CN103675911B - A kind of method based on compressional wave and converted shear wave joint inversion intercept and gradient - Google Patents

Abstract

The present invention relates to a kind of method based on compressional wave and converted shear wave joint inversion intercept and gradient.The method comprises the following steps: obtain compressional wave, converted shear wave angle domain road collection, obtains compressional wave and shear wave velocity and result of log interpretation simultaneously; Wavelet is extracted to compressional wave and converted shear wave angle domain road collection and obtains angle domain reflection coefficient section, and utilize the compressional wave that obtains and shear wave velocity to ask for transverse and longitudinal wave velocity to compare section; Utilize derive with intercept and gradient be parameter converted shear wave reflection coefficient formula and longitudinal wave reflection coefficient approximate expression, adopt singular value decomposition method to carry out joint inversion, obtain intercept and gradient; Utilize result of log interpretation to carry out demarcation to the intercept be finally inversed by and gradient and set up fluid identification pattern, fluid detection is carried out to non-drilling area.The present invention makes full use of the Multi-component earthquake wave field information received, and decreases the multi-solution of refutation process, improves the stability of refutation process, for petroleum-gas prediction provides a kind of method reliably.

Description

Intercept and gradient joint inversion method based on longitudinal wave and converted shear wave

Technical Field

The invention belongs to the field of geophysical exploration, and particularly relates to a method for jointly inverting intercept and gradient based on longitudinal waves and converted shear waves.

Background

Amplitude variation with offset (AmplifiedeuVersusOffet) is a technology for obtaining lithological parameters such as speed, density and the like by researching the variation relation of reflection amplitude with offset so as to predict oil and gas reservoirs. Oil and gas identification and fluid prediction can be effectively carried out by extracting corresponding intercept (P) and gradient (G) or combination parameters of the intercept (P) and the gradient (G) by using Shuey (1985) formula. The intercept (P) section, namely the vertical incidence reflection coefficient section, has higher signal-to-noise ratio and resolution, is vertically incident and reflected at the position close to zero offset, does not generate converted transverse waves, and can be regarded as a longitudinal wave section. The gradient (G) section is related to Poisson's ratio, contains information of lithologic changes of strata above and below a reflecting interface, and intercept (P) and gradient (G) are good oil and gas indication and fluid prediction parameters.

However, the traditional method for extracting the intercept (P) and the gradient (G) only depends on longitudinal wave data, and is difficult to overcome the instability in the extraction process and the multi-solution of the extraction result; on the other hand, the multi-component seismic data records abundant seismic wave field information, so that the defect of a method of simply utilizing longitudinal waves can be overcome, the multi-solution property of the inversion process is reduced, the stability of the inversion result is improved, and a reliable method is provided for oil and gas prediction. The invention comprehensively utilizes the combination of longitudinal waves and converted shear waves to extract the intercept (P) and the gradient (G), fully utilizes the received multi-component seismic wave field information, reduces the multi-solution property of the inversion process, improves the stability of the inversion process and provides a reliable method for oil-gas prediction.

Disclosure of Invention

In order to overcome the defect that intercept (P) and gradient (G) parameters are inverted by using compressional wave data alone, the invention provides a method for jointly inverting the intercept (P) and the gradient (G) based on compressional wave and converted shear wave; the method is characterized in that longitudinal wave and converted shear wave field information is fully utilized, on the basis of a longitudinal wave reflection coefficient approximate expression (Shuey, 1985) with intercept (P) and gradient (G) as parameters, a converted shear wave reflection coefficient approximate expression with the intercept (P) and the gradient (G) as parameters is further deduced, and the intercept (P) and the gradient (G) are obtained by performing joint inversion by using the obtained longitudinal wave and converted shear wave angle gathers and adopting a singular value decomposition method.

In order to achieve the purpose, the technical scheme of the invention is as follows:

a method for jointly inverting intercept and gradient based on compressional waves and converted shear waves comprises the following steps:

step 1: obtaining the angle domain gathers of the longitudinal wave and the converted transverse wave, and simultaneously obtaining the velocity of the longitudinal wave and the transverse wave and the well logging interpretation result

Step 2: extracting sub-waves from the gathers of longitudinal wave and converted transverse wave angle domain to obtain angle domain reflection coefficient profile, and obtaining transverse-longitudinal wave velocity ratio profile by using the obtained longitudinal wave and transverse wave velocities

And step 3: by using the deduced converted transverse wave reflection coefficient formula and longitudinal wave reflection coefficient approximate formula which take the intercept (P) and the gradient (G) as parameters, the singular value decomposition method is adopted to carry out joint inversion to obtain the intercept (P) and the gradient (G)

And 4, step 4: calibrating inverted intercept (P) and gradient (G) by using well logging interpretation results to establish a fluid identification mode, and carrying out fluid detection on an undrilled area

Compared with the prior art, the invention has the following beneficial effects: the invention comprehensively utilizes the combination of longitudinal waves and converted shear waves to extract the intercept (P) and the gradient (G), fully utilizes the received multi-component seismic wave field information, reduces the multi-solution property of the inversion process, improves the stability of the inversion process and provides a reliable method for oil-gas prediction.

Drawings

FIG. 1 is a flow chart of a method for extracting intercept (P) and gradient (G) by combining compressional waves and converted shear waves;

FIG. 2 is a graph of layered model data for combined extraction of intercept (P) and gradient (G) for compressional and converted shear waves;

FIG. 3a is a compressional angle domain gather (0-38) synthesized directly from the Zoeppritz equation;

FIG. 3b is a converted shear wave angle domain gather (0-38) synthesized directly from the Zoeppritz equation;

FIG. 4a is a longitudinal wave angle domain reflection coefficient obtained by removing the wavelets from the longitudinal wave angle domain gather of FIG. 3a using least squares pulse deconvolution principles;

FIG. 4b is a converted shear wave angle domain reflection coefficient obtained by removing the wavelets of the converted shear wave angle domain gather of FIG. 3b using the least squares pulse deconvolution principle;

FIG. 5a is a comparison of intercept (P) extracted jointly by compressional and converted shear and intercept (P) extracted with compressional alone with the true value of intercept (P);

FIG. 5b is a comparison of the gradient (G) extracted using compressional and converted shear waves in combination with the gradient (G) extracted using compressional waves only with the true value of the gradient (G);

FIG. 6a is a compressional angle domain gather (0-38) synthesized directly from the Zoeppritz equation with 25dB background noise added;

FIG. 6b is a converted shear wave angle domain gather (0-38) synthesized directly from the Zoeppritz equation with the addition of 25dB background noise;

FIG. 7a is a longitudinal wave angle domain reflection coefficient obtained by removing the wavelets from the longitudinal wave angle domain gather of FIG. 6a using least squares pulse deconvolution principles;

FIG. 7b is a converted shear wave angle domain reflection coefficient obtained by removing the wavelets of the converted shear wave angle domain gather of FIG. 6b using the least squares pulse deconvolution principle;

FIG. 8a is a comparison of intercept (P) extracted jointly by compressional and converted shear waves and intercept (P) extracted with compressional waves alone with the true value of intercept (P) in the case of random noise added to the input angle gather;

FIG. 8b compares the gradient (G) extracted with the compressional and converted shear waves in combination with the true value of the gradient (G) extracted with the compressional waves alone, with the addition of random noise in the input angle gathers.

Detailed Description

As shown in fig. 1, the method for jointly inverting intercept (P) and gradient (G) based on compressional and converted shear waves comprises the following steps:

(1) acquiring longitudinal wave and converted transverse wave angle domain gathers, and simultaneously acquiring longitudinal wave and transverse wave speeds and well logging interpretation results; the specific method comprises the following steps:

extracting pre-stack seismic angle gather data by using a seismic imaging method based on seismic data acquired in the field;

acquiring longitudinal wave and transverse wave speeds and well logging interpretation results based on well logging data of field well logging;

(2) extracting sub-waves from the longitudinal wave and converted transverse wave angle domain gathers to obtain an angle domain reflection coefficient profile, and solving a transverse-longitudinal wave velocity ratio profile by using the obtained longitudinal wave and transverse wave velocities; the specific method comprises the following steps:

the method comprises the steps of obtaining wavelets in longitudinal wave and converted transverse wave angle channel sets by using a prestack seismic geological calibration method (strong cautious nations, prestack seismic inversion method and influence factor research [ J ].2013), and then removing the wavelets in the longitudinal wave and converted transverse wave angle domain channel sets by using a least square pulse deconvolution principle to obtain reflection coefficients in the longitudinal wave and converted transverse wave angle domains.

By using a Zoeppritz equation (see the seismic wave theory and method, Sunjun, equation 3-2-7 on page 65), by using the logging longitudinal wave velocity, the transverse wave velocity and the density parameters, the forward longitudinal wave and converted transverse wave well-side angle domain seismic channels can be obtained, and the wavelet in the actually obtained angle channel set can be estimated by comparing the longitudinal wave and converted transverse wave angle channel set with the actually obtained prestack longitudinal wave and converted transverse wave angle channel set. Then, under the known condition of the wavelets, a least square pulse deconvolution method (Leshuchun's seismic data processing method, page 67) is adopted to remove the wavelets of the gathers in the angle domain of the longitudinal waves and the converted transverse waves, and the reflection coefficients in the angle domain of the longitudinal waves and the converted transverse waves are obtained.

And dividing the transverse wave velocity and the longitudinal wave velocity obtained by the logging data to obtain a transverse-longitudinal wave velocity ratio profile.

(3) Performing joint inversion by using a derived conversion transverse wave reflection coefficient formula (formula 1.5) and a derived longitudinal wave reflection coefficient approximate formula (formula 1.1) of Shuey (1985) with intercept (P) and gradient (G) as parameters and adopting a singular value decomposition method to obtain the intercept (P) and the gradient (G); the specific method comprises the following steps:

shuey (1985) rewrites the longitudinal wave reflection coefficient approximation derived from Aki and Richard (1980) into a formula with intercept (P) and gradient (G) as parameters:

$R_{\mathrm{pp}}\left(\mathrm{\θ}\right)=P+G{\sin }^2\mathrm{\θ}+\frac{1}{2}\frac{\mathrm{\Δ}{V}_p}{V_p}({\tan }^2\mathrm{\θ}-{\sin }^2\mathrm{\θ})---\left(1.1\right)$

wherein, $P=\frac{1}{2}(\frac{\mathrm{\Δ}{v}_p}{v_p}+\frac{\mathrm{\Δ\ρ}}{\mathrm{\ρ}}),$ $G=\frac{1}{2}\frac{\mathrm{\Δ}{v}_p}{v_p}-2(2\frac{\mathrm{\Δ}{v}_p}{v_p}+\frac{\mathrm{\Δ\ρ}}{\mathrm{\ρ}})+\frac{\mathrm{\Δ\σ}}{{(1-\mathrm{\σ})}^2};$ V_p、ΔV_pthe average velocity of longitudinal waves of the upper and lower layers and the difference between the velocities of longitudinal waves of the upper and lower layers are provided, and theta is the average of the incident angle and the transmission angle of the longitudinal waves.

The intercept (P) and the gradient (G) are taken as parameters to derive an approximate formula (formula 1.5) of the reflection coefficient of the converted shear wave. Aki and Richard et al (1980) give an approximation of the converted shear wave reflection coefficient:

wherein theta,Respectively, the average value of the incident angle and the transmission angle of the longitudinal wave, the average value of the reflection angle and the transmission angle of the converted wave, V_p、V_sRho is the average value of longitudinal wave velocity, transverse wave velocity and density of the upper and lower layers, respectively, and DeltaV_p、ΔV_sAnd Δ ρ is the difference between the longitudinal wave velocity, the transverse wave velocity and the density of the upper and lower layers.

In isotropic media, the relationship between poisson's ratio and the velocity ratio of longitudinal and transverse waves:

$\frac{V_s^2}{V_p^2}=\frac{1-2\mathrm{\σ}}{2(1-\mathrm{\σ})}---\left(1.3\right)$

wherein, V_p、V_sAnd sigma is the average value of longitudinal wave velocity, transverse wave velocity and Poisson's ratio of the upper and lower layers, and is V_p、V_sAs a function of (c). Differentiating two sides of the formula 1.2 and finishing to obtain:

$\frac{\mathrm{\Δ}{V}_s}{V_s}=\frac{\mathrm{\Δ}{V}_p}{V_p}-\frac{1}{4}\frac{V_p^2}{V_s^2}\frac{\mathrm{\Δ\σ}}{{(1-\mathrm{\σ})}^2}---\left(1.4\right)$

substituting formula 1.4 into formula 1.2 and finishing to obtain the following formula:

in order to overcome the defect that parameters of intercept (P) and gradient (G) are inverted by using longitudinal wave data alone, the information of longitudinal wave and converted shear wave fields is fully utilized, a formula 1.1 and a formula 1.5 are combined, and an SVD (singular value decomposition) method is used for jointly inverting the intercept and the gradient:

$\left[\begin{array}{c}{R}_{\mathrm{pp}}\left({\mathrm{\θ}}_1\right)\\ .\\ .\\ .\\ R_{\mathrm{pp}}\left({\mathrm{\θ}}_n\right)\\ R_{\mathrm{ps}}\left({\mathrm{\θ}}_1\right)\\ .\\ .\\ .\\ R_{\mathrm{ps}}\left({\mathrm{\θ}}_n\right)\end{array}\right]\left[\begin{array}{ccc}1& B\left({\mathrm{\θ}}_1\right)& C\left({\mathrm{\θ}}_1\right)\\ .& .& .\\ .& .& .\\ .& .& .\\ 1& B\left({\mathrm{\θ}}_n\right)& C\left({\mathrm{\θ}}_n\right)\\ D\left({\mathrm{\θ}}_1\right)& E\left({\mathrm{\θ}}_1\right)& F\left({\mathrm{\θ}}_1\right)\\ .& .& .\\ .& .& .\\ .& .& .\\ D\left({\mathrm{\θ}}_n\right)& E\left({\mathrm{\θ}}_n\right)& F\left({\mathrm{\θ}}_n\right)\end{array}\right]\left[\begin{array}{c}\\ P\\ G\\ \frac{\mathrm{\Δ}{V}_p}{V_p}\end{array}\right]$

wherein, B (theta)_i)，C(θ_i) (i 1, 2, 3.. n) are different angles θ, respectively_iN) (i ═ 1, 2, 3.. n) Rpp equation G, Δ V_p/V_pCoefficient of front, D (theta)_i)、E(θ_i)、F(θ_i) (i ═ 1, 2, 3.. n) at different angles θ_iN) (i 1, 2, 3.. n) Rps equation P, G, Δ V_p/V_pCoefficient of front, Rpp (θ)_i)、Rps(θ_i) (i 1, 2, 3.. n) are different angles θ_i（i＝1，2，3......n) longitudinal and converted shear reflection coefficients.

The above expression for the inversion intercept and gradient can be summarized as:

y＝Ax

wherein y is an angle domain quasi-reflection coefficient matrix, x is an attribute parameter matrix to be extracted, and A is an attribute parameter x front coefficient matrix.

The attribute parameter x matrix can be obtained by AVO inversion using singular value method, i.e.

x＝UΛ^-1V^Ty

U, V is AA obtained by singular value decomposition of matrix A^TΛ is the AA obtained by singular value decomposition of the matrix A^TA matrix of singular values.

(4) Calibrating the inverted intercept (P) and gradient (G) by using the well logging interpretation result to establish a fluid identification mode, and carrying out fluid detection on the non-drilled area; the specific method comprises the following steps:

firstly, calibrating an inverted intercept (P) and gradient (G) section and a logging interpretation result of a well passing through the section, namely correspondingly calibrating an oil-gas layer and a water layer on the intercept (P) and gradient (G) section on the inverted intercept (P) and gradient (G) section according to the logging interpretation result, and predicting an oil layer and a gas layer on the intercept and gradient section of an undrilled well area according to the display phenomena of the oil-gas layer and the water layer on the intercept (P) and gradient (G) section to obtain a favorable well area.

By testing the four-layer model proposed by mahmoudianandsmograve (2004), fig. 5 is a comparison graph of combined compressional and converted shear extraction intercept (P) and gradient (G) properties and the use of a single compressional extraction intercept (P) and gradient (G) properties with the true values of intercept (P) and gradient (G), the comparison finding that the jointly inverted intercept (P) and gradient (G) have higher accuracy, particularly the gradient (G) properties; FIG. 8 is a plot of the combined compressional and converted shear extracted intercept (P) and gradient (G) properties and the use of a single compressional to extract the contrast of the intercept (P) and gradient (G) properties with the true values of the intercept (P) and gradient (G) in the presence of noise in the input angle gather data, showing that the contrast results in a higher accuracy of the jointly inverted intercept (P) and gradient (G), particularly the gradient (G) properties, and that the combination has some suppression of background noise.

Claims (4)

1. A method for jointly inverting intercept and gradient based on compressional waves and converted shear waves is characterized by comprising the following steps:

step 1: acquiring a longitudinal wave angle domain gather, a converted transverse wave angle domain gather, and simultaneously acquiring longitudinal wave and transverse wave speeds and well logging interpretation results;

step 2: extracting sub-waves from the longitudinal wave and converted shear wave angle domain gathers to obtain an angle domain reflection coefficient profile, and solving a shear-longitudinal wave velocity ratio profile by using the obtained longitudinal wave and shear wave velocities;

and step 3: performing joint inversion by using a converted shear wave reflection coefficient formula and a longitudinal wave reflection coefficient approximate formula which take intercept P and gradient G as parameters and adopting a singular value decomposition method to obtain the intercept P and the gradient G; the specific method comprises the following steps:

approximate formula of longitudinal wave reflection coefficient with intercept P and gradient G as parameters:

$R_{pp}\left(\mathrm{\θ}\right)=P+G{\sin }^2\mathrm{\θ}+\frac{1}{2}\frac{{\mathrm{\ΔV}}_p}{V_p}({\tan }^2\mathrm{\θ}-{\sin }^2\mathrm{\θ})---\left(1.1\right)$

wherein, $P=\frac{1}{2}(\frac{{\mathrm{\ΔV}}_p}{V_p}+\frac{\mathrm{\Δ}\mathrm{\ρ}}{\mathrm{\ρ}}),G=\frac{1}{2}\frac{{\mathrm{\ΔV}}_p}{V_p}-2(2\frac{{\mathrm{\ΔV}}_p}{V_p}+\frac{\mathrm{\Δ}\mathrm{\ρ}}{\mathrm{\ρ}})+\frac{\mathrm{\Δ}\mathrm{\σ}}{{(1-\mathrm{\σ})}^2};$

V_prho and sigma are respectively the average values of the longitudinal wave speed, the density and the Poisson ratio of the upper layer and the lower layer; Δ V_pAnd delta rho and delta sigma are respectively the longitudinal wave velocity difference, the density difference and the Poisson's ratio difference of the upper layer and the lower layer; theta is the average of the incident angle and the transmission angle of the longitudinal wave;

approximate formula of reflection coefficient of converted shear wave with intercept P and gradient G as parameters:

wherein theta,Respectively, the average value of the incident angle and the transmission angle of the longitudinal wave, the average value of the reflection angle and the transmission angle of the converted wave, V_p、V_sRho is the average value of the longitudinal wave velocity, the transverse wave velocity and the density of the upper layer and the lower layer respectively,ΔV_p、ΔV_sand the delta rho is the longitudinal wave velocity difference, the transverse wave velocity difference and the density difference of the upper layer and the lower layer respectively;

by utilizing a longitudinal wave and converted shear wave angle domain reflection coefficient profile obtained after the extraction of the wavelet and a calculated shear-longitudinal wave velocity ratio profile, the following system equation is solved by a Singular Value Decomposition (SVD) method to carry out joint inversion, and intercept P and gradient G are obtained:

$\left[\begin{array}{c}{R}_{pp}\left({\mathrm{\θ}}_1\right)\\ \begin{array}{c}.\\ .\\ .\end{array}\\ R_{pp}\left({\mathrm{\θ}}_n\right)\\ R_{ps}\left({\mathrm{\θ}}_1\right)\\ \begin{array}{c}.\\ .\\ .\end{array}\\ R_{ps}\left({\mathrm{\θ}}_n\right)\end{array}\right]=\left[\begin{array}{ccc}1& B\left({\mathrm{\θ}}_1\right)& C\left({\mathrm{\θ}}_1\right)\\ \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}\\ 1& B\left({\mathrm{\θ}}_n\right)& C\left({\mathrm{\θ}}_n\right)\\ D\left({\mathrm{\θ}}_1\right)& E\left({\mathrm{\θ}}_1\right)& F\left({\mathrm{\θ}}_1\right)\\ \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}& \begin{array}{c}.\\ .\\ .\end{array}\\ D\left({\mathrm{\θ}}_n\right)& E\left({\mathrm{\θ}}_n\right)& F\left({\mathrm{\θ}}_n\right)\end{array}\right]\left[\begin{array}{c}P\\ G\\ \frac{{\mathrm{\ΔV}}_p}{V_p}\end{array}\right]$

wherein, B (theta)_i)，C(θ_i) N, each of which is a different angle θ_i，i＝1，2，3......n，R_ppG in the equation,A coefficient of front; d (theta)_i)、E(θ_i)、F(θ_i) N, i being different angles θ, 1, 2, 3_i，i＝1，2，3......n，R_psP, G in the equation,A coefficient of front; r_pp(θ_i)、R_ps(θ_i) N, i being different angles θ, 1, 2, 3_iN, i 1, 2, 3.. n, and converted shear wave reflection coefficients;

and 4, step 4: and calibrating the inverted intercept P and the gradient G by using the well logging interpretation result to establish a fluid identification mode, and carrying out fluid detection on the non-drilled area.

2. The method for jointly inverting the intercept and the gradient based on the compressional wave and the converted shear wave according to claim 1, wherein the specific method in the step 1 is as follows:

extracting pre-stack seismic angle gather data by using a seismic imaging method based on seismic data acquired in the field;

and obtaining longitudinal wave and transverse wave speeds and well logging interpretation results based on the well logging data of the field well logging.

3. The method for jointly inverting the intercept and the gradient based on the compressional wave and the converted shear wave according to claim 2, wherein the specific method in the step 2 is as follows:

obtaining wavelets in the longitudinal wave and converted transverse wave angle domain channel sets by using a prestack seismic geological calibration method, and then removing the wavelets in the longitudinal wave and converted transverse wave angle domain channel sets by using a least square pulse deconvolution method to obtain reflection coefficients in the longitudinal wave and converted transverse wave angle domains;

and dividing the transverse wave velocity and the longitudinal wave velocity obtained by the logging data to obtain a transverse-longitudinal wave velocity ratio profile.

4. The method for jointly inverting the intercept and the gradient based on the compressional wave and the converted shear wave according to claim 3, wherein the specific method in the step 4 is as follows:

firstly, calibrating inverted intercept P and gradient G sections and well logging interpretation results of wells passing through the sections, namely correspondingly calibrating oil-gas layers and water layers on the intercept P and gradient G sections on the inverted intercept P and gradient G sections according to the well logging interpretation results, and predicting oil layers and gas layers on the intercept and gradient sections of an undrilled well area according to display phenomena of the oil-gas layers and the water layers on the intercept P and gradient G sections to obtain a favorable well area.