Reservoir physical property detection method and device, electronic equipment and storage medium

Abstract

The invention provides a reservoir physical property detection method and device, electronic equipment and a storage medium, and belongs to the technical field of oil exploration and development. The method comprises the following steps: acquiring real-time longitudinal wave azimuth AVO gather data of a well area; the longitudinal wave azimuth AVO gather data are input into the constructed AVO inversion model, reservoir physical property parameters of the target well area are obtained through inversion, the reservoir physical property parameters comprise transverse wave velocity reflectivity, longitudinal wave velocity reflectivity and density reflectivity, and limiting conditions and weight coefficients are added into an inversion target function of the AVO inversion model. According to the method, starting from the most basic mathematical level, limiting conditions and design weight coefficients are added to an original objective function of least square, and a coefficient matrix is improved, so that the sensitivity of a solution of the inversion objective function to errors is reduced, the influence of data errors on a result is determined, noise interference is dealt with, and the accuracy of the result is improved. And the sensitivity of the solution of the inversion objective function to errors and noise is greatly reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of oil exploration and development, especially the technical field of AVO prestack inversion, and specifically relates to a method, device, electronic device and storage medium for detecting reservoir physical properties. Background Art

[0002] The main problem faced in oilfield exploration from beginning to end is to determine the oil and gas content and the amount of oil and gas in the exploration target layer. Based on these indicators, decisions can be made on the investment in subsequent detailed exploration, the number of wells drilled, the scale of oilfield construction, and the laying of oil pipelines, etc., which is a plan for several years or even longer. The difficulties in judging the oil and gas content and calculating the amount of oil and gas lie in that we can only obtain one or at most two reservoir parameters from seismic data.

[0003] Existing research has shown that acoustic impedance is related to lithology, porosity, fluid composition and saturation. However, theoretically, it is impossible to judge the oil and gas content only by acoustic impedance, let alone calculate the amount of oil and gas. Only guessed and estimated values can be obtained. Further, at present, a second parameter, shear wave impedance, has been introduced, which reduces the uncertainty of the oil and gas content and the amount of oil and gas, but the judgment and estimation are still inaccurate, seriously affecting the oilfield exploration and development work.

[0004] At present, the prestack inversion modules provided by mainstream software on the market, such as prestack inversion software like Hampton-Russell and Jason, are already backward because they still only calculate two parameters, and the calculation results of the second parameter are still not very accurate, unable to meet the requirements of oil and gas exploration. In addition, there are also other characteristic software on the market. These characteristic software often have strong locality, that is, they can only be applied to a certain area, or certain strata, or certain lithology combinations, and do not have universality. In some cases, the results of prestack inversion will mislead the interpreters' interpretation of reservoir physical properties because they consider problems from the perspective of rock physics.

[0005] To improve the accuracy of prestack inversion, those skilled in the art have made more and more research and practice on AVO prestack inversion and have also achieved effective research results. Among them, prestack seismic inversion for inverting the P-wave to S-wave velocity ratio, etc. has emerged and has become one of the important geophysical technologies in oil exploration and development. The P-wave to S-wave velocity ratio containing shear wave information often has stronger lithology discrimination and fluid identification capabilities than impedance. It is a commonly used parameter for lithology characterization and fluid identification. The density parameter is an effective parameter for describing reservoir physical properties. In the conventional identification of oil and gas layers by logging, density and resistivity are important discriminant bases for distinguishing oil and water layers and are key parameters for predicting the lateral distribution law of reservoirs and fluids, and have important reference significance in reservoir physical property prediction.

[0006] In the existing pre-stack seismic inversion technology, simultaneous inversion of three pre-stack parameters is one of the most commonly used technologies. This technology effectively utilizes the AVO information contained in pre-stack seismic data, and simultaneously inverts three parameters through multiple partial stacking data volumes, providing elastic parameters or parameter combinations that are more effective for lithology and fluid identification. However, there are still some problems in simultaneously inverting three parameters. For example, the AVO inversion problem is highly ill-conditioned and requires reasonable regularization. The value of the regularization coefficient is usually an estimated value, resulting in a certain error in the inversion results of the three pre-stack parameters. The ill-conditioned equation is also highly sensitive to errors, leading to poor accuracy in the inversion of the three pre-stack parameters. At the same time, noise also has a great impact on the inversion results of this technology, seriously affecting the acquisition of reservoir physical property parameters, and thus unable to accurately track and interpret the reservoir physical properties.

[0007] In summary, how to reduce the ill-conditioned degree of the inversion problem and improve the inversion accuracy of the three pre-stack parameters is a technical difficulty that those skilled in the art need to overcome urgently. Summary of the Invention

[0008] The purpose of the embodiments of the present invention is to provide a reservoir physical property detection method, device, electronic device and storage medium, so as to solve the technical problems in the prior art that the inversion error of the three pre-stack parameters is relatively large due to the use of an estimated regularization coefficient during regularization, and the three pre-stack parameter inversion is highly sensitive to data errors and noise, resulting in poor accuracy in solving the three parameters.

[0009] To achieve the above purpose, the first aspect of the embodiments of the present invention provides a reservoir physical property detection method, and the method includes:

[0010] Obtain real-time longitudinal wave azimuth AVO gather data in the well area;

[0011] Input the longitudinal wave azimuth AVO gather data into the constructed AVO inversion model, and invert the reservoir physical property parameters in the well area. The reservoir physical property parameters include shear wave velocity reflectivity, longitudinal wave velocity reflectivity, and density reflectivity;

[0012] When constructing the AVO inversion model, historical longitudinal wave azimuth AVO gather data is used as the data source;

[0013] The determination process of the inversion objective function of the AVO inversion model is as follows: Based on the linear equations constructed by multi-channel reflection coefficients, determine the objective function S=(AX - b) T (AX - b) of the least squares inversion; add a constraint condition to the objective function, and the objective function is modified to S=(AX - b) T (AX - b)+λX T X, where the magnitude of the first parameter λ cannot be estimated in advance; for the first term (AX - b) of the objective function T(AX - b) is added with weight coefficient P and the second term X T X is added with weight coefficient Q, and the objective function is modified to S = (AX - b) T P(AX - b) + X T QX;

[0014] Correspondingly, the solution X of the AVO inversion model is X = (A T PA + Q) -1 (A T Pb);

[0015] Among them, the weight coefficient P represents the credibility of the first term of the objective function, the weight coefficient Q represents the credibility of the second term of the objective function, A represents the coefficient matrix of the linear equation system, X represents the reservoir physical property parameters to be solved, and b represents the historical P-wave azimuth AVO gather data.

[0016] Optionally, the matrix form of the linear equation system is:

[0017]

[0018] Among them, R PP (θ n ) represents the P-wave reflection coefficient when the reflection angle is θ n When, R PP (θ m ) represents the P-wave reflection coefficient when the reflection angle is θ m When, R PP (θ f ) represents the P-wave reflection coefficient when the reflection angle is θ f When, γ is the pre-estimated ratio of the shear wave velocity to the P-wave velocity, R Vp Represents the P-wave velocity reflectivity, R Vs Represents the shear wave velocity reflectivity, R ρ Represents the density reflectivity.

[0019] Optionally, when constructing the AVO inversion model, the weight coefficient P is the inverse matrix of the covariance matrix of the P-wave azimuth AVO gather data.

[0020] Optionally, when constructing the AVO inversion model, the weight coefficient Q is calculated based on the synthetic pre-stack gather corresponding to the well area of the historical P-wave azimuth AVO gather data, and is the inverse matrix of the covariance matrix of the solution X.

[0021] The second aspect of the embodiments of the present invention provides a reservoir physical property detection device, and the device includes:

[0022] A data acquisition module, configured to acquire real-time P-wave azimuth AVO gather data of the well area;

[0023] An AVO inversion module for inputting the longitudinal wave azimuth AVO gather data into a constructed AVO inversion model to inversely obtain reservoir physical property parameters in the well area, where the reservoir physical property parameters include shear wave velocity reflectivity, longitudinal wave velocity reflectivity, and density reflectivity;

[0024] Among them, historical longitudinal wave azimuth AVO gather data is used as the data source when constructing the AVO inversion model;

[0025] The determination process of the inversion objective function of the AVO inversion model is as follows:

[0026] Based on the linear equations constructed by multi-channel reflection coefficients, determine the objective function of least squares inversion S=(AX - b) T (AX - b);

[0027] Add a constraint condition to the objective function, and the objective function is modified to S=(AX - b) T (AX - b)+λX T X, where the magnitude of the first parameter λ cannot be estimated in advance;

[0028] For the first term (AX - b) T (AX - b) of the objective function, add a weight coefficient P and for the second term X T X, add a weight coefficient Q, and the objective function is modified to S=(AX - b) T P(AX - b)+X T QX;

[0029] Correspondingly, the solution X of the AVO inversion model =(A T PA + Q) -1 (A T Pb);

[0030] Among them, the weight coefficient P represents the credibility of the first term of the objective function, the weight coefficient Q represents the credibility of the second term of the objective function, A represents the coefficient matrix of the linear equations, X represents the reservoir physical property parameters to be solved, and b represents the historical longitudinal wave azimuth AVO gather data.

[0031] Optionally, in the AVO inversion module, the matrix form of the linear equations is:

[0032]

[0033] Among them, R PP (θ n ) represents the longitudinal wave reflection coefficient when the reflection angle is θ n and R PP (θ m ) represents the longitudinal wave reflection coefficient when the reflection angle is θm Longitudinal wave reflection coefficient at time R PP (θ f ) represents that the reflection angle is θ f Longitudinal wave reflection coefficient at time, γ is the ratio of the pre-estimated shear wave velocity to the longitudinal wave velocity, R Vp Represents the longitudinal wave velocity reflectivity, R Vs Represents the shear wave velocity reflectivity, R ρ Represents the density reflectivity.

[0034] Optionally, when constructing the AVO inversion model, the weight coefficient P is the inverse matrix of the covariance matrix of the longitudinal wave azimuth AVO gather data.

[0035] Optionally, when constructing the AVO inversion model, the weight coefficient Q is calculated according to the synthetic pre-stack gather in the well area corresponding to the historical longitudinal wave azimuth AVO gather data, and is the inverse matrix of the covariance matrix of the solution X.

[0036] A third aspect of the embodiments of the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, it implements a reservoir physical property detection method as described in the first aspect of the embodiments of the present invention.

[0037] A fourth aspect of the embodiments of the present invention provides a storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements a reservoir physical property detection method as described in the first aspect of the embodiments of the present invention.

[0038] It is known that the maximum reflection angle of general AVO gather data is about 30 degrees. Therefore, the coefficient matrix A of the linear equation system constructed based on the multi-channel reflection coefficient is a rank-deficient matrix or an ill-conditioned matrix, and the condition number can reach 300. This linear equation system cannot be solved by the least squares method. In the prior art, generally, reasonable regularization is performed according to the ill-conditioned degree of the inversion problem. However, the regularization coefficient mostly uses estimated values and is not accurate. Starting from the most basic mathematical level, the present invention improves the coefficient matrix by adding constraint conditions to the original objective function of the least squares and designing the weight coefficient, so that the sensitivity of the solution X to errors becomes smaller, and the influence of data errors on the result is determined to cope with noise interference, greatly reducing the sensitivity of the solution X to errors and noise. The three parameters obtained by synchronous inversion have better accuracy. Through the high-precision solution of the three parameters, the oil and gas bearing property and oil and gas content of the well area can be judged more accurately, and even basic data support can be provided for quantitatively estimating porosity and fluid saturation, etc.

[0039] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent specific implementation part. Description of the Drawings

[0040] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and form a part of the specification. Together with the following detailed description, they are used to explain the embodiments of the present invention, but do not constitute a limitation to the embodiments of the present invention. In the accompanying drawings:

[0041] Figure 1 is an exemplary flowchart of a reservoir physical property detection method;

[0042] Figure 2 is the inversion result of the density reflectivity when the signal-to-noise ratio is 1 / 1, where part (2a) is the conventional inversion result, part (2b) is the inversion result after regularization, and part (2c) is the inversion result of the weight quantification achieved in the embodiments of the present invention;

[0043] Figure 3 is the inversion result of the density reflectivity when the signal-to-noise ratio is 3 / 1, where part (3a) is the conventional inversion result, part (3b) is the inversion result after regularization, and part (3c) is the inversion result of the weight quantification achieved in the embodiments of the present invention.

[0044] Figure 4 is the inversion result of the density reflectivity without noise, where part (4a) is the conventional inversion result, part (4b) is the inversion result after regularization, and part (4c) is the inversion result of the weight quantification achieved in the embodiments of the present invention;

[0045] Figure 5 is a schematic diagram of the consistency between the inversion result beside the wellbore and the logging data interpretation;

[0046] Figure 6 is a schematic diagram of the composition of a reservoir physical property detection device. Detailed Description of the Embodiments

[0047] The following will describe in detail the specific embodiments of the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention.

[0048] Method Embodiment 1

[0049] This embodiment provides a reservoir physical property detection method for synchronous three-parameter inversion in a target well area. After obtaining the reservoir physical property parameters, the petrophysical model can be used to judge the oil and gas content in the reservoir of the target well area and quantitatively calculate the oil and gas content.

[0050] Specifically, as shown in Figure 1 a reservoir physical property detection method specifically includes the following implementation steps:

[0051] S100. Obtain the real-time longitudinal wave azimuth AVO gather data of the well area. This real-time longitudinal wave azimuth AVO gather data can also be referred to as the actual measurement value or the actual observation data.

[0052] S200. Input the longitudinal wave azimuth AVO gather data into the constructed AVO inversion model, and inversely obtain the reservoir physical property parameters of the target well area. The reservoir physical property parameters include the shear wave velocity reflectivity R Vs , the longitudinal wave velocity reflectivity R Vp and the density reflectivity R ρ . The inversion objective function of the AVO inversion model incorporates constraint conditions and weight coefficients.

[0053] Among them, the constructed AVO inversion model is the AVO synchronous three-parameter inversion model. When constructing this model, the historical longitudinal wave azimuth AVO gather data is used as the data source.

[0054] In addition, the inversion objective function of this model is determined through the following steps:

[0055] S01. Based on the linear equations constructed by multi-channel reflection coefficients, determine the objective function of the least squares inversion. At this time, the equations of the objective function are the original equations and are expressed as S = (AX - b) T (AX - b). Among them, A represents the coefficient matrix of the linear equations, X represents the reservoir physical property parameters to be solved, and b represents the historical longitudinal wave azimuth AVO gather data, which can also be referred to as the historical measurement value or the historical observation data.

[0056] Because the maximum reflection angle of a general AVO gather is about 30 degrees, the coefficient matrix of the linear equations constructed based on multi-channel reflection coefficients is a rank-deficient matrix or an ill-conditioned matrix, and the condition number reaches 300. Therefore, the equations of the objective function at this time cannot be solved by the least squares method.

[0057] S02. Add constraint conditions to the objective function. After adding the constraint conditions, the objective function is modified to S = (AX - b) T (AX - b) + λX T X.

[0058] By adding the constraint condition λX T X, the coefficient matrix is improved, and the sharp expansion effect of noise in the ill-conditioned matrix is suppressed, making it possible for the equations of the objective function to obtain meaningful solutions. This depends on the value of the first parameter λ, but the size of the first parameter λ cannot be estimated in advance. Therefore, the objective function after adding the constraint conditions still needs to be improved.

[0059] S03. Add the weight coefficient P to the first term (AX - b) T (AX - b) of the objective function and the second term X TAdd the weight coefficient Q. After adding the weight coefficient, the objective function is modified to S = (AX - b) T P(AX - b)+X T QX. Among them, the weight coefficient P represents the credibility of the first term (AX - b) T (AX - b) of the objective function, that is, the contribution degree of the first term to the objective function, and the weight coefficient Q represents the credibility of the second term X T X of the objective function, that is, the contribution degree of the second term to the objective function.

[0060] The contributions of the two terms of the objective function to the objective function are objectively determined by the weight coefficient P and the weight coefficient Q, so that the solution of the objective function is always within a reasonable range.

[0061] After determining the final inversion objective function through the above steps S01 - S03, solve the objective function equations based on the least squares method, and correspondingly obtain the solution X=(A T PA + Q) -1 (A T Pb).

[0062] Method Embodiment Two

[0063] Method Embodiment Two is a preference of Method Embodiment One. Specifically, this embodiment provides a reservoir physical property detection method for synchronous three - parameter inversion in a target well area. After obtaining the reservoir physical property parameters, the petrophysical model can be used to judge the oil - gas bearing property in the reservoir of the target well area and quantitatively calculate the oil - gas content.

[0064] Combined with Figure 1 as shown, a reservoir physical property detection method specifically includes the following implementation steps:

[0065] S100. Obtain the real - time longitudinal wave azimuth AVO gather data of the well area. This real - time longitudinal wave azimuth AVO gather data can also be called the actual measurement value or the actual observation data.

[0066] S200. Input the longitudinal wave azimuth AVO gather data into the constructed AVO inversion model, and invert to obtain the reservoir physical property parameters of the target well area. The reservoir physical property parameters include the shear wave velocity reflectivity R Vs 、the longitudinal wave velocity reflectivity R Vp and the density reflectivity R ρ .

[0067] Among them, the constructed AVO inversion model is an AVO synchronous three - parameter inversion model. When constructing this model, the historical longitudinal wave azimuth AVO gather data is used as the data source.

[0068] In addition, the inversion objective function of this model is determined through the following steps:

[0069] S01. Determine the objective function of the least squares inversion based on the linear equations constructed from multi-channel reflection coefficients. At this time, the equation system of the objective function is the original equation system and is expressed as S = (AX - b). T (AX - b).

[0070] Specifically, the matrix form of the linear equations constructed from multi-channel reflection coefficients is specifically as follows:

[0071]

[0072] Among them, A represents the coefficient matrix of the linear equations, X represents the reservoir physical property parameters to be solved, and b represents the historical P-wave azimuth AVO gather data, which can also be called historical measurement values or historical observation data;

[0073]

[0074] R PP (θ n ) represents the P-wave reflection coefficient when the reflection angle is θ n , R PP (θ m ) represents the P-wave reflection coefficient when the reflection angle is θ m , R PP (θ f ) represents the P-wave reflection coefficient when the reflection angle is θ f , γ is the ratio of the pre-estimated shear wave velocity to the P-wave velocity, R Vp represents the P-wave velocity reflectivity, R Vs represents the shear wave velocity reflectivity, R ρ represents the density reflectivity.

[0075] Because the maximum reflection angle of a general AVO gather is about 30 degrees, the coefficient matrix of the linear equations constructed from multi-channel reflection coefficients is a rank-deficient matrix or an ill-conditioned matrix, and the condition number reaches 300. Therefore, the equation system of the objective function cannot be solved by the least squares method at this time.

[0076] S02. Add a constraint condition to the objective function. After adding the constraint condition, the objective function is modified to S = (AX - b) T (AX - b) + λX T X.

[0077] After adding the constraint condition λX T X, the coefficient matrix is improved, and the sharp expansion effect of noise in the ill-conditioned matrix is suppressed, making it possible for the equation system of the objective function to obtain a meaningful solution, which depends on the value of the first parameter λ. However, the size of the first parameter λ cannot be estimated in advance. Therefore, the objective function after adding the constraint condition still needs to be improved.

[0078] S03. For the first term of the objective function (AX - b) T (AX - b), add the weight coefficient P and the second term X T For X, add the weight coefficient Q. After adding the weight coefficients, the objective function is modified to S = (AX - b) T P(AX - b)+X T QX. Among them, the weight coefficient P represents the credibility of the first term of the objective function, that is, the contribution degree of the first term (AX - b) T (AX - b) to the objective function, and the weight coefficient Q represents the credibility of the second term X T X of the objective function, that is, the contribution degree of the second term to the objective function.

[0079] The contributions of the two terms of the objective function to the objective function are objectively determined by the weight coefficient P and the weight coefficient Q, so that the solution of the objective function is always within a reasonable range.

[0080] After determining the inversion objective function through the above steps S01 - S03, solve the objective function equations based on the least squares method, and correspondingly obtain the solution X = (A T PA + Q) -1 (A T Pb).

[0081] Method Embodiment Three

[0082] Method Embodiment Three is a preference of Method Embodiment Two. Specifically, this embodiment provides a reservoir physical property detection method for synchronous three - parameter inversion in a target well area. After obtaining the reservoir physical property parameters, the petrophysical model can be used to judge the oil - gas bearing property in the reservoir of the target well area and quantitatively calculate the oil - gas content.

[0083] Combined with Figure 1 shown, a reservoir physical property detection method specifically includes the following implementation steps:

[0084] S100. Obtain the real - time longitudinal wave azimuth AVO gather data of the well area. This real - time longitudinal wave azimuth AVO gather data can also be called the actual measurement value or the actual observation data.

[0085] S200. Input the longitudinal wave azimuth AVO gather data into the constructed AVO inversion model, and inversely obtain the reservoir physical property parameters of the target well area. The reservoir physical property parameters include the shear wave velocity reflectivity R Vs 、the longitudinal wave velocity reflectivity R Vp and the density reflectivity R ρ .

[0086] Among them, the constructed AVO inversion model is an AVO synchronous three - parameter inversion model. When constructing this model, the historical longitudinal wave azimuth AVO gather data is used as the data source.

[0087] In addition, the inversion objective function of the model is determined through the following steps:

[0088] S01. Based on the linear equations constructed from multi-channel reflection coefficients, determine the objective function of least squares inversion. At this time, the equation system of the objective function is the original equation system and is expressed as S = (AX - b) T (AX - b).

[0089] Specifically, the linear equations constructed based on multi-channel reflection coefficients are specifically as follows:

[0090]

[0091] Among them, A represents the coefficient matrix of the linear equations, X represents the reservoir physical property parameters to be solved, and b represents the historical P-wave azimuth AVO gather data, which can also be called historical measurement values or historical observation data;

[0092]

[0093] R PP (θ n ) represents the P-wave reflection coefficient when the reflection angle is θ n When, R PP (θ m ) represents the P-wave reflection coefficient when the reflection angle is θ m When, R PP (θ f ) represents the P-wave reflection coefficient when the reflection angle is θ f When, γ is the ratio of the pre-estimated shear wave velocity to the P-wave velocity, R Vp Represents the P-wave velocity reflectivity, R Vs Represents the shear wave velocity reflectivity, R ρ Represents the density reflectivity.

[0094] Because the maximum reflection angle of a general AVO gather is about 30 degrees, the coefficient matrix of the linear equations constructed based on multi-channel reflection coefficients is a rank-deficient matrix or an ill-conditioned matrix, and the condition number reaches 300. Therefore, the equation system of the objective function at this time cannot be solved by the least squares method.

[0095] S02. Add a constraint condition to the objective function. After adding the constraint condition, the objective function is modified to S = (AX - b) T (AX - b) + λX T X.

[0096] By adding the constraint condition λX TAfter X, the coefficient matrix is improved, and the sharp expansion effect of noise in the ill-conditioned matrix is suppressed, making it possible to obtain a meaningful solution to the objective function system of equations, which depends on the value of the first parameter λ. However, the size of the first parameter λ cannot be estimated in advance. Therefore, it is still necessary to improve the objective function after adding the constraint condition.

[0097] S03. For the first term of the objective function (AX - b) T (AX - b) add the weight coefficient P and the second term X T X add the weight coefficient Q. After adding the weight coefficients, the objective function is modified to S = (AX - b) T P(AX - b) + X T QX. Among them, the weight coefficient P represents the credibility of the first term of the objective function, that is, the contribution degree of the first term (AX - b) T (AX - b) to the objective function, and the weight coefficient Q represents the credibility of the second term X of the objective function T X, that is, the contribution degree of the second term to the objective function.

[0098] Specifically, the weight coefficient P is the inverse matrix of the covariance matrix of the longitudinal wave azimuth AVO gather data, that is, the inverse matrix of the covariance matrix of the measured value b. The weight coefficient Q is the inverse matrix of the covariance matrix of the solution X and can be calculated from the synthetic pre-stack gather corresponding to the well area of the historical longitudinal wave azimuth AVO gather data.

[0099] The principle of defining the weight coefficient P as the inverse matrix of the covariance matrix of the longitudinal wave azimuth AVO gather data and the weight coefficient Q as the inverse matrix of the covariance matrix of the solution X is explained as follows.

[0100] First, consider the Mahalanobis distance between a point X in the n-dimensional space and a distribution as:

[0101] D M =[(X - μ) T S -1 (X - μ)] 1 / 2 ;

[0102] Among them, μ is the average value of the distribution, and S is the covariance matrix of the distribution;

[0103] Based on the above principle, the objective function can be further modified to: S = (AX - b) T P(AX - b) + (X - Xavg) TQ(X - Xavg), where Xavg represents the average value of the solution X and takes the value of 0 in the reflectivity signal. Taking a one-dimensional variable as an example, if the noise of the measured value b is large, that is, the noise of the well area observation data is large, at this time, the variance of the observation data is large, the reciprocal of the variance is small, that is, the weight coefficient P is small. Then, the first term of the objective function is characterized by the weight coefficient P with low credibility, and correspondingly, its contribution to the objective function is small. Similarly, if the change of the solution X is small, that is, the reservoir physical property parameters in the well area change little, at this time, the variance of the reservoir physical property parameters is small, the reciprocal of the variance is large, that is, the weight coefficient Q is large. Then, the second term of the objective function is characterized by the weight coefficient Q with high credibility, and correspondingly, its contribution to the objective function is large. It can be seen that the contributions of the two terms of the objective function to the objective function are objectively determined by the weight coefficient P and the weight coefficient Q, so that the solution of the objective function is always within a reasonable range.

[0104] After determining the inversion objective function through the above steps S01 - S03, solve the objective function system of equations based on the least squares method, and correspondingly obtain the solution X = (A T PA + Q) -1 (A T Pb).

[0105] The following content takes geophysical inversion data as an example to illustrate the feasibility of the reservoir physical property detection method provided by the above method embodiments.

[0106] Based on the matrix form of the linear equation system , take appropriate values for the reflection angle θ n , the reflection angle θ m , the reflection angle θ f , and the ratio γ of the shear wave velocity to the compressional wave velocity. For example, take the reflection angle θ n as 10 degrees, the reflection angle θ m as 20 degrees, the reflection angle θ f as 30 degrees, and γ as 1 / 2. The coefficient matrix obtained at this time is:

[0107]

[0108] Furthermore, perform SVD decomposition on the coefficient matrix, and the decomposition result is:

[0109]

[0110] It can be seen from the decomposition result that one value in the middle triangular matrix is close to zero, indicating that the rank of the matrix is close to 2 at this time. The linear equation system at this time is nearly linearly dependent, or in other words, the matrix determinant must be very small, and it is an ill-conditioned matrix or a rank-deficient matrix.

[0111] To illustrate the feasibility of the reservoir physical property detection method provided by the above method embodiments, the verification idea is as follows: Using the forward modeling method, according to the above linear equations, the measured value b is calculated from the determined solution X, and then after adding noise to the measured value b, it is used as the input for the prestack synchronous three-parameter inversion, and the solution X is inverted. Finally, after statistical results, the feasibility and accuracy of the method embodiments are judged.

[0112] The following description gives a specific application example of the above verification idea. Assume a normal distribution with a mean of 0 and a standard deviation of 0.1, and randomly extract the longitudinal wave velocity reflectivity R Vp , the shear wave velocity reflectivity R Vs and the density reflectivity R ρ (that is, assume the solution X is known and is the true value), substitute them into the above linear equations to forward calculate a series of measured values b, add noise to the measured values b as the measured values to be inverted, and then use three prestack synchronous three-parameter inversion methods to invert the solution X respectively. The first prestack synchronous three-parameter inversion method is to directly solve without any processing. The second prestack synchronous three-parameter inversion method is the regularization inversion, but the value of the regularization coefficient λ' is roughly estimated. For example, in this application example, it is set to 0.001. The third prestack synchronous three-parameter inversion method uses the method implemented by the above method embodiments, that is, both terms of the objective function have reasonable weight coefficients, that is, the weights have been quantified. Taking the most common inversion parameter density reflectivity R ρ as an example, to investigate the feasibility of this algorithm, the verification results are as Figures 2 to 4 shown.

[0113] Figure 2 The inversion result of the density reflectivity when the signal-to-noise ratio is 1 / 1 (that is, 50% of the measured value is noise). Among them, part (2a) is the inversion result using the first prestack synchronous three-parameter inversion method. The inversion result has almost no relationship with the true value, the correlation coefficient is 0.0279, and the value range is very large, far exceeding the reasonable range, and the inversion fails completely; part (2b) is the inversion result using the second prestack synchronous three-parameter inversion method. The inversion result has a very weak linear relationship with the true value, the correlation coefficient is 0.1475, and the inversion is not successful either; part (2c) is the inversion result of the reservoir physical property detection method implemented by the method embodiments. The inversion result obtained after weight quantification has an obvious linear relationship with the true value, and the correlation coefficient is 0.5268. It can be seen that the inversion result obtained by using the method embodiments under this noise condition is the best.

[0114] Figure 3The inversion results of density reflectivity when the signal-to-noise ratio is 3 / 1 (i.e., 25% of the measured value is noise). Among them, the inversion results in part (3a) are obtained by using the first prestack synchronous three-parameter inversion method, and the inversion results are still invalid, with a correlation coefficient of only 0.0467; the inversion results in part (3b) are obtained by using the second prestack synchronous three-parameter inversion method, and the linear relationship between the inversion results and the true value has been improved, with a correlation coefficient of 0.3503, and the inversion results still need to be improved; the inversion results in part (3c) are the inversion results of the reservoir physical property detection method implemented by the method embodiment. After weight quantification, the linear relationship between the inversion results and the true value is further enhanced, with a correlation coefficient of 0.6859. At this time, the obtained inversion results can be used for signal interpretation.

[0115] Figure 4 The inversion results of density reflectivity without noise. Among them, due to the absence of noise, the inversion results obtained by using the first prestack synchronous three-parameter inversion method are correct, as shown in part (4a). At this time, the inversion results are the same as the true value; the inversion results obtained by using the second prestack synchronous three-parameter inversion method are shown in part (4b). Since the regularization parameter λ' is relatively large without noise, the results are not accurately inverted; due to weight quantification, that is, the weight is automatically adjusted according to the noise level, the linearity between the inversion results of the reservoir physical property detection method implemented by the method embodiment and the true value is very good, as shown in part (4c).

[0116] Finally, the real-time longitudinal wave azimuth AVO gather data of the target well area is used for testing, and the test results are as Figure 5 shown. It can be seen from the figure that the real-time longitudinal wave azimuth AVO gather data is input into the AVO inversion model constructed in the method embodiment, and three parameters are inverted. Through the three parameters, the oil and gas content and the amount of oil and gas in the reservoir are interpreted. At this time, the oil and gas-bearing sections characterized by the interpretation of the three parameters are consistent with the oil and gas-bearing sections obtained by logging data interpretation.

[0117] Device embodiment

[0118] Combined with Figure 6 shown, this embodiment provides a reservoir physical property detection device, which includes a data acquisition module and an AVO inversion module connected in sequence.

[0119] The data acquisition module is used to acquire the real-time longitudinal wave azimuth AVO gather data of the well area.

[0120] The AVO inversion module is used to input the longitudinal wave azimuth AVO gather data into the constructed AVO inversion model, and invert the reservoir physical property parameters of the well area. The reservoir physical property parameters include shear wave velocity reflectivity, longitudinal wave velocity reflectivity, and density reflectivity.

[0121] Among them, when the AVO inversion model is constructed, the historical longitudinal wave azimuth AVO gather data is used as the data source.

[0122] The determination process of the inversion objective function of the AVO inversion model is as follows:

[0123] Based on the linear equations constructed from multi-channel reflection coefficients, determine the objective function of least squares inversion \(S=(AX - b)\) T (AX - b);

[0124] Add a constraint condition to the objective function, and the objective function is modified to \(S=(AX - b)\) T (AX - b)+\(\lambda X\) T X, where the magnitude of \(\lambda\) cannot be estimated in advance;

[0125] For the first term \((AX - b)\) of the objective function T (AX - b) add a weight coefficient \(P\) and for the second term \(X\) T X add a weight coefficient \(Q\), and the objective function is modified to \(S=(AX - b)\) T P(AX - b)+X T QX;

[0126] Correspondingly, the solution \(X=(A\) T PA + Q) -1 (A T Pb);

[0127] Among them, the weight coefficient \(P\) represents the credibility of the first term of the objective function, the weight coefficient \(Q\) represents the credibility of the second term of the objective function, \(A\) represents the coefficient matrix of the linear equations, \(X\) represents the reservoir physical property parameters to be solved, and \(b\) represents the historical P-wave azimuth AVO gather data.

[0128] Optionally, in the AVO inversion module, the matrix form of the linear equations is:

[0129]

[0130] Among them, R PP (\(\theta\) n ) represents the P-wave reflection coefficient when the reflection angle is \(\theta\) n At this time, \(R\) PP (\(\theta\) m ) represents the P-wave reflection coefficient when the reflection angle is \(\theta\) m At this time, \(R\) PP (\(\theta\) f ) represents the P-wave reflection coefficient when the reflection angle is \(\theta\) f At this time, \(\gamma\) is the ratio of the pre-estimated shear wave velocity to the P-wave velocity, \(R\) Vp Represents the P-wave velocity reflectivity, \(R\) Vs Represents the shear wave velocity reflectivity, \(R\) ρ Represents the density reflectivity.

[0131] Optionally, when constructing the AVO inversion model, the weight coefficient P is the inverse matrix of the covariance matrix of the P-wave azimuth AVO gather data.

[0132] Optionally, when constructing the AVO inversion model, the weight coefficient Q is calculated based on the pre-stack gather synthesized in the well area corresponding to the historical P-wave azimuth AVO gather data, and is the inverse matrix of the covariance matrix of the solution X.

[0133] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0134] On the other hand, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the reservoir physical property detection method described in any one of Method Embodiments 1 to 3.

[0135] Among them, the memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash RAM. The memory includes at least one storage chip.

[0136] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0137] In another aspect, the present invention also provides a machine-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the reservoir physical property detection method described in any one of Method Embodiments 1 to 3 is implemented.

[0138] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the essence of the above technical solution, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. A reservoir physical property detection method, characterized in that Including: Obtaining real-time vertical wave azimuth AVO gather data of a well area; Inputting the vertical wave azimuth AVO gather data into a constructed AVO inversion model to inversely obtain reservoir physical property parameters of the well area, where the reservoir physical property parameters include shear wave velocity reflectivity, compressional wave velocity reflectivity, and density reflectivity; When constructing the AVO inversion model, using historical vertical wave azimuth AVO gather data as a data source; The determination process of the inversion objective function of the AVO inversion model is as follows: Based on the linear equations constructed from multi-channel reflection coefficients, the objective function of least squares inversion is determined as S = (AX - b) T (AX - b); Add a constraint condition to the objective function, and the objective function is modified to S = (AX - b) T (AX - b) + λX T X, where the magnitude of the first parameter λ cannot be estimated in advance; For the first term (AX - b) of the objective function T (AX - b) add a weight coefficient P and the second term X T X add a weight coefficient Q, and the objective function is modified to S = (AX - b) T P(AX - b) + X T QX; Accordingly, the solution X of the AVO inversion model is X = (A T PA + Q) -1 (A T Pb); Among them, the weight coefficient P represents the credibility of the first term of the objective function, the weight coefficient Q represents the credibility of the second term of the objective function, A represents the coefficient matrix of the linear equation system, X represents the reservoir physical property parameters to be solved, and b represents the historical vertical wave azimuth AVO gather data.

2. The reservoir physical property detection method according to claim 1, wherein The matrix form of the linear equation system is: Among them, R PP (θ n ) represents the longitudinal wave reflection coefficient when the reflection angle is θ n When, R PP (θ m ) represents the longitudinal wave reflection coefficient when the reflection angle is θ m When, R PP (θ f ) represents the longitudinal wave reflection coefficient when the reflection angle is θ f When, γ is the ratio of the pre-estimated shear wave velocity to the longitudinal wave velocity, R Vp represents the longitudinal wave velocity reflectivity, R Vs represents the shear wave velocity reflectivity, R ρ represents the density reflectivity.

3. A reservoir physical property detection method according to claim 1, characterized in that When constructing the AVO inversion model, the weight coefficient P is the inverse matrix of the covariance matrix of the vertical wave azimuth AVO gather data.

4. The reservoir physical property detection method according to claim 1, characterized in that When constructing the AVO inversion model, the weight coefficient Q is calculated based on the synthetic pre-stack gather of the well area corresponding to the historical vertical wave azimuth AVO gather data, and is the inverse matrix of the covariance matrix of the solution X.

5. A reservoir physical property detection device, characterized in that, The device includes: A data acquisition module for obtaining real-time vertical wave azimuth AVO gather data of a well area; An AVO inversion module for inputting the vertical wave azimuth AVO gather data into a constructed AVO inversion model to inversely obtain reservoir physical property parameters of the well area, where the reservoir physical property parameters include shear wave velocity reflectivity, compressional wave velocity reflectivity, and density reflectivity; Among them, when constructing the AVO inversion model, using historical vertical wave azimuth AVO gather data as a data source; The determination process of the inversion objective function of the AVO inversion model is as follows: Based on the linear equations constructed from multiple reflection coefficients, determine the objective function S of the least squares inversion as S = (AX - b) T (AX - b); Add a constraint to the objective function, and the objective function is modified to S = (AX - b) T (AX - b) + λX T X, where the magnitude of the first parameter λ cannot be estimated in advance; For the first term of the objective function (AX - b) T (AX - b) add the weight coefficient P and the second term X T For X add the weight coefficient Q, and the objective function is modified to S = (AX - b) T P(AX - b)+X T QX; Accordingly, the solution X of the AVO inversion model is X = (A T PA + Q) -1 (A T Pb); Among them, the weight coefficient P represents the credibility of the first term of the objective function, the weight coefficient Q represents the credibility of the second term of the objective function, A represents the coefficient matrix of the linear equation system, X represents the reservoir physical property parameters to be solved, and b represents the historical vertical wave azimuth AVO gather data.

6. The reservoir physical property detection device according to claim 5, wherein In the AVO inversion module, the matrix form of the linear equation system is: Among them, R PP (θ n ) represents the longitudinal wave reflection coefficient when the reflection angle is θ n When, R PP (θ m ) represents the longitudinal wave reflection coefficient when the reflection angle is θ m When, R PP (θ f ) represents the longitudinal wave reflection coefficient when the reflection angle is θ f When, γ is the ratio of the pre-estimated shear wave velocity to the longitudinal wave velocity, and R Vp represents the longitudinal wave velocity reflectivity, and R Vs represents the shear wave velocity reflectivity, and R ρ represents the density reflectivity.

7. The reservoir physical property detection device according to claim 5, characterized in that, When constructing the AVO inversion model, the weight coefficient P is the inverse matrix of the covariance matrix of the vertical wave azimuth AVO gather data.

8. A reservoir physical property detection device according to claim 5, characterized in that, When constructing the AVO inversion model, the weight coefficient Q is calculated based on the synthetic pre-stack gather of the well area corresponding to the historical vertical wave azimuth AVO gather data, and is the inverse matrix of the covariance matrix of the solution X.

9. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a reservoir physical property detection method as described in any one of claims 1 to 4.

10. A storage medium, on which a computer program is stored, characterized in that, When the computer program is executed by the processor, it implements a reservoir physical property detection method as described in any one of claims 1 to 4.

Images