A high-resolution seismic processing method

Abstract

A high-resolution seismic processing method. It mainly solves the problem that the existing seismic processing methods have a narrow frequency range, insufficient resolution and cannot identify thin reservoirs. It includes the following steps: obtaining seismic data, using the well logging acoustic wave velocity curve and density curve, and using the well-seismic correlation coefficient constrained inverse distance weighted algorithm to interpolate and calculate the longitudinal wave impedance model; constructing a high-frequency seismic wavelet, and performing convolution operation with the longitudinal wave impedance model to obtain high-frequency synthetic seismic data; performing frequency fusion calculation with the measured seismic data to obtain broadband seismic data. This high-resolution seismic processing method can obtain broadband and high-resolution seismic data, so that the seismic data can improve the resolution of thin reservoirs with a thickness of more than 3m, and improve the exploration and development effect.

Description

Technical Field

[0001] The invention relates to the field of oil field exploration and development, in particular to a high-resolution seismic processing method. Background Art

[0002] Seismic data is essential basic information for oilfield exploration and development. Seismic data can be used to interpret the structure of underground oil-bearing strata and reservoir structure information, which is of great significance for improving the success rate of oilfield exploration and expanding development effects.

[0003] Seismic data is obtained through field explosives excitation, geophone reception, and a series of indoor seismic processing. The seismic processing link is crucial, and the seismic processing method directly affects the seismic resolution. The seismic processing methods involved in the current public patents and published articles generally lack high-frequency signals, have a narrow frequency range, and insufficient resolution, resulting in the inability to identify small structures and thin reservoirs based on seismic data, affecting oil and gas exploration and development. Summary of the invention

[0004] In order to overcome the shortcomings of existing seismic processing methods, such as narrow frequency range, insufficient resolution and inability to identify thin reservoirs, the present invention provides a high-resolution seismic processing method, which can obtain broadband and high-resolution seismic data, so that the seismic data can improve the resolution of thin reservoirs with a thickness of more than 3m, thereby improving the exploration and development effects.

[0005] The technical solution of the present invention is: a high-resolution seismic processing method, comprising the following steps:

[0006] Step 1: Obtain seismic data;

[0007] Step 2: Obtain an acoustic wave velocity curve and a density curve by logging, and multiply the acoustic wave velocity curve and the density curve to obtain a longitudinal wave impedance curve;

[0008] Step 3: Convert the P-wave impedance curve obtained in step 2 from the depth domain to the time domain through well seismic calibration to obtain synthetic seismic record data of each well in the study area, and linearly fit the synthetic seismic record data with the measured seismic data near the well to obtain the correlation coefficient;

[0009] Step 4: Based on the time domain P-wave impedance curve obtained in step 3, the P-wave impedance model is obtained by using the well-seismic correlation coefficient constrained inverse distance weighted algorithm. The specific formula is:

[0010]

[0011]

[0012] P(x0, y0, z0) is the longitudinal wave impedance value of the spatial interpolation point, x0, y0, z0 are the spatial coordinates of the interpolation point, i is the well point number, x i ,y i ,z i is the spatial coordinate of the well point with serial number i, n is the total number of wells, P(x i ,y i ,z i ) is the longitudinal wave impedance value of the well point, λ i is the weight coefficient of the well point with sequence number i, d i is the distance between the spatial interpolation point and the well numbered i, μ i is the correlation coefficient between the synthetic record data of the well with sequence number i and the measured seismic data beside the well;

[0013] Step 5: construct a high-frequency seismic wavelet, and perform convolution operation on the high-frequency seismic wavelet and the longitudinal wave impedance model to obtain high-frequency synthetic seismic data;

[0014] Step 6: Perform Fourier transform on the high-frequency synthetic seismic data to generate a spectrum function S(f), and perform bandpass filtering on it;

[0015] At the same time, the measured seismic data is Fourier transformed to generate the spectrum function Y(f), and low-pass filtered;

[0016] Step 7: Perform frequency fusion calculation on the high-frequency synthetic seismic data and the measured seismic data to obtain the spectrum function of the fused seismic, and perform inverse Fourier transform to generate broadband seismic data.

[0017] Further, in step 1, the frequency range of the seismic data is recorded as f1 to f2, where f1 is the low-frequency value of the seismic data and f2 is the high-frequency value;

[0018] In step 5, the frequency range of the high-frequency seismic wavelet is recorded as f3-f4, where f3 is the low-frequency value of the high-frequency seismic wavelet, f4 is the high-frequency value of the high-frequency seismic wavelet, and f4 is 2-3 times of f2.

[0019] Furthermore, in step 6, the high-frequency synthetic seismic data is band-pass filtered to retain the data between f2 and f4. The formula for band-pass filtering is:

[0020] S(f) 高 =S(g)D(g)

[0021] Where D(f) is the bandpass filter function, f is the frequency;

[0022] The high-frequency synthetic seismic data is low-pass filtered to retain data below f2. The formula for low-pass filtering is:

[0023] Y(f) 低 =Y(g)M(g)

[0024] Where M(f) is the low-pass filter function, f is the frequency.

[0025] Furthermore, in step 7, the calculation formula of the spectrum function of the fused earthquake is:

[0026] Z(f)=S(f) 高 +Y(g) 低 .

[0027] Furthermore, in step seven, the frequency range of the broadband seismic data is g1-g4.

[0028] The present invention has the following beneficial effects: by adopting the above scheme, by processing the seismic data, broadband and high-resolution seismic data are obtained, the broadband seismic data has a better correspondence with the lithological interpretation results of the drilling, and the resolution ability is higher, which can improve the resolution ability of thin reservoirs with a thickness of more than 3 meters and improve the exploration and development effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a flow chart of the new high-resolution seismic processing method;

[0030] Figure 2 is the longitudinal wave impedance model profile;

[0031] Figure 3 It is a high-frequency seismic wavelet;

[0032] Figure 4 It is high-frequency synthetic seismic data;

[0033] Figure 5 It is the bandwidth comparison between broadband seismic data and measured seismic data;

[0034] Figure 6 It is a comparison of the resolution capabilities of broadband seismic data and measured seismic data. DETAILED DESCRIPTION

[0035] The present invention will be further described below in conjunction with the accompanying drawings:

[0036] Depend on Figure 1 As shown, a high-resolution seismic processing method comprises the following steps:

[0037] Step 1: Obtain seismic data through field seismic acquisition and indoor seismic processing. The frequency range of seismic data is recorded as g1~g2, where f1 is the low-frequency value of the seismic data and g2 is the high-frequency value.

[0038] Step 2: Obtain the acoustic wave velocity curve and density curve through well logging, and multiply the acoustic wave velocity curve and the density curve to obtain the longitudinal wave impedance curve.

[0039] Step 3: Convert the P-wave impedance curve obtained in step 2 from the depth domain to the time domain through well seismic calibration to obtain synthetic seismic record data for each well in the study area, and linearly fit the synthetic seismic record data with the measured seismic data near the well to obtain the correlation coefficient.

[0040] Step 4: Based on the time domain P-wave impedance curve obtained in step 3, the P-wave impedance model is obtained by using the well-seismic correlation coefficient constrained inverse distance weighted algorithm. The specific formula is:

[0041]

[0042]

[0043] P(x0, y0, z0) is the longitudinal wave impedance value of the spatial interpolation point, x0, y0, z0 are the spatial coordinates of the interpolation point, i is the well point number, x i ,y i ,z i is the spatial coordinate of the well point with serial number i, n is the total number of wells, P(x i ,y i ,z i ) is the longitudinal wave impedance value of the well point, λ i is the weight coefficient of the well point with sequence number i, d i is the distance between the spatial interpolation point and the well numbered i, μ i is the correlation coefficient between the synthetic record data of well number i and the measured seismic data near the well.

[0044] Step 5: construct a high-frequency seismic wavelet. The frequency range of the high-frequency seismic wavelet is recorded as f3~f4, where f3 is the low-frequency value of the high-frequency seismic wavelet, f4 is the high-frequency value of the high-frequency seismic wavelet, and f4 is 2-3 times of f2. The high-frequency seismic wavelet is convolved with the longitudinal wave impedance model to obtain high-frequency synthetic seismic data.

[0045] Step 6: Perform Fourier transform on the high-frequency synthetic seismic data to generate a spectrum function S(f), and perform bandpass filtering on it to retain the data between f2 and f4. The specific formula is:

[0046] S(f) 高 =S(f)D(f)

[0047] Where D(f) is the bandpass filter function, f is the frequency;

[0048] At the same time, the measured seismic data is Fourier transformed to generate the spectrum function Y(f), and low-pass filtered to retain data below f2. The specific formula is:

[0049] Y(f) 低 =Y(f)M(f)

[0050] Where M(f) is the low-pass filter function, f is the frequency.

[0051] Step 7: Perform frequency fusion calculation on the high-frequency synthetic seismic data and the measured seismic data. The specific calculation formula is:

[0052] Z(f)=S(f) 高 +Y(f) 低 ,

[0053] The spectrum function of the fused earthquake is obtained and inverse Fourier transform is performed to generate broadband seismic data. The broadband seismic data has a frequency range of f1 to f4 and has a relatively high resolution.

[0054] Example:

[0055] In a block of Daqing Oilfield, measured seismic data were obtained through field seismic acquisition and indoor seismic processing. The frequency range of measured seismic data is 6Hz to 80Hz. The acoustic velocity curve and density curve were obtained through well logging, and the longitudinal wave impedance curve was obtained by multiplying the acoustic velocity curve and the density curve. The longitudinal wave impedance curve obtained above was converted from the depth domain to the time domain through well seismic calibration, and the correlation coefficient between the synthetic record data of each well in the study area and the measured seismic data beside the well was obtained.

[0056] Based on the time domain P-wave impedance curve obtained above, the P-wave impedance model is obtained by using the well-seismic correlation coefficient constrained inverse distance weighted algorithm. The P-wave impedance model profile is shown in Figure 2 .

[0057] Construct a high-frequency seismic wavelet, see Figure 3 The high-frequency seismic wavelet uses a 0-phase Ricker wavelet, with a main frequency of 100 Hz and a bandwidth of 30 Hz to 200 Hz. The high-frequency seismic wavelet is convoluted with the longitudinal wave impedance model to obtain high-frequency synthetic seismic data, see Figure 4 .

[0058] The high-frequency synthetic seismic data is Fourier transformed to generate the spectrum function S(f), and then band-pass filtered to retain the data between f2 and f4. The specific formula is:

[0059] S(f) 高 =S(f)D(f)

[0060] Where D(f) is the bandpass filter function,

[0061] f is the frequency;

[0062] At the same time, the measured seismic data is Fourier transformed to generate the spectrum function Y(f), and low-pass filtered to retain data below f2. The specific formula is:

[0063] Y(f) 低 =Y(g)M(g)

[0064] Where M(f) is the low-pass filter function, f is the frequency.

[0065] Finally, the high-frequency synthetic seismic data and the measured seismic data are frequency fused and calculated to obtain broadband seismic data. The broadband seismic data has a frequency range of 6 to 200 Hz and has a high resolution, which can improve the resolution of thin reservoirs with a thickness of more than 3 meters.

[0066] Figure 5 and Figure 6 The bandwidth and resolution of broadband seismic data are compared with those of measured seismic data. It can be seen that the frequency range of broadband seismic data is wider; the broadband seismic data has a better correspondence with the lithology (sandstone and mudstone) interpretation results of drilling, and has a higher resolution, and can distinguish thin sand bodies of more than 3m.

Claims

1. A high-resolution seismic processing method, characterized in that The following steps are involved: Step 1: Obtain seismic data; The frequency range of seismic data is recorded as f1~f2, where f1 is the low-frequency value of the seismic data and f2 is the high-frequency value; Step 2: Obtain an acoustic wave velocity curve and a density curve by logging, and multiply the acoustic wave velocity curve and the density curve to obtain a longitudinal wave impedance curve; Step 3: Convert the P-wave impedance curve obtained in step 2 from the depth domain to the time domain through well seismic calibration to obtain synthetic seismic record data of each well in the study area, and linearly fit the synthetic seismic record data with the measured seismic data near the well to obtain the correlation coefficient; Step 4: Based on the time domain P-wave impedance curve obtained in step 3, the P-wave impedance model is obtained by using the well-seismic correlation coefficient constrained inverse distance weighted algorithm. The specific formula is: P(x0, y0, z0) is the longitudinal wave impedance value of the spatial interpolation point, x0, y0, z0 are the spatial coordinates of the interpolation point, i is the well point number, x i ,y i ,z i is the spatial coordinate of the well point with serial number i, n is the total number of wells, P(x i ,y i ,z i ) is the longitudinal wave impedance value of the well point, λ i is the weight coefficient of the well point with serial number i, d i is the distance between the spatial interpolation point and the well numbered i, μ i is the correlation coefficient between the synthetic record data of the well with sequence number i and the measured seismic data beside the well; Step 5: construct a high-frequency seismic wavelet, and perform convolution operation on the high-frequency seismic wavelet and the longitudinal wave impedance model to obtain high-frequency synthetic seismic data; The frequency range of the high-frequency seismic wavelet is recorded as f3-f4, where f3 is the low-frequency value of the high-frequency seismic wavelet, f4 is the high-frequency value of the high-frequency seismic wavelet, and f4 is 2-3 times of f2; Step 6: Perform Fourier transform on the high-frequency synthetic seismic data to generate a spectrum function S(f), and perform bandpass filtering on it; At the same time, the measured seismic data is Fourier transformed to generate the spectrum function Y(f), and low-pass filtered; Step 7: Perform frequency fusion calculation on the high-frequency synthetic seismic data and the measured seismic data to obtain the spectrum function of the fused seismic, and perform inverse Fourier transform to generate broadband seismic data.

2. The high-resolution seismic processing method according to claim 1, characterized in that: In step 6, the high-frequency synthetic seismic data is band-pass filtered to retain the data between f2 and f4. The formula for band-pass filtering is: S(f) 高 =S(f)D(f) Where D(f) is the bandpass filter function, f is the frequency; The high-frequency synthetic seismic data is low-pass filtered to retain data below f2. The formula for low-pass filtering is: Y(f) 低 =Y(f)M(f) Where M(f) is the low-pass filter function, f is the frequency.

3. The high-resolution seismic processing method according to claim 2, characterized in that: In step 7, the calculation formula of the spectrum function of the fused earthquake is: Z(f)=S(f) 高 +Y(f) 低 。 4. The high-resolution seismic processing method according to claim 3, characterized in that: In step 7, the frequency range of broadband seismic data is f1-f4.

Images