Method of low-frequency seismic data enhancement for improving characterization precision of deep carbonate reservoir

Abstract

A method of low-frequency seismic data enhancement for improving the characterization precision of a deep carbonate reservoir includes: first performing inversions on an input seismic data set to obtain the corresponding reflection coefficients and average seismic wavelet; then constructing a seismic wavelet with rich low-frequency information; and finally, performing convolution on the seismic wavelet with rich low-frequency information and the reflection coefficients to obtain seismic data with rich low-frequency information and enhanced low-frequency energy. In the present invention, changes of the seismic data in a work area in transverse and longitudinal directions are taken into consideration, and processing parameters can be quickly determined according to actual conditions of the work area to obtain an optimal processing effect. In this way, the characterization quality of geological anomalies, such as a fault, a fracture system, or the like, in a deep carbonate reservoir can be improved significantly.

Claims

1. A method of a low-frequency seismic data enhancement for improving a characterization precision of a deep carbonate reservoir, comprising the following steps: a) enhancing low-frequency seismic data by: (1) inputting a seismic data set, performing a first inversion using statistical information of seismic data to obtain seismic wavelets, and then performing a second inversion using an optimized sparse regularization method to obtain seismic reflection coefficients; (2) constructing a seismic wavelet w.sub.b with rich low-frequency information by the following sub-steps: (2-1) calculating an average seismic wavelet w.sub.a of the seismic data set based on the seismic wavelets obtained by the first inversion in step (1); (2-2) performing an N-point Fourier transform on the average seismic wavelet w.sub.a, and obtaining a normalized amplitude spectrum S.sub.a of the average seismic wavelet w.sub.a, wherein N represents a number of sampling points contained in a seismic trace; (2-3) determining a reference position n=N/2, and setting a control parameter P.sub.a and a control parameter P.sub.b; (2-4) calculating an amplitude spectrum S.sub.b according to $S_b\left(j\right)=\{\begin{array}{cc}{S}_b\left(j\right)=0.5\left\{1+\cos \left[\pi \left(\frac{2j}{{\mathrm{nP}}_a}-1\right)\right]\right\}{P}_b,& 1\le j<\frac{P_a\left(n-1\right)}{2}\\ S_b\left(j\right)=S_a\left(j\right),& j>h\\ S_b\left(j\right)=P_b,& \mathrm{otherwise}\end{array},$ wherein j is a sampling point number, and h is a sampling point number corresponding to a maximum value in the normalized amplitude spectrum S.sub.a; the control parameter P.sub.a has a value range of $\left[0.01,\frac{2h}{n-1}\right),$ when a value of the control parameter P.sub.a decreases, a number of enhanced low-frequency components increases, and when the value of the control parameter P.sub.a increases, the number of the enhanced low-frequency components decreases; the control parameter P.sub.b has a value range of [1, 3], when a value of the control parameter P.sub.b decreases, the energy of the enhanced low-frequency components decreases, and when the value of the control parameter P.sub.b increases, the energy of the enhanced low-frequency components increases; (2-5) calculating a conversion coefficient C according to $C=S_{a}o\frac{1}{s_b},$ wherein ∘ represents an elementary product operation; (2-6) calculating a temporary seismic wavelet w according to
w=real(ift(ft(w.sub.a)∘C)), wherein ft(□) represents a Fourier transform, ift(□) represents an inverse Fourier transform, and real(□) represents an operation of taking a real part; (2-7) calculating a scale transformation coefficient λ according to $\lambda =\frac{sum\left(S_a\right)}{\mathrm{sum}\left(s_b\right)},$ wherein sum(□) represents a summation operation; and (2-8) calculating the seismic wavelet w.sub.b with the rich low-frequency information according to $w_b=\frac{w}{\lambda };$ (3) convoluting the seismic wavelet w.sub.b with the rich low-frequency information and the seismic reflection coefficients obtained by the second inversion in step (1), to obtain a seismic data set with the rich low-frequency information and enhanced low-frequency energy; and b) generating and displaying a seismic image of the deep carbonate reservoir based on enhanced low-frequency seismic data.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1A-1D show an average seismic wavelet w.sub.a and a corresponding seismic wavelet w.sub.b with rich low-frequency information of seismic data in a certain work area according to an embodiment of the present invention. FIG. 1A shows the average seismic wavelet w.sub.a of the input seismic data set, wherein the ordinate represents amplitude and the abscissa represents time in milliseconds (ms). FIG. 1B shows an amplitude spectrum of the average seismic wavelet w.sub.a, wherein the ordinate represents amplitude and the abscissa represents frequency in hertz (Hz). FIG. 1C shows the seismic wavelet w.sub.b with rich low-frequency information, wherein the ordinate represents amplitude and the abscissa represents time in milliseconds (ms). FIG. 1D shows an amplitude spectrum of the seismic wavelet w.sub.b with rich low-frequency information, wherein the ordinate represents amplitude and the abscissa represents frequency in hertz (Hz).

(2) FIGS. 2A-2B show a comparison of seismic sections before and after low-frequency enhancement is performed on the seismic data set of the certain work area using the seismic wavelet w.sub.b with rich low-frequency information in FIG. 1C according to an embodiment of the present invention. FIG. 2A shows an original seismic section before the low-frequency-information enhancement, wherein the abscissa represents a trace number and the ordinate represents time in seconds-(s). FIG. 2B shows a seismic section after performing the low-frequency-information enhancement using the method according to the present invention, wherein the abscissa represents a trace number and the ordinate represents time in seconds-(s).

(3) FIGS. 3A-3B show a comparison of time slices extracted from variance attributes of the seismic data set calculated before and after the low-frequency-information enhancement in the certain work area according to an embodiment of the present invention, which illustrates the difference in imaging quality of deep geological abnormal information in the seismic data before and after performing the enhancement using the method according to the present invention. FIG. 3A shows the time slice of the variance attribute of the original seismic data at 4,452 ms before the low-frequency-information enhancement, wherein the abscissa represents an InLine line number and the ordinate represents an XLine line number. FIG. 3B shows the time slice of the variance attribute of the seismic data at 4,452 ms after performing the low-frequency-information enhancement using the method according to the present invention, wherein the abscissa represents an InLine line number and the ordinate represents an XLine line number.

(4) FIG. 4 shows an exemplary method of low-frequency seismic data enhancement for effectively improving the characterization precision of a deep carbonate reservoir.

DETAILED DESCRIPTION OF THE EMBODIMENTS

(5) As shown in FIG. 4, and according to an embodiment of the present invention, a method of low-frequency seismic data enhancement for effectively improving the characterization precision of a deep carbonate reservoir includes the following steps:

(6) (1) a seismic data set is input, and in this embodiment, first an inversion is performed using statistical information of seismic data to obtain the seismic wavelet of each seismic trace in the input data set, and then an inversion is performed using an optimized sparse regularization method to obtain the reflection coefficient of each seismic trace in the input data set;

(7) (2) a seismic wavelet w.sub.b with rich low-frequency information is constructed by the following sub-steps:

(8) 2-1 an average seismic wavelet w.sub.a of the input seismic data set is calculated based on the seismic wavelets obtained by the inversion in step (1);

(9) 2-2 an N-point Fourier transform is performed on the average seismic wavelet w.sub.a, and a normalized amplitude spectrum S.sub.a of the average seismic wavelet w.sub.a is obtained, wherein N represents the number of sampling points contained in a seismic trace, and in this embodiment, N=1750;

(10) 2-3 a reference position n=N/2 is determined, and a control parameter P.sub.a=0.07 and a control parameter P.sub.b=1.5 are set;

(11) 2-4 a new amplitude spectrum S.sub.b is calculated according to the following formula:

(12) $S_b\left(j\right)=\{\begin{array}{cc}{S}_b\left(j\right)=0.5\left\{1+\cos \left[\pi \left(\frac{2j}{{\mathrm{nP}}_a}-1\right)\right]\right\}{P}_b,& 1\le j<\frac{P_a\left(n-1\right)}{2}\\ S_b\left(j\right)=S_a\left(j\right),& j>h\\ S_b\left(j\right)=P_b,& \mathrm{otherwise}\end{array},$

(13) wherein j is a sampling point number, and h is a sampling point number corresponding to the maximum value in the normalized amplitude spectrum S.sub.a;

(14) 2-5 a conversion coefficient C is calculated according to the following formula:

(15) $C=S_{a}o\frac{1}{s_b},$

(16) wherein ∘ represents an elementary product operation;

(17) 2-6 a temporary seismic wavelet w is calculated according to the following formula:
w=real(ift(ft(w.sub.a)∘C)),

(18) wherein ft(g) represents Fourier transform, ift(g) represents inverse Fourier transform, and real(g) represents the real part of complex number;

(19) 2-7 a scale transformation coefficient λ is calculated according to the following formula:

(20) $\lambda =\frac{sum\left(S_a\right)}{\mathrm{sum}\left(s_b\right)},$

(21) wherein sum(g) represents a summation operation; and

(22) 2-8 the seismic wavelet w.sub.b with rich low-frequency information is calculated according to the following formula

(23) $w_b=\frac{w}{\lambda };$

(24) (3) a convolution operation is performed on the seismic wavelet w.sub.b with rich low-frequency information and the reflection coefficients obtained by the inversion in step (1), so as to finally obtain a seismic data set with rich low-frequency information and enhanced low-frequency energy.

(25) FIG. 1C and FIG. 1D show a seismic wavelet w.sub.b with rich low-frequency information and a normalized amplitude spectrum thereof corresponding to control parameters P.sub.a=0.07 and P.sub.b=1.5 based on the average seismic wavelet w.sub.a of the seismic data set of a certain work area according to an embodiment of the present invention.

(26) FIG. 2A and FIG. 2B show a comparison of seismic sections before and after low-frequency compensation is performed on the seismic data set of the certain work area using the seismic wavelet w.sub.b with rich low-frequency information in FIG. 1C according to an embodiment of the present invention, which illustrates that the imaging quality of a deep carbonate fracture system in FIG. 2B after low-frequency-information enhancement is better than the imaging quality of the original seismic section in FIG. 2A (for example, the areas outlined by the white dashed rectangles in FIG. 2A and FIG. 2B).

(27) FIG. 3A and FIG. 3B show a comparison of time slices at 4,452 ms separately extracted from a variance-attribute data set of an original seismic data set in a certain work area and a variance-attribute data set of a seismic data set with enhanced low-frequency information according to an embodiment of the present invention, wherein the seismic data set with enhanced low-frequency information is obtained by performing low-frequency-information enhancement on the original seismic data set using the method of the present invention, and the variance-attribute data set of the original seismic data set and the variance-attribute data set of the seismic data set with enhanced low-frequency information are obtained by respectively calculating the variance attributes of the original seismic data set and the variance attributes of the seismic data set with enhanced low-frequency information. A comparison between FIGS. 3B and 3A indicate that, after performing the low-frequency-information enhancement using the method according to the present invention, the time slice of the variance attribute in FIG. 3B shows more and richer geological abnormal information more clearly.

(28) The present invention has the following advantages: (1) the seismic wavelet with rich low-frequency information is constructed by using the average seismic wavelet of the whole input seismic data set, and changes of the seismic data in the work area in transverse and longitudinal directions are taken into consideration; (2) only 2 control parameters are involved, wherein the number of enhanced low-frequency components can be controlled by the parameter P.sub.a, while the energy of the enhanced low-frequency components can be controlled by the parameter P.sub.b. In a practical application in seismic data processing, the processing parameters can be quickly determined according to actual conditions of the data of the work area, so as to obtain the optimal processing effect.

(29) The above embodiments are only used for illustrating the present invention, the implementation steps of the method, or the like, may be changed, and such equivalent changes and modifications based on the technical solution of the present invention shall fall within the scope of protection of the present invention.