A method for making a time-depth conversion template for shallow conglomerate

Abstract

The present invention relates to the technical field of fine velocity modeling of the surface layer in oil geological exploration, and particularly relates to a method for making a time-depth conversion template for shallow conglomerate. The making method includes the following steps: (1) Selecting the time and depth data of the target area as research samples, and performing layering, which are, from top to bottom, a low-velocity layer, a velocity-decreasing layer, and a high-velocity layer; (2) Obtaining the thickness information according to the depth data in step (1), performing regional spatial interpolation, and obtaining and calculating the thickness of the low-velocity layer and the vertical delay time of each sample; (3) Performing normalization correction on the time-depth data of the velocity-decreasing layer; (4) Using a high-order polynomial based on the least squares method to fit the time-depth data of the velocity-decreasing layer corrected in step (3), determining the parameters, and the obtained high-order polynomial is the time-depth conversion template of the velocity-decreasing layer. The present invention has the advantage of high simulation accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of precise velocity modeling of the surface layer in oil geological exploration, and specifically relates to a method for making a time-depth conversion template for shallow conglomerate. Background Art

[0002] In the southern margin of the Junggar Basin, thick low- and high-velocity conglomerate layers are generally developed in the shallow surface layer. The complex velocity structure changes lead to poor seismic imaging quality of the shallow surface layer and the underlying strata, and it is difficult to accurately determine the structural form and high points of the underlying target layer. It is difficult to accurately obtain the spatial distribution and velocity variation relationship of the low- and high-velocity conglomerate layers, and only pattern-based interpretation and velocity definition can be used for processing. Therefore, the overall imaging accuracy is uncontrollable, resulting in large displacements of the high points of the target layer structure and large errors in the well-seismic time-depth relationship.

[0003] For velocity modeling of shallow low- and high-velocity conglomerates, the commonly used method is stripping and constant-velocity filling, that is, after low-velocity stripping, correction is performed using delay time, and high-velocity is filled using constant velocity. Both of these methods will result in different velocity errors, thus affecting the accuracy of seismic migration imaging and not meeting the requirements of high-precision seismic imaging and deep geological target determination at the present stage.

[0004] At the present stage, there are mainly two methods for modeling shallow low-velocity conglomerates in seismic imaging. The first is surface layer tomography inversion in the static correction calculation link. Affected by the accuracy of the first arrival time and the limitations of the method, the model accuracy and vertical depth are limited to a certain extent. The second is constant-velocity modeling in the migration imaging or time-domain to depth conversion process. This method cannot accurately reflect the longitudinal and transverse changes of the velocity and thickness of the conglomerate layer, which can directly lead to the accuracy of its final results.

[0005] Static correction means that in the process of onshore seismic data processing, the seismic data usually needs to be corrected to a unified reference plane, and this reference plane is generally a horizontal plane. The theories of seismic exploration interpretation assume that the excitation point and the receiving point are on a horizontal plane, and the formation velocity is uniform. However, in reality, the ground is often uneven, the depths of each excitation point may also be different, and the wave velocity in the low-velocity zone is very different from the wave velocity in the formation. Therefore, it will surely affect the shape of the measured travel-time curve. To eliminate these influences, the original seismic data needs to be subjected to terrain correction, excitation depth correction, low-velocity zone correction, etc. These corrections are the same for different seismic interfaces at the same observation point, but the model accuracy and vertical depth are limited to a certain extent.

[0006] At present, there are also methods for making dune curves or loess curves for thick desert areas and thick loess areas, but such methods are only applicable to surface homogeneous media (such as deserts and loess), and it is difficult to adapt to the thick gravel areas (inhomogeneous media, where the vertical velocity changes significantly with the media) in the southern margin of the front mountain.

[0007] Therefore, it is very necessary to develop a method for making the time-depth conversion template of shallow thick conglomerate that can solve the above technical problems. Summary of the Invention

[0008] The object of the present invention is to provide a method for making the time-depth conversion template of shallow conglomerate in view of the problems existing in the prior art. This method establishes the vertical velocity variation law of the gravel layer by using the result data such as deep micro-logging, conventional micro-logging, and drill logging, so as to accurately describe the vertical velocity variation characteristics of the gravel layer and the seismic result calibration in the time domain or depth domain.

[0009] The present invention is realized by the following technical solutions:

[0010] A method for making the time-depth conversion template of shallow conglomerate includes the following steps:

[0011] (1) Select the time and depth data of the target area as the research samples, calculate the vertical velocity, and then layer according to the vertical velocity and lithology changes. From top to bottom, they are the low-velocity layer, the velocity-decreasing layer, and the high-velocity layer;

[0012] (2) Obtain the thickness information according to the depth data in step (1), and perform regional spatial interpolation on the velocity and thickness information of the low-velocity layer to obtain and calculate the thickness of the low-velocity layer and the vertical delay time of each sample;

[0013] (3) Use the thickness of the low-velocity layer and the vertical delay time obtained in step (2) to perform normalization correction on the time-depth data of the velocity-decreasing layer;

[0014] (4) Use the high-order polynomial based on the least squares method to fit the time-depth data of the velocity-decreasing layer corrected in step (3), determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the velocity-decreasing layer.

[0015] Preferably, the research samples in step (1) come from at least one of deep micro-logging, conventional micro-logging, and drill logging.

[0016] Preferably, the vertical medium stratification of the research samples in step (1).

[0017] Preferably, in step (1), the low-velocity layer is a floating soil loose gravel layer, the velocity-decreasing layer is a sand-mud-gravel mixed layer, and the high-velocity layer is a cemented gravel layer.

[0018] Preferably, the method of spatial interpolation in step (2) is the inverse distance weighted method based on triangular mesh.

[0019] Preferably, the calculation method of the vertical delay time in step (2) is vertical delay time = thickness of the low-velocity layer / average velocity of the low-velocity layer.

[0020] Preferably, in step (3), the corrected time-depth data of the deceleration layer are superimposed and displayed in the same coordinate system, and the consistency of the time-depth relationship at different positions is analyzed. If they are consistent, it indicates that the data is reasonable; if not, the deceleration layer data needs to be reselected for correction.

[0021] Preferably, the normalization correction in step (3) means that the time and depth data of the deceleration layer are respectively subtracted by the low-speed layer thickness and the vertical delay time obtained in step (2).

[0022] Preferably, the high-order polynomial in step (4) is:

[0023] H(T) = a0 + a1T + a2T 2 + a3T 3 +......+ a k T k k ∈ (4, 5...10);

[0024] Take different values of k, substitute them into the high-order polynomial for fitting, and calculate the error between the measured value and the fitting function. The value with the minimum error is the final k value.

[0025] Preferably, it further includes step (5): Referring to the methods in steps (2)-(4), normalize and correct the time-depth data of the high-speed layer using the low-speed layer and the deceleration layer, fit the corrected time-depth data of the high-speed layer using a high-order polynomial based on the least squares method to determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the high-speed layer.

[0026] More preferably, in step (5), perform regional spatial interpolation on the velocity and thickness information of the low-speed layer and the deceleration layer, obtain and calculate the low-speed layer thickness and the vertical delay time of each sample, then normalize and correct the time-depth data of the high-speed layer, fit the corrected time-depth data of the high-speed layer using a high-order polynomial based on the least squares method to determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the high-speed layer.

[0027] The present invention performs normalization correction to enable its statistical samples to avoid spatial errors caused by vertical changes in the medium.

[0028] When correcting the deceleration layer, the deceleration layer is the target layer and the low-speed layer is the correction layer.

[0029] When correcting the high-speed layer, the high-speed layer is the target layer and the low-speed layer and the deceleration layer are the correction layers.

[0030] Based on the high-order polynomial fitting of the least squares method, the velocity gradient of the deceleration layer gradually decreases with the increase of the burial depth in the vertical direction. In the past, it was difficult to accurately describe the vertical velocity change with linear or second-order polynomial fitting. The high-order polynomial fitting method can better reflect the vertical velocity change law, and the final result has a higher coincidence rate with the sample.

[0031] The beneficial effects of the present invention are as follows:

[0032] The overall technical solution of the present invention mainly includes two aspects. One is the normalization correction at the initial delay of the low-velocity conglomerate layer and the high-velocity conglomerate layer, that is, the vertical delay time of the medium overlying the low-velocity conglomerate or the high-velocity conglomerate is corrected through micro-logging and drill-logging data, so that the initial delay time of the same medium layer is reset to zero. The other is the statistical analysis of the time-depth samples of the same medium layer and the fitting of the time-depth template, that is, the vertical delay time and depth samples in the low-velocity and high-velocity conglomerate layers are respectively analyzed and statistically processed in the specified spatially varying area. According to the vertical velocity change relationship, high-order polynomial template fitting is performed using the least squares method. The obtained template can accurately describe the vertical velocity change of the conglomerate layer, achieving the accurate conversion of the vertical delay time, thickness, and velocity of the conglomerate layer. By using the method of the present invention, when obtaining one parameter variable of the vertical delay time and thickness of the surface conglomerate layer, the other can be accurately calculated, which plays an important role in the calibration and velocity fine modeling of shallow seismic imaging in the time domain or depth domain. Brief Description of the Drawings

[0033] Figure 1 is a flow chart of a method for making a time-depth conversion template for shallow conglomerate provided by the present invention;

[0034] Figure 2 is the depth-time sample analysis of the deep micro-logging, drill-logging original low-velocity layer, deceleration layer, and high-velocity layer in the target area;

[0035] Figure 3 is the depth-time sample of the deceleration layer before normalization correction;

[0036] Figure 4 is the depth-time sample of the deceleration layer after normalization correction and the final velocity curve obtained by fitting. Detailed Embodiments

[0037] The present invention will be further described below in conjunction with specific embodiments, and the advantages and features of the present invention will become clearer with the description. However, these embodiments are only exemplary and do not constitute any limitation to the scope of the present invention. Those skilled in the art should understand that the details and forms of the technical solution of the present invention can be modified or replaced without departing from the spirit and scope of the present invention, but these modifications and replacements all fall within the protection scope of the present invention.

[0038] Embodiment 1

[0039] See Figure 1 , a method for making a time-depth conversion template for shallow conglomerate provided by an embodiment of the present invention includes the following steps:

[0040] Step 101: Select the deep micro-logging and drill logging data of the target area as research samples (time-depth) (see Figure 2 ), conduct a fine analysis of the vertical velocity of each well, and combine the lithology changes for layer identification. The thickness h0 of the low-velocity layer and the average velocity v0 of the low-velocity layer, the thickness h1 of the velocity-decreasing layer, and the average velocity v1 of the velocity-decreasing layer of each well can be obtained. The velocities and thicknesses of different layers are divided according to the slope of the time-depth relationship curve;

[0041] Step 102: Taking the velocity-decreasing layer as the fitting target, search for the three points closest to the position of the measured physical point i, and perform regional spatial interpolation on v0 and h0 respectively using the inverse distance weighted average method based on triangular grids to obtain the thickness h i0 and the average velocity v i0 at the position of each measured physical point i, and then the vertical delay time t i0 =h i0 / v i0 can be calculated;

[0042] Step 103: Screen the samples (see Figure 3 , the sample space of this embodiment is the velocity-decreasing layer), according to the layering result in Step 101, screen all the time t in and depth h in samples within the thickness h1 of the velocity-decreasing layer (i is the measured physical point, and n is the vertical sample number of point i). Use the thickness h i0 of the low-velocity layer and the vertical delay time t i0 obtained in Step 102 to normalize and correct the time t in and depth h in samples of the velocity-decreasing layer in each measured sample point, that is, the corrected sample time T in =t in -t i0 , and the sample depth H in =h in -h i0 ;

[0043] Step 104: Perform high-order polynomial fitting based on the least squares method on the normalized and corrected time-depth samples (T in , H in ). The fitting process is briefly described as follows:

[0044] Substitute n time-depth samples (T in , H in ) into

[0045] H(T)=a0 + a1T + a2T 2 + a3T 3 +......+ a k T k k∈(4, 5...10)

[0046] n equations are obtained. The error between the measured value and the fitting function based on the least squares method is used as the objective function, that is:

[0047] ε = ∑[H i -(a0 + a1T + a2T 2 + a3T 3 +......+ a k T k )] 2

[0048] To optimize the objective function ε, the partial derivatives of each coefficient are set to 0, and the "gradient descent method" or "solving linear equations" can be used to find all the coefficients a0, a1, a2.........a k .

[0049] During the fitting process, the power of the polynomial traverses the given power range from low to high, and the fitting polynomial with the smallest objective function error is calculated as the final result (see Figure 4 ). The time-depth conversion template model fitted in this area is as follows:

[0050] H = 7.69757E - 10T 5 - 2.03265E - 07T 4 - 2.13693E - 06T 3 + 7.03704E - 03T 2 + 1..1990T - 0.9927

[0051] In the above formula, H represents the sample depth, T is the vertical delay time corresponding to the depth. Taking the deep micro-log SKS1001 as an example, the vertical delay time corresponding to the measured depth of 230m is 125.8ms. Using this time-depth conversion template model to calculate, when the given delay time is 125.8ms, the calculated vertical depth is 232.3m, and the actual depth error is 2.3m, and the relative error rate is 1%.

[0052] Example 2

[0053] Step 101: The sample data is the same as in Example 1. Repeat steps 101 - step 103 in the example to obtain the corrected sample time T in = t in - t i0 , sample depth H in = h in - hi0 ;

[0054] Step 104: Perform a second-order polynomial fitting based on the least squares method on the normalized and corrected time-depth samples (T in , H in ). Substitute the samples into the formula H(T) = a0 + a1T + a2T 2 . The fitting calculation process is the same as that in Step 104 of Embodiment 1, and all coefficients a0, a1, and a2 can be obtained. The time-depth conversion template model for the second-order polynomial fitting in this area is as follows:

[0055] H = 0.003926112533463T 2 + 1.451079792895790T - 3.478484,094343120

[0056] In the above formula, H represents the sample depth, and T is the vertical delay time corresponding to the depth. Taking the deep micro-log SKS1001 as an example, the vertical delay time corresponding to the measured depth of 230 m is 125.8 ms. Using this time-depth conversion template model for calculation, when the given delay time is 125.8 ms, the calculated vertical depth is 241.2 m, and the actual depth error is 11.2 m, and the relative error rate is 4.9%.

[0057] The above detailed description is a specific description of one feasible embodiment of the present invention. This embodiment is not intended to limit the patent scope of the present invention. Any equivalent implementation or modification made without departing from the present invention shall be included within the scope of the technical solution of the present invention.

Claims

1. A method for making a time-depth conversion template for shallow conglomerate, characterized in that, It includes the following steps: (1) Select the time and depth data of the target area as the research samples, calculate the vertical velocity, and then layer according to the vertical velocity and lithology changes. From top to bottom, they are the low-velocity layer, the velocity-decreasing layer, and the high-velocity layer; (2) Obtain the thickness information according to the depth data in step (1), and perform regional spatial interpolation on the velocity and thickness information of the low-velocity layer to obtain and calculate the low-velocity layer thickness and vertical delay time of each sample; (3) Use the low-velocity layer thickness and vertical delay time obtained in step (2) to perform normalization correction on the time-depth data of the velocity-decreasing layer; (4) Use a high-order polynomial based on the least squares method to fit the time-depth data of the velocity-decreasing layer corrected in step (3), determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the velocity-decreasing layer; The normalization correction described in step (3) means subtracting the low-velocity layer thickness and vertical delay time obtained in step (2) from the time and depth data of the velocity-decreasing layer respectively.

2. The manufacturing method according to claim 1, wherein The research samples described in step (1) come from at least one of deep micro-logging, conventional micro-logging, and drill logging.

3. The manufacturing method according to claim 1, characterized in that, In step (1), the low-velocity layer is a floating soil loose gravel layer, the velocity-decreasing layer is a sand-mud-gravel mixed layer, and the high-velocity layer is a cemented gravel layer.

4. The manufacturing method according to claim 1, characterized in that The method of spatial interpolation described in step (2) is the inverse distance weighted method based on a triangular grid.

5. The manufacturing method according to claim 1, characterized in that, The high-order polynomial described in step (4) is H(T) = a0 + a1T + a2T 2 + a3T 3 +......+ a k T k k ∈ (4, 5...10); Take different values of k, substitute them into the high-order polynomial for fitting, and calculate the error between the measured value and the fitting function. The value with the smallest error is the final k value.

6. The manufacturing method according to claim 1, wherein It also includes step (5): Referring to the methods in steps (2)-(4), use the low-velocity layer and the velocity-decreasing layer to perform normalization correction on the time-depth data of the high-velocity layer, use a high-order polynomial based on the least squares method to fit the time-depth data of the corrected high-velocity layer, determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the high-velocity layer.

7. The manufacturing method according to claim 6, characterized in that, In step (5), perform regional spatial interpolation on the velocity and thickness information of the low-velocity layer and the velocity-decreasing layer, obtain and calculate the low-velocity layer thickness and vertical delay time of each sample, then perform normalization correction on the time-depth data of the high-velocity layer, use a high-order polynomial based on the least squares method to fit the time-depth data of the corrected high-velocity layer, determine the parameters, and the obtained high-order polynomial is the time-depth conversion template of the high-velocity layer.

8. The manufacturing method according to claim 1, characterized in that, The calculation method of the vertical delay time described in step (2) is vertical delay time = low-velocity layer thickness / low-velocity layer average velocity.

9. The manufacturing method according to claim 1, wherein In step (3), the time-depth data of the corrected velocity-decreasing layer are superimposed and displayed in the same coordinate system, and the consistency of the time-depth relationship at different positions is analyzed. If they are consistent, it means the data is reasonable; if not, the velocity-decreasing layer data needs to be reselected for correction.

Images