Clastic rock reservoir seismic sequence high-resolution division method and system

Abstract

The invention provides a clastic rock reservoir seismic sequence high-resolution division method and system, and the method comprises the steps: S1, obtaining a sensitive logging curve of a well closest to a seismic channel, carrying out the processing of the sensitive logging curve through a cumulative prediction error method, obtaining a cumulative prediction error curve, carrying out the generalized S transformation of the sensitive logging curve, and obtaining a time-frequency distribution feature; s2, performing long-term and short-term cycle sequence interface identification by using the cumulative prediction error curve and the time-frequency distribution characteristics to obtain long-term and short-term cycle sequence interfaces; and S3, calibrating the long-term and short-term cycle sequence interfaces to well-side seismic traces, carrying out seismic sequence division on the well-side seismic traces, carrying out well-seismic time-frequency matching analysis in a time-frequency domain by adopting an atomic matching pursuit technology, identifying seismic hidden sequence interfaces, and obtaining the high-resolution seismic sequence framework of the clastic rock reservoir after all seismic traces are calculated. According to the method, the high-resolution seismic sequence framework is obtained, and powerful support is provided for improving the precision and accuracy of clastic rock reservoir prediction.

Description

Technical Field

[0001] The invention belongs to the technical field of oil and gas exploration, and in particular to a method and system for high-resolution division of seismic sequences of clastic rock reservoirs. Background Art

[0002] The focus of oil and gas exploration and development has gradually shifted to dense, low-permeability, deep and other fields. The target scale has gradually decreased, and the difficulty of searching for favorable targets has increased. Carrying out fine reservoir prediction under the system domain framework can more accurately predict the hidden lithological targets within the system domain. High-resolution sequence framework is the basis for conducting seismic inversion in areas without well control and quantitative reservoir prediction using deterministic sedimentary modeling methods. Therefore, the accuracy of sequence division and interpretation will directly affect the effectiveness and accuracy of comprehensive interpretation of seismic data and seismic reservoir prediction. How to create a high-precision isochronous sequence stratigraphic framework is particularly important.

[0003] In the actual process of dividing sequences, manual division based on logging lithology interpretation accounts for a large proportion, but this method is inefficient, highly subjective, and difficult to guarantee accuracy. Based on this, some scholars have proposed the integrated prediction error filtering analysis method (INPEFA) for stratigraphic interface division and isochronous stratigraphic correlation. This method processes and analyzes the logging curves, extracts the spectral trend lines reflecting the stratigraphic superposition and cyclic structure, and completes the isochronous correlation of the stratigraphic layers based on the inflection points and change trends, laying the foundation for accurate stratigraphic division and true isochronous correlation. However, this method is difficult to avoid the problem of "one-hole view" of logging data, and cannot provide accurate lateral distribution of the sequence. Summary of the invention

[0004] In order to solve the problems existing in the prior art, the present invention provides a method and system for high-resolution division of seismic sequences of clastic reservoirs. The method uses the cumulative prediction error method and the atomic matching pursuit method to perform special processing on seismic data to obtain a high-resolution seismic sequence framework, which provides strong support for improving the precision and accuracy of clastic reservoir prediction.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for high-resolution division of seismic sequences of clastic reservoirs, the specific steps of which include:

[0006] S1 obtains the sensitive logging curve of the well closest to the seismic track, processes the sensitive logging curve using the cumulative prediction error method to obtain the cumulative prediction error curve, and performs generalized S transformation on the sensitive logging curve to obtain the time-frequency distribution characteristics;

[0007] S2 uses the cumulative prediction error curve and time-frequency distribution characteristics to identify the long-term and short-term cyclic sequence interfaces and obtains the long-term and short-term cyclic sequence interfaces;

[0008] S3 calibrates the long- and short-term cyclic sequence interfaces to the seismic traces near the well, divides the seismic sequences of the seismic traces near the well, and uses the atomic matching pursuit technology to perform well-seismic time-frequency matching analysis in the time-frequency domain to identify the seismic implicit sequence interfaces. After all seismic traces are calculated, a high-resolution seismic sequence framework for the clastic reservoir is obtained.

[0009] Furthermore, in S1, the sensitive logging curve is a logging curve that has a good reflection of the stratigraphic superposition relationship and the cyclic structure.

[0010] Furthermore, in S1, the logging curve is processed by using the cumulative prediction error method to obtain the cumulative prediction error curve. The specific steps are as follows:

[0011] 1) Perform maximum entropy spectrum estimation on the sensitive logging curve x(t) to obtain the maximum entropy spectrum S(f) and the maximum entropy spectrum coefficient a k ;

[0012] 2) Using the maximum entropy spectrum coefficient a k Perform curve prediction to obtain discretized predicted logging curve

[0013]

[0014] Where n is the sampling point number, n>M, is the predicted value of the curve at the nth sampling point;

[0015] 3) Estimate the cumulative prediction error of the logging data and obtain the cumulative prediction error curve INPEFA(n):

[0016]

[0017] Where x(i) is the original logging curve value at the i-th sampling point, is the predicted logging curve value at the i-th sampling point.

[0018] Furthermore, in S2, the local extreme value of high-frequency energy on the time-frequency distribution diagram corresponding to the inflection point of the cumulative prediction error curve is the sequence interface.

[0019] Furthermore, in S3, the atomic matching pursuit technique was used to perform a step-by-step refined well-seismic time-frequency matching analysis in the time-frequency domain, with the third-order sequence interface and the long-term cyclic sequence interface as constraints.

[0020] Furthermore, in S3, the nearest well is selected for each seismic trace, and S1 and S2 are performed to obtain the long- and short-term cyclic sequence interfaces, and then S3 is performed to identify the seismic implicit sequence interface under the well-seismic time-frequency matching analysis.

[0021] Furthermore, in S3, the specific steps are as follows:

[0022] 1) Use atomic matching pursuit technology to calculate the time-frequency distribution of the seismic trace near the well and obtain the seismic time-frequency distribution map;

[0023] 2) Using the long-term cyclic sequence interface, the earthquake time-frequency distribution map is divided into multiple regions in the time direction;

[0024] 3) Pick up two frequency separation lines in the frequency direction and divide the earthquake time-frequency distribution diagram into three parts: low frequency, medium frequency and high frequency;

[0025] 4) Use the amplitudes of low, medium and high frequencies in the seismic time-frequency distribution diagram to identify the stratigraphic sequence of the well side channel and obtain the seismic implicit stratigraphic interface.

[0026] Furthermore, in S3, the method of picking up the two frequency separation lines is as follows: draw a separation line from the low frequency and the high frequency, move toward the center, calculate the amplitude average A in each grid, and when the variance σ of each grid 2 At maximum, the frequency separation line is determined by:

[0027]

[0028] Furthermore, in S3, the amplitude averages of low frequency, medium frequency and high frequency are calculated, and the amplitude averages are converted into numerical values ​​of the three primary colors RGB. The low frequency amplitude average corresponds to the R color value, the medium frequency amplitude average corresponds to the G color value, and the high frequency amplitude average corresponds to the B color value. Different colors are used to represent the stratigraphic characteristics of the seismic record.

[0029] The present invention also provides a high-resolution seismic sequence division system for clastic reservoirs, comprising:

[0030] The time-frequency distribution feature acquisition module is used to obtain the sensitive logging curve of the well closest to the seismic track, process the sensitive logging curve using the cumulative prediction error method to obtain the cumulative prediction error curve, and perform generalized S transformation on the sensitive logging curve to obtain the time-frequency distribution feature;

[0031] The sequence interface recognition module is used to identify the long-term and short-term cyclic sequence interfaces by using the cumulative prediction error curve and time-frequency distribution characteristics to obtain the long-term and short-term cyclic sequence interfaces;

[0032] The seismic sequence framework construction module is used to calibrate the long- and short-term cyclic sequence interfaces to the seismic traces near the well, divide the seismic sequences based on the seismic traces near the well, use the atomic matching pursuit technology to perform well-seismic time-frequency matching analysis in the time-frequency domain, and identify the seismic implicit sequence interfaces. After all the seismic traces are calculated, a high-resolution seismic sequence framework for the clastic reservoir is obtained.

[0033] Compared with the prior art, the present invention has at least the following beneficial effects:

[0034] The present invention proposes a high-resolution seismic sequence division method for clastic reservoirs, which uses the cumulative prediction error method to calculate the INPEFA curve from the sensitive logging curve, and combines the generalized S transform to identify the long-term and short-term cyclic sequence interfaces; then, the atomic matching pursuit technology is used to divide the wellside seismic trace sequence, and the logging sequence interface picked up by the INPEFA curve is used. On the basis of the vertical high-resolution sequence division of the logging data, the seismic data is interpreted with high resolution to obtain a high-precision spatial seismic sequence framework, thereby improving the rate and quality of sequence division, and mapping the logging sequence division to the seismic data to highlight the seismic implicit sequence interface characteristics, so that the horizontal and vertical resolutions of the sequence are unified, providing an important basis for the combination of geological and land object data, providing a good basic support for seismic comprehensive interpretation and high-precision reservoir prediction, and providing technical support for clastic oil and gas exploration. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a flow chart of the high-resolution division method of seismic sequences in clastic reservoirs.

[0036] Figure 2 It is the cumulative prediction error curve and the time-frequency distribution diagram of atomic matching pursuit.

[0037] Figure 3 It is a high-resolution seismic sequence framework map. DETAILED DESCRIPTION

[0038] The specific embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings. The illustrative examples and descriptions of the present invention are used to explain the present invention, but are not intended to limit the present invention.

[0039] like Figure 1 As shown, the present invention provides a high-resolution (fourth to fifth order) division method of clastic reservoir seismic sequence based on cumulative prediction error, and the specific steps are as follows:

[0040] Step 1: Select sensitive logging curves that are sensitive to the stratigraphic superposition relationship and cyclic structure, such as natural gamma GR, natural potential SP, etc.;

[0041] Step 2: Use the cumulative prediction error method to process the sensitive logging curve to obtain the cumulative prediction error curve, as follows:

[0042] First, the maximum entropy spectrum of the sensitive logging curve x(t) is estimated using formula (1) to obtain the maximum entropy spectrum S(f) and the maximum entropy spectrum coefficient a k :

[0043]

[0044] Where S(f) is the maximum entropy spectrum, fk is the frequency, a k is the coefficient, t is the time, j is the imaginary part, and M is the entropy spectrum order.

[0045] Secondly, the maximum entropy spectrum coefficient a k Substitute into formula (2), carry out curve prediction, and obtain the discretized prediction logging curve

[0046]

[0047] Where n is the sampling point number, n>M, is the predicted value of the curve at the nth sampling point.

[0048] Next, the cumulative prediction error of the logging data is estimated to obtain the cumulative prediction error curve INPEFA(n), which can reflect the trend change characteristics of the formation sedimentary cycle:

[0049]

[0050] Where x(i) is the original logging curve value at the i-th sampling point, is the predicted logging curve value at the i-th sampling point.

[0051] Step 3: Perform generalized S transformation on the sensitive logging curve, extract the time-frequency characteristics therein, and obtain the time-frequency distribution characteristics of the sensitive logging curve at the wellbore;

[0052] Step 4: Compare the cumulative prediction error curve and the time-frequency distribution characteristics. If the inflection point of the cumulative prediction error curve corresponds to the local extreme point of high-frequency energy of the time-frequency distribution characteristics, it is the sequence interface, and the identification and comparative analysis of the long-term and short-term cycle sequence interface are completed;

[0053] Step 5: Calibrate the long-term and short-term cyclic sequence interfaces interpreted from the sensitive logging curves to the seismic traces near the wells, and divide the seismic sequences on the seismic traces near the wells (generally, the long-term cycles divided on the logging curves correspond to the third-order seismic sequences, and the short-term cycles correspond to the fourth and fifth-order seismic sequences). Using the third-order sequence interface and the long-term cyclic sequence interface as constraints, the atomic matching pursuit technology is used to perform a step-by-step refined well-seismic time-frequency matching analysis in the time-frequency domain;

[0054] Atomic matching pursuit is a signal processing technique that is often used for time-frequency analysis of different signals. The idea of ​​the matching pursuit algorithm is to project the original signal onto a series of time-frequency atoms, that is, to represent the original signal as a linear combination of these time-frequency atoms, and use these time-frequency atoms to accurately express the original signal. The formula is as follows:

[0055]

[0056] In the formula, Wavelet Atom The Wigner distribution function, a n is the atomic coefficient, t is the time, f is the frequency, and N is the order;

[0057] Firstly, the atomic matching pursuit technique is used to calculate the time-frequency distribution of the seismic traces near the well and obtain the time-frequency distribution diagram.

[0058] Secondly, the time-frequency distribution diagram is divided into multiple regions in the time direction using the long-term cyclic sequence interface picked up in step 4.

[0059] Then, two frequency separation lines are picked up in the frequency direction to divide the time-frequency distribution diagram into three parts: low frequency, medium frequency, and high frequency. At this time, the time-frequency distribution diagram becomes a grid with multiple rows and three columns.

[0060] The method of picking up the two frequency separation lines is: draw a separation line from the low frequency and high frequency, move it toward the center, calculate the amplitude average A in each grid, and when the variance σ of each grid is 2 When the maximum is reached, the frequency separation line is determined.

[0061]

[0062] Where μ is the average value of A and N is the number of sample points.

[0063] Finally, the amplitude of low, medium and high frequencies in the time-frequency distribution diagram is used to identify the sequence of the well bypass and obtain the seismic implicit sequence interface. Specifically, the average amplitude of low frequency, medium frequency and high frequency is calculated, and the average amplitude is converted into the value of the three primary colors RGB. The low frequency amplitude average corresponds to the R color value, the medium frequency amplitude average corresponds to the G color value, and the high frequency amplitude average corresponds to the B color value.

[0064] The RGB primary color values ​​are between 0 and 255. The specific setting standard is: the maximum value of the amplitude averages of the three frequency bands corresponds to a color value of 255, the minimum value corresponds to a color value of 0, and the intermediate value is standardized so that the corresponding color value is between 0 and 255. The corresponding color of the seismic trace near the well is obtained, and different colors are used to represent the stratigraphic characteristics of the seismic record.

[0065] Step 6: For each seismic trace, first select the nearest well and perform steps 2 to 4 to obtain the long- and short-term cyclic sequence interfaces. Then perform step 5 to identify the seismic implicit sequence interfaces under the well-seismic time-frequency matching analysis. After all seismic traces are calculated, a high-resolution seismic sequence framework for the clastic reservoir can be established.

[0066] Figure 1This is a flow chart of the high-resolution division method of seismic stratigraphic sequences in clastic reservoirs. This method mainly uses the cumulative prediction error method and atomic matching pursuit technology to perform high-resolution sequence division. First, the INPEFA curve is calculated from the well logging curve using the cumulative prediction error method, and the long-term and short-term cycle variation characteristics are identified by combining the generalized S transform. Then, the atomic matching pursuit technology is used to divide the stratigraphic sequence of the seismic trace near the well. Finally, the seismic implicit sequence interface characteristics are calculated to obtain a high-resolution seismic sequence framework.

[0067] Figure 2 The cumulative prediction error curve (INPEFA curve) and the atomic matching pursuit time-frequency distribution diagram are used to map the well logging sequence interface picked up by the INPEFA curve onto the seismic data under the well-seismic time-frequency matching analysis to obtain the sequence interface of the seismic record.

[0068] Figure 3 It is a high-resolution seismic sequence framework map, which conducts well-seismic time-frequency matching analysis on seismic records of different cycles and performs high-resolution seismic sequence division under a step-by-step refinement strategy. Figure 3 It is difficult to trace the fourth-order sequence interface on the medium-level seismic section, but it can be clearly identified on the seismic cycle section (shown by the ellipse).

Claims

1. A high-resolution seismic sequence division method for clastic reservoirs, characterized in that: The specific steps include: S1 obtains the sensitive logging curve of the well closest to the seismic track, processes the sensitive logging curve using the cumulative prediction error method to obtain the cumulative prediction error curve, and performs generalized S transformation on the sensitive logging curve to obtain the time-frequency distribution characteristics; S2 uses the cumulative prediction error curve and time-frequency distribution characteristics to identify the long-term and short-term cyclic sequence interfaces and obtains the long-term and short-term cyclic sequence interfaces; S3 calibrates the long- and short-term cyclic sequence interfaces to the seismic traces near the well, divides the seismic sequences of the seismic traces near the well, and uses the atomic matching pursuit technology to perform well-seismic time-frequency matching analysis in the time-frequency domain to identify the seismic implicit sequence interfaces. After all seismic traces are calculated, a high-resolution seismic sequence framework for the clastic reservoir is obtained.

2. A high-resolution seismic sequence division method for clastic reservoirs according to claim 1, characterized in that: In S1, the sensitive logging curve is a logging curve that has a good reflection of the stratigraphic superposition relationship and cyclic structure.

3. The high-resolution seismic sequence division method for clastic reservoirs according to claim 1 is characterized in that: In S1, the cumulative prediction error method is used to process the logging curve to obtain the cumulative prediction error curve. The specific steps are as follows: 1) Perform maximum entropy spectrum estimation on the sensitive logging curve x(t) to obtain the maximum entropy spectrum S(f) and the maximum entropy spectrum coefficient a k ; 2) Using the maximum entropy spectrum coefficient a k Perform curve prediction to obtain discretized predicted logging curve Where n is the sampling point number, n>M, is the predicted value of the curve at the nth sampling point; 3) Estimate the cumulative prediction error of the logging data and obtain the cumulative prediction error curve INPEFA(n): Where x(i) is the original logging curve value at the i-th sampling point, is the predicted logging curve value at the i-th sampling point.

4. The high-resolution seismic sequence division method for clastic reservoirs according to claim 1 is characterized in that: In S2, the local extreme value of high-frequency energy on the time-frequency distribution diagram corresponding to the inflection point of the cumulative prediction error curve is the sequence interface.

5. The high-resolution seismic sequence division method for clastic reservoirs according to claim 1 is characterized in that: In S3, the third-order sequence interface and the long-term cyclic sequence interface are used as constraints, and the atomic matching pursuit technology is used to perform a step-by-step refined well-seismic time-frequency matching analysis in the time-frequency domain.

6. The high-resolution seismic sequence division method for clastic reservoirs according to claim 1, characterized in that: In S3, the nearest well is selected for each seismic trace, and S1 and S2 are performed to obtain the long- and short-term cyclic sequence interfaces. Then S3 is performed to identify the seismic implicit sequence interface under the well-seismic time-frequency matching analysis.

7. The high-resolution seismic sequence division method for clastic reservoirs according to claim 1, characterized in that: In S3, the specific steps are as follows: 1) Use atomic matching pursuit technology to calculate the time-frequency distribution of the seismic trace near the well and obtain the seismic time-frequency distribution map; 2) Using the long-term cyclic sequence interface, the earthquake time-frequency distribution map is divided into multiple regions in the time direction; 3) Pick up two frequency separation lines in the frequency direction and divide the earthquake time-frequency distribution diagram into three parts: low frequency, medium frequency and high frequency; 4) Use the amplitudes of low, medium and high frequencies in the seismic time-frequency distribution diagram to identify the stratigraphic sequence of the well bypass and obtain the seismic implicit stratigraphic interface.

8. A high-resolution seismic sequence division method for clastic reservoirs according to claim 7, characterized in that: In S3, the two frequency separation lines are picked up as follows: draw a separation line from the low frequency and high frequency, move it toward the center, calculate the amplitude average A in each grid, and when the variance σ of each grid is 2 At maximum, the frequency separation lines are determined by:

9. A high-resolution seismic sequence division method for clastic reservoirs according to claim 7, characterized in that: In S3, the amplitude averages of low frequency, medium frequency and high frequency are calculated and converted into the values ​​of the three primary colors RGB. The low frequency amplitude average corresponds to the R color value, the medium frequency amplitude average corresponds to the G color value, and the high frequency amplitude average corresponds to the B color value. Different colors are used to represent the stratigraphic characteristics of the seismic record.

10. A high-resolution seismic sequence division system for clastic reservoirs, characterized in that: include: The time-frequency distribution feature acquisition module is used to obtain the sensitive logging curve of the well closest to the seismic track, process the sensitive logging curve using the cumulative prediction error method to obtain the cumulative prediction error curve, and perform generalized S transformation on the sensitive logging curve to obtain the time-frequency distribution feature; The sequence interface recognition module is used to identify the long-term and short-term cyclic sequence interfaces by using the cumulative prediction error curve and time-frequency distribution characteristics to obtain the long-term and short-term cyclic sequence interfaces; The seismic sequence framework construction module is used to calibrate the long- and short-term cyclic sequence interfaces to the seismic traces near the well, divide the seismic sequences based on the seismic traces near the well, use the atomic matching pursuit technology to perform well-seismic time-frequency matching analysis in the time-frequency domain, and identify the seismic implicit sequence interfaces. After all the seismic traces are calculated, a high-resolution seismic sequence framework for the clastic reservoir is obtained.

Images