Earthquake frequency-division inversion sandstone prediction method

Abstract

The invention discloses an earthquake frequency division inversion sandstone prediction method, and solves the problem of low sandstone prediction precision caused by mutual interference of different frequency information of full-frequency earthquake data in an existing sandstone prediction method. The method comprises the following steps: S1, preferably selecting a sandstone sensitivity curve; s2, obtaining a seismic data volume and a frequency range thereof; s3, performing generalized S time-frequency transformation and band-pass filtering operation on the seismic data volumes to obtain m seismic data volumes with different bandwidths; s4, m Ricker seismic wavelets are constructed; s5, sparse pulse inversion is carried out by using the Ricker seismic wavelets and the seismic data volume with the same bandwidth, and a wave impedance data volume with the same bandwidth is obtained; s6, extracting m wave impedance curves at the well points, and performing correlation calculation on the wave impedance curves and the sandstone sensitivity curve; s7, obtaining a fusion wave impedance data body; and S8, obtaining a target curve data volume to realize sandstone prediction. The method can improve sandstone prediction precision and improve oil-gas exploration and development effects.

Description

Technical Field

[0001] This invention relates to the field of reservoir sandstone prediction technology, and in particular to a seismic frequency division inversion method for predicting sandstone. Background Technology

[0002] Seismic data is frequently used in lithology prediction, which can improve the accuracy of sandstone prediction and is beneficial for the precise exploration and efficient development of clastic oil layers in my country.

[0003] Currently, the most commonly used seismic data is full-frequency seismic data, which has a certain frequency range, generally 0–80 Hz. Due to the mutual interference between different frequencies of full-frequency seismic data, effective signals are masked and there is a high degree of ambiguity, which affects the accuracy of sandstone prediction. Summary of the Invention

[0004] This invention addresses the problem of low prediction accuracy in existing sandstone prediction methods due to interference from different frequencies in full-frequency seismic data. It provides a seismic frequency-division inversion method for predicting sandstone. This method divides earthquakes into frequencies at certain intervals and then comprehensively utilizes seismic information from different frequency bands to improve prediction accuracy and enhance oil and gas exploration and development.

[0005] The present invention solves its problem through the following technical solution: the seismic frequency division inversion prediction method for sandstone includes the following steps:

[0006] S1: Obtain multiple logging curves related to reservoir properties through well logging, and select the sandstone-sensitive curve from the multiple logging curves;

[0007] S2: Seismic data volumes are obtained through field seismic acquisition, and the frequency range of the acquired seismic data volumes is statistically analyzed as f1~f2;

[0008] S3: Perform generalized S-time-frequency transform and bandpass filtering on the seismic data volume obtained in step S2 to obtain m seismic data volumes with different bandwidths, where m is equal to the integer part of (f1~f2) / 10;

[0009] S4: Construct m Reck seismic wavelets, where the bandwidth of each wavelet is the same as the bandwidth of each seismic data volume;

[0010] S5: Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, sparse pulse inversion is carried out to obtain the same bandwidth of wave impedance data volume, resulting in m wave impedance data volumes with different frequency ranges.

[0011] S6: Based on the obtained m wave impedance data volumes of different frequency ranges, extract m wave impedance curves at the well point, calculate the correlation between the wave impedance curves and the sandstone sensitivity curve to obtain m correlation coefficients, and normalize the correlation coefficients to obtain m weight coefficients.

[0012] S7: Based on the obtained m weight coefficients, perform multivariate regression calculations to obtain a fused wave impedance data volume;

[0013] S8: Using the sandstone sensitive curve and the fused wave impedance data volume, the waveform indication inversion algorithm is used to obtain the target curve data volume, thereby realizing sandstone prediction.

[0014] Furthermore, the method for selecting the sandstone-sensitive curve from the multiple logging curves in step S1 is as follows:

[0015] Multiple logging curves related to reservoir properties were obtained through well logging. Cross plot analysis was performed on different logging curves and lithological data to determine the most sensitive and effective logging curve for distinguishing sandstone from non-sandstone, and the value range of the logging curve for sandstone was obtained.

[0016] Furthermore, the logging curves related to reservoir properties include spontaneous potential, resistivity, gamma, sonic transit time, and density curves.

[0017] Furthermore, the method for performing generalized S-time-frequency transform and bandpass filtering on the seismic data volume in step S3 to obtain m seismic data volumes with different bandwidths is as follows:

[0018] A generalized S-time-frequency transform operation is performed on the seismic data volume to form a frequency domain seismic data volume.

[0019] Then, through bandpass filtering, seismic waves of a specified bandwidth are allowed to pass through and be output, resulting in m seismic data volumes with different bandwidths. The bandwidths of each seismic data volume are 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m, where m is equal to the integer part of (f1~f2) / 10.

[0020] Furthermore, the method for constructing m Reck seismic wavelets in step S4 is as follows:

[0021] By manually constructing m Reck seismic wavelets, each wavelet has a time width of 80-100ms and a frequency bandwidth of 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m.

[0022] Furthermore, the method for obtaining the wave impedance data volumes of m different frequency ranges in step S5 is as follows:

[0023] Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, with the longitudinal sampling frequency of the seismic data volume set to 1ms, sparse pulse inversion was carried out. That is, under the constraint of the low-frequency model, the Lake seismic wavelet and the seismic data volume were used to perform deconvolution calculation to obtain the wave impedance data volume. A total of m wave impedance data volumes with different frequency widths were obtained, namely 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m.

[0024] Furthermore, step S6 involves adjusting the wave impedance curve A. i The method for calculating the correlation between G(t) and the sandstone sensitivity curve G(t) to obtain m correlation coefficients is as follows:

[0025] Based on wave impedance curve A i The correlation coefficient is calculated using the correlation coefficient formula between G(t) and the sandstone sensitivity curve G(t).

[0026] The formula for correlation coefficient data is:

[0027]

[0028] In the formula: Cov is the covariance; D is the variance; t is the time depth of the curve;

[0029] i represents different wave impedance curves, with values ​​of 1, 2, ..., m.

[0030] Furthermore, the method for obtaining the m weight coefficients is as follows: based on the obtained m correlation coefficients, a normalization calculation is performed, that is, the m correlation coefficients are summed, and then the ratio of each correlation coefficient to the sum is calculated to obtain the m weight coefficients.

[0031] Furthermore, the specific method for obtaining the fused impedance data volume in step S7 includes the following steps: using a multiple regression algorithm, multiplying each impedance data volume by its respective weight coefficient, and then adding all the results together to obtain the fused impedance data volume.

[0032] Furthermore, the specific method for obtaining the target curve data volume using the sandstone sensitive curve and the fused wave impedance data volume in step S8 is as follows:

[0033] Using the sandstone sensitive curve and the fused wave impedance data volume, a correspondence between the sandstone sensitive curve and the fused wave impedance curve at the same location is established. Then, based on the correspondence and using the fused wave impedance data volume as an indicator, the data volume of the target curve is obtained.

[0034] Furthermore, step S8 utilizes the obtained target curve data volume to achieve sandstone prediction. The specific method is as follows:

[0035] Based on the value range of sandstone within the target curve, the target curve data volume is transformed into a sandstone distribution data volume; using the obtained sandstone distribution data volume, the distribution of sandstone is predicted.

[0036] Compared with the above-mentioned background technology, the present invention has the following beneficial effects:

[0037] This invention relates to a seismic frequency division inversion method for predicting sandstone. It divides earthquakes into frequencies at certain intervals and then comprehensively utilizes seismic information from different frequency bands. Through a seismic frequency division inversion algorithm, it achieves precise prediction of sandstone formation, thereby improving the effectiveness of oil and gas exploration and development.

[0038] The seismic frequency division inversion prediction method for sandstone in this invention is applicable to clastic oil layers, and the calculation results are reasonable and highly accurate. Attached Figure Description

[0039] Figure 1 This is a flowchart of the earthquake frequency division inversion prediction method for sandstone according to the present invention;

[0040] Figure 2 These are seismic data volumes of different bandwidths in the NSQ block according to an embodiment of the present invention; (wherein figures a, b, c, d, e, f, and g are seismic data volumes at bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively;)

[0041] Figure 3 These are the Ricker seismic wavelets of different bandwidths in the NSQ block of this invention; (wherein figures a, b, c, d, e, f, and g are Ricker seismic wavelets with bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively;)

[0042] Figure 4 This is the wave impedance data body of the NSQ block with different bandwidths in the embodiment of the present invention; (wherein figures a, b, c, d, e, f, and g are the wave impedance data bodies at bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively;)

[0043] Figure 5 This is the sandstone prediction result for the NSQ block in this embodiment of the invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0045] like Figure 1As shown, a seismic frequency division inversion method for predicting sandstone includes the following steps:

[0046] S1: Obtain multiple logging curves related to reservoir properties through well logging, and select the sandstone-sensitive curve from the multiple logging curves;

[0047] Multiple logging curves related to reservoir properties were obtained through well logging. Cross plot analysis was performed on different logging curves and lithological data to determine the most sensitive and effective logging curve for distinguishing sandstone from non-sandstone, and the value range of the logging curve for sandstone was obtained.

[0048] The logging curves related to reservoir properties include curves for spontaneous potential, resistivity, gamma, sonic transit time, and density.

[0049] S2: Seismic data volumes are obtained through field seismic acquisition, and the frequency range of the acquired seismic data volumes is statistically analyzed as f1~f2;

[0050] S3: Perform generalized S-time-frequency transform and bandpass filtering on the seismic data volume obtained in step S2 to obtain m seismic data volumes with different bandwidths. The specific method is as follows:

[0051] A generalized S-time-frequency transform operation is performed on the seismic data volume to form a frequency domain seismic data volume.

[0052] Then, through bandpass filtering, seismic waves of a specified bandwidth are allowed to pass through and be output, resulting in m seismic data volumes with different bandwidths. The bandwidths of each seismic data volume are 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m, where m is equal to the integer part of (f1~f2) / 10.

[0053] S4: Construct m Reck seismic wavelets, each with the same bandwidth as the seismic data volume: 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m; each wavelet has a time width of 80-100ms.

[0054] S5: Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, sparse pulse inversion is carried out to obtain the same bandwidth of wave impedance data volume, resulting in m wave impedance data volumes with different frequency ranges.

[0055] The method for obtaining the wave impedance data volumes of m different frequency ranges in step S5 is as follows:

[0056] Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, with the longitudinal sampling frequency of the seismic data volume set to 1ms, sparse pulse inversion was carried out. That is, under the constraint of the low-frequency model, the Lake seismic wavelet and the seismic data volume were used to perform deconvolution calculation to obtain the wave impedance data volume. A total of m wave impedance data volumes with different frequency widths were obtained, namely 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m.

[0057] S6: Based on the obtained m wave impedance data volumes of different frequency ranges, extract m wave impedance curves at the well point, calculate the correlation between the wave impedance curves and the sandstone sensitivity curve to obtain m correlation coefficients, and normalize the correlation coefficients to obtain m weight coefficients.

[0058] Step S6 involves the wave impedance curve A. i The method for calculating the correlation between G(t) and the sandstone sensitivity curve G(t) to obtain m correlation coefficients is as follows:

[0059] Based on wave impedance curve A i The correlation coefficient is calculated using the correlation coefficient formula between G(t) and the sandstone sensitivity curve G(t).

[0060] The formula for correlation coefficient data is:

[0061]

[0062] In the formula: Cov is the covariance; D is the variance;

[0063] t represents the time depth of the curve;

[0064] i represents different wave impedance curves, with values ​​of 1, 2, ..., m; the correlation coefficient R is calculated. i (t);

[0065] The method for obtaining m weight coefficients is as follows: based on the obtained m correlation coefficients, normalization calculation is performed, that is, the m correlation coefficients are summed, and then the ratio of each correlation coefficient to the sum is calculated to obtain m weight coefficients.

[0066] S7: Based on the obtained m weight coefficients, perform multivariate regression calculations to obtain a fused wave impedance data volume;

[0067] A multivariate regression algorithm is used to multiply each wave impedance data volume by its respective weight coefficient, and then all the results are added together to obtain the fused wave impedance data volume.

[0068] S8: Using the sandstone sensitive curve and the fused wave impedance data volume, the waveform indication inversion algorithm is used to obtain the target curve data volume, thereby realizing sandstone prediction.

[0069] Using the sandstone sensitive curve and the fused wave impedance data volume, a correspondence between the sandstone sensitive curve and the fused wave impedance curve at the same location is established. Then, based on the correspondence and using the fused wave impedance data volume as an indicator, the data volume of the target curve is obtained.

[0070] Based on the value range of sandstone within the target curve, the target curve data volume is transformed into a sandstone distribution data volume; using the obtained sandstone distribution data volume, the distribution of sandstone is predicted.

[0071] Example 1

[0072] To make the objectives, technical solutions, and advantages of this invention clearer, the following description, using the NSQ block of Daqing Oilfield as an example, will be further detailed with reference to the accompanying drawings. The seismic frequency division inversion prediction method for sandstone includes the following steps:

[0073] Step 1: In the NSQ block of Daqing Oilfield, seismic data volumes were obtained through field seismic acquisition, and the frequency range of the seismic data volumes was statistically determined to be 6–90 Hz.

[0074] Step 2: In the NSQ block of Daqing Oilfield, multiple logging curves such as resistivity, spontaneous potential, gamma, sonic transit time, and density are obtained through well logging. From these multiple logging curves, the deep lateral resistivity curve is selected as the sandstone sensitive curve.

[0075] Step 3: Perform generalized S-time-frequency transform and bandpass filtering on the seismic data volume to obtain 8 seismic data volumes with different bandwidths, namely 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz. Figure 2 These are seismic data volumes of different bandwidths in the NSQ block. Figures a, b, c, d, e, f, and g represent seismic data volumes with bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively.

[0076] Step 4: Construct 8 Lake seismic wavelets, with bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz respectively. Figure 3 These are the Reck seismic wavelets of different bandwidths in the NSQ block, where figures a, b, c, d, e, f, and g represent the Reck seismic wavelets with bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively.

[0077] Step 5: Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, sparse pulse inversion is performed to obtain wave impedance data volume, resulting in a total of 7 wave impedance data volumes with different frequency ranges. Figure 4 These are impedance data volumes for different bandwidths of the NSQ block, where figures a, b, c, d, e, f, and g represent impedance data volumes for bandwidths of 10-20Hz, 20-30Hz, 30-40Hz, 40-50Hz, 50-60Hz, 60-70Hz, and 70-80Hz, respectively.

[0078] Step 6: Based on the 7 acoustic impedance data volumes, extract 7 acoustic impedance curves at the well points, calculate the correlation between the acoustic impedance curves and the sandstone sensitivity curves to obtain 7 correlation coefficients, and normalize the correlation coefficients to obtain 7 weight coefficients.

[0079] Step 7: Using the weighting coefficients, perform multivariate regression calculations to obtain a fused wave impedance data volume.

[0080] Step 8: Using the sandstone sensitive curve and the fused wave impedance data volume, the waveform indication inversion algorithm is used to obtain the target curve data volume. Based on the value range of sandstone in the target curve, the target curve data volume is transformed into a sandstone distribution data volume, thereby realizing sandstone prediction. Figure 5 This is the prediction result for sandstone in the NSQ block.

[0081] The embodiments of this disclosure have been described above. These descriptions are exemplary and not exhaustive, and are not limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical applications, or technical improvements to the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

[0082] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the implementation methods of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of the present invention.

Claims

1. A method for predicting sandstone using seismic frequency division inversion, characterized in that: Includes the following steps: S1: Obtain multiple logging curves related to reservoir properties through well logging, and select the sandstone-sensitive curve from the multiple logging curves; S2: Seismic data volumes are obtained through field seismic acquisition, and the frequency range of the acquired seismic data volumes is statistically analyzed as f1~f2; S3: Perform generalized S-time-frequency transform and bandpass filtering on the seismic data volume obtained in step S2 to obtain m seismic data volumes with different bandwidths, where m is equal to the integer part of (f1~f2) / 10; S4: Construct m Reck seismic wavelets, where the bandwidth of each wavelet is the same as the bandwidth of each seismic data volume; S5: Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, sparse pulse inversion is carried out to obtain the same bandwidth of wave impedance data volume, resulting in m wave impedance data volumes with different frequency ranges. S6: Based on the obtained m wave impedance data volumes of different frequency ranges, extract m wave impedance curves at the well point, calculate the correlation between the wave impedance curves and the sandstone sensitivity curve to obtain m correlation coefficients, and normalize the correlation coefficients to obtain m weight coefficients. S7: Based on the obtained m weight coefficients, perform multivariate regression calculations to obtain a fused wave impedance data volume; S8: Using the sandstone sensitive curve and the fused wave impedance data volume, the waveform indication inversion algorithm is used to obtain the data volume of the target curve, thereby realizing sandstone prediction.

2. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The method for selecting the sandstone-sensitive curve from multiple logging curves in step S1 is as follows: Multiple logging curves related to reservoir properties were obtained through well logging; Cross-plot analysis was performed on different logging curves and lithological data to determine the most sensitive and effective logging curve for distinguishing between sandstone and non-sandstone, and the value range of the logging curve for sandstone was obtained.

3. A method for predicting sandstone using seismic frequency division inversion according to claim 1 or 2, characterized in that: The logging curves related to reservoir properties include: spontaneous potential, resistivity, gamma, sonic transit time, and density curves.

4. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The method for performing generalized S-time-frequency transform and bandpass filtering on the seismic data volume in step S3 to obtain m seismic data volumes with different bandwidths is as follows: A generalized S-time-frequency transform operation is performed on the seismic data volume to form a frequency domain seismic data volume. By using bandpass filtering, seismic waves of a specified bandwidth are allowed to pass through and be output, resulting in m seismic data volumes with different bandwidths. The bandwidths of each seismic data volume are 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m, where m is equal to the integer part of (f1~f2) / 10.

5. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The method for constructing m Reik seismic wavelets in step S4 is as follows: m Reck seismic wavelets were manually constructed, each with a time width of 80-100ms; the longitudinal sampling frequency was set to 1ms; and the bandwidths of each seismic wavelet were 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m.

6. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The method for obtaining the wave impedance data volumes of m different frequency ranges in step S5 is as follows: Using the same bandwidth of the Lake seismic wavelet and the seismic data volume, with the longitudinal sampling frequency of the seismic data volume set to 1ms, sparse pulse inversion was carried out. That is, under the constraint of the low-frequency model, the Lake seismic wavelet and the seismic data volume were used to perform deconvolution calculation to obtain the wave impedance data volume. A total of m wave impedance data volumes with different frequency widths were obtained, namely 10-20Hz, 20-30Hz, 30-40Hz, ..., 10(m-1)-10m.

7. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: Step S6 involves the wave impedance curve A. i The method for calculating the correlation between G(t) and the sandstone sensitivity curve G(t) to obtain m correlation coefficients is as follows: Based on wave impedance curve A i The correlation coefficient is calculated using the correlation coefficient formula between G(t) and the sandstone sensitivity curve G(t). The formula for correlation coefficient data is: In the formula: The function Cov represents the covariance; the function D represents the variance; and t represents the time depth of the curve. i represents different wave impedance curves, with values ​​of 1, 2, ..., m; the correlation coefficient R is calculated. i (t).

8. A method for predicting sandstone by seismic frequency division inversion according to claim 1 or 7, characterized in that: The method for obtaining the m weight coefficients is as follows: Based on the obtained m correlation coefficients, a normalization calculation is performed, which involves summing the m correlation coefficients. Then, the ratio of each correlation coefficient to the sum is calculated to obtain m weight coefficients.

9. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The specific method for obtaining the fused impedance data volume in step S7 is as follows: A multivariate regression algorithm is used to multiply each wave impedance data volume by its respective weight coefficient, and then all the results are added together to obtain the fused wave impedance data volume.

10. The method for predicting sandstone by seismic frequency division inversion according to claim 1, characterized in that: The specific method for obtaining the target curve data volume using the sandstone sensitive curve and the fused wave impedance data volume in step S8 is as follows: Using the sandstone sensitive curve and the fused wave impedance data volume, a correspondence between the sandstone sensitive curve and the fused wave impedance curve at the same location is established. Then, based on the correspondence and using the fused wave impedance data volume as an indicator, the data volume of the target curve is obtained. Step S8 utilizes the obtained target curve data volume to achieve sandstone prediction. The specific method is as follows: Based on the value range of sandstone within the target curve, the target curve data volume is transformed into a sandstone distribution data volume; using the obtained sandstone distribution data volume, the distribution of sandstone is predicted.

Images