specific combination mode, and bipolar apparent high resolution seismic data is obtained, and the flow enters step 108.

[0100] In step 108, the thin sand layer is predicted and the sand body distribution rule is analyzed by using the seismic bipolar data body to track and describe the thin sand body.

[0101] The above is only the preferred specific embodiment of the present application, but the protection scope of the present application is not limited to this, any skilled person in the art can replace or change the technical solution and the inventive concept of the present application according to the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A method for predicting fluvial thin sand bodies based on seismic bipolar transformation, characterized in that, Includes the following steps: S1. Preprocessing of 3D seismic data volume; S2. Perform seismic phase transformation on the preprocessed 3D seismic data volume; S3. Estimating the adaptive piecewise time-varying wavelet of the three-dimensional seismic data volume after phase transformation; S4. Calculate the orthogonal reflection coefficients of seismic data volumes with different phases; S5. The convolution of the pseudo-positive reflection coefficient and the time-varying wavelet yields a single-phase seismic data volume; S6. Reconstruct the wave peaks and troughs from the convolution results of different phases to obtain a bipolar seismic data volume; S7. Use seismic bipolar data volumes to trace and describe thin sand bodies.

2. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 1, characterized in that, Step S1, the preprocessing described, improves seismic resolution through spectral decomposition and high-frequency gain signal reconstruction, including: S11. Perform spectral decomposition of seismic data and calculate the gain of signals with lower energy in the high-frequency part; S12. Reconstruct low, medium and high frequency signals to achieve the purpose of widening the frequency band.

3. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 1, characterized in that, The seismic phase transformation in step S2 includes 90°, 180°, and 270° seismic phase transformations through Fourier transform.

4. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 1, characterized in that, The time-varying wavelet estimation in step S3 includes: S31. Data preparation: Select the three-dimensional seismic data volume after phase transformation; S32. Time-frequency analysis: Obtain the frequency characteristics of seismic signals as a function of time through short-time Fourier transform or wavelet transform; S33. Wavelet Extraction: Based on the results of time-frequency analysis, identify the dominant frequency at each time point, and use the dominant frequency information in combination with the time-domain waveform of the seismic signal to extract the corresponding time-varying wavelet.

5. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 1, characterized in that, Step S4, the calculation of the reflection coefficient, includes: In the formula: R – Reflection coefficient; ρ1v1 – Wave impedance of medium 1, g / s / cm 2 10 4 ρ2v2 – Wave impedance of medium 2, g / s·cm 2 10 4 .

6. A method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to any one of claims 2-5, characterized in that, The three-dimensional seismic data volume is a standard segy format file.

7. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 6, characterized in that, The adaptive piecewise time-varying wavelet estimation includes 0°, 90°, 180°, and 270° adaptive piecewise time-varying wavelet estimation.

8. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 7, characterized in that, The calculation of orthoreflection coefficients for different phase seismic data types includes the calculation of orthoreflection coefficients for 0°, 90°, 180°, and 270° phase seismic data types.

9. The method for predicting fluvial thin sand bodies based on seismic bipolar transformation according to claim 1, characterized in that, The positive reflection coefficient of the earthquake data volume is calculated using a bio-inspired computational positive reflection coefficient processing technique, which selects two or more algorithms to be fused together based on different geological conditions.

10. A method for predicting fluvial thin sand bodies based on seismic bipolar transformation, characterized in that, include: Input a 3D seismic data volume in quasi-Segy format; The original 3D seismic data was preprocessed by reconstructing high-frequency gain signals through spectral decomposition to improve the resolution of seismic data and initially enhance the resolution of thin sand bodies in the seismic data. For the preprocessed seismic data with improved resolution, perform 90°, 180°, and 270° seismic phase transformations; Adaptive piecewise time-varying wavelet estimation at 0°, 90°, 180°, and 270°; Calculation of orthogonal reflection coefficients for seismic data with different phases at 0°, 90°, 180°, and 270°; High-resolution seismic data with a single phase is obtained by convolving the quasi-positive reflection coefficient with a time-varying wavelet. Different phase convolution results are reconstructed into wave peaks and troughs according to a specific combination to obtain bipolar apparent high-resolution seismic data; Thin sand bodies were described by using seismic bipolar data volumes.

Images