Reservoir fluid identification method based on joint of seismic elastic attribute, resistivity and polarizability

Abstract

The application discloses a reservoir fluid identification method based on joint of seismic elastic properties, resistivity and polarizability, and comprises the following steps: obtaining seismic detection data and electromagnetic detection data of a region to be detected and identified, extracting resistivity, polarizability and seismic wave velocity to obtain a first physical property model; obtaining a second physical property model by using K-means clustering analysis and regression analysis according to logging data containing resistivity, polarizability and seismic wave velocity and combining with rock physical experiment analysis; correcting the first physical property model by using the second physical property model and performing joint inversion; obtaining model parameters containing resistivity, polarizability and velocity after inversion, and performing reservoir fluid prediction according to the model parameters. Through the above scheme, the application has the advantages of simple logic, accuracy and reliability and has high practical value and popularization value in the field of geophysical comprehensive detection technology.

Description

Technical Field

[0001] This invention relates to the field of geophysical integrated exploration technology, and in particular to a reservoir fluid identification method based on the combined properties of seismic elasticity, resistivity and polarizability. Background Technology

[0002] Earth science and resource development have always been a focus of human attention, particularly in oil and gas exploration, reservoir evaluation, and development. To better develop underground resources, petroleum engineers and geologists have been seeking more accurate and efficient reservoir fluid detection technologies. Over the past few decades, geophysical exploration technologies have been continuously evolving, and scientists are constantly trying to combine multiple exploration techniques to obtain more crucial information.

[0003] Seismic exploration is a commonly used geophysical exploration method that identifies the presence of oil and gas reservoirs or mineral deposits by detecting the reflection and propagation characteristics of underground sound waves. Seismic waves undergo refraction and reflection as they propagate through different reservoir rocks. The propagation speed and direction of these waves are related to the physical properties of the underground rocks (such as density and Poisson's ratio). Therefore, seismic data can be used to understand the underground structures and the presence of reservoir fluids. However, traditional seismic exploration methods can only provide information on the elastic properties of underground rocks, offering limited understanding of fluid types and distribution patterns.

[0004] To overcome the limitations of traditional seismic exploration methods, researchers have begun to explore combining other geophysical exploration techniques to improve the accuracy of reservoir fluid detection. Resistivity logging and electromagnetic exploration are commonly used geophysical exploration methods that determine the distribution of fluids (such as water, oil, and gas) by measuring the resistivity of underground rocks. Different types of fluids have different electrical conductivity in rocks, so resistivity logging can be used to distinguish between different types of reservoir fluids. However, resistivity methods also have limitations; they are mainly suitable for identifying aquifers and are less effective for unconventional oil and gas reservoirs.

[0005] Another promising geophysical exploration method is polarizability logging and electromagnetic exploration (EMA) to extract polarizability parameters. Polarizability refers to the ability of a rock to polarize under the influence of an applied electric field. When fluids are present in the rock, the polarizability of the fluid differs from that of the rock; therefore, the presence of fluids can be identified by measuring the rock's polarizability. Polarizability logging and EMA polarizability extraction methods have certain advantages for detecting unconventional oil and gas reservoirs; however, they typically do not provide information on the reservoir's elastic properties.

[0006] In summary, traditional seismic exploration methods, electromagnetic exploration methods, resistivity logging, and polarizability logging each have their own advantages and limitations. Therefore, petroleum engineers and geologists have begun to explore the combined use of these geophysical exploration techniques to obtain more comprehensive reservoir fluid information. However, to date, there are no technical introductions on the integrated analysis and processing of data from seismic exploration, electromagnetic exploration, resistivity logging, and polarizability logging; further improvements are needed to enhance the accuracy and reliability of reservoir fluid detection.

[0007] In Peng Guomin's article "Research on Joint Inversion Method of CSEM and Seismic Based on Cross-Gradient," it is mentioned that this joint inversion method establishes a joint inversion objective function for controlled-source electromagnetic and seismic synchronous inversion. The objective function is solved using the nonlinear conjugate gradient (NLCG) algorithm, and an adaptive regularization method is adopted to adjust the model regularization factor based on data fitting between two adjacent iterations. Specifically, the following three methods are used in the joint inversion of controlled-source electromagnetic and seismic data: adaptive correction method, electromagnetic frequency division multi-scale and multi-grid inversion method, and cross-gradient weighted method.

[0008] However, this technique involves multiple steps and algorithms, including establishing a joint objective function, solving it using a nonlinear conjugate gradient algorithm, and employing adaptive regularization methods. The complexity of these steps increases the difficulty of implementation and places high demands on the operator. Specifically, the method updates model parameters by fitting data from two adjacent iterations; this dependency may lead to error accumulation during iteration, affecting the accuracy of the inversion results. Furthermore, the method uses adaptive regularization to adjust the model's regularization factor. However, choosing an appropriate regularization factor remains a challenge. An inappropriate choice may result in overly smooth or overly complex inversion results, affecting imaging quality. Moreover, this method is primarily applied to joint inversion problems involving controlled-source electromagnetic and seismic data. Therefore, it may not be applicable to inversion problems in other fields or with different data types, limiting its applicability. The computational complexity of the nonlinear conjugate gradient algorithm is high, especially for high-dimensional problems or large-scale datasets. This can lead to a lengthy inversion process, limiting the method's real-time performance and practicality.

[0009] Therefore, there is an urgent need to propose a simple, accurate, and reliable reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability. Summary of the Invention

[0010] To address the aforementioned problems, the present invention aims to provide a reservoir fluid identification method based on the combined analysis of seismic elastic properties, resistivity, and polarizability. Accurate location and identification of reservoir fluids are crucial in oil and gas exploration. Traditional seismic exploration techniques can obtain the elastic properties of subsurface rocks, but their ability to detect fluids is limited; while resistivity and polarizability measurement techniques, although sensitive to fluids, cannot directly provide elastic information about rock reservoirs. The technical solution adopted in this invention is as follows:

[0011] A reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability includes the following steps:

[0012] Seismic and electromagnetic detection data of the area to be detected and identified are acquired, and resistivity, polarizability and seismic wave velocity are extracted to obtain the first physical property model.

[0013] Based on well logging data containing resistivity, polarizability, and seismic wave velocity, and combined with rock physics experimental analysis, a second physical property model was obtained using K-means clustering analysis and regression analysis.

[0014] The first physical property model is modified using the second physical property model, and a joint inversion is performed.

[0015] The model parameters containing resistivity, polarizability, and velocity are obtained after inversion, and reservoir fluid prediction is performed based on the model parameters.

[0016] Furthermore, the K-means clustering analysis includes the following steps:

[0017] Acquire well logging data containing resistivity, polarizability, and seismic wave velocity, and construct a dataset X = {x1, x2, ..., x} to be classified. n}; where n∈k; where k represents the number of clusters;

[0018] Initialize the dataset to be classified as X = {x1, x2, ..., x...} n The cluster centers c1, c2, ..., c of} k ;

[0019] The dataset to be classified is X = {x1, x2, ..., x...} n The expression for class allocation is as follows:

[0020]

[0021] Where x represents the dataset to be classified, X = {x1, x2, ..., x} n Data within}; C j Represents the j-th family; c j This indicates the updated class center;

[0022] Calculate the average value of the data points contained in any family, and use this average value as the updated family center c. j Its expression is:

[0023]

[0024] Where, N j This represents the number of data points within the j-th class;

[0025] The clustering criterion function is constructed as follows:

[0026]

[0027] Furthermore, K-means clustering analysis was used to cluster resistivity and seismic wave velocity, and regression analysis was used to adjust the resistivity or velocity model.

[0028] Preferably, the regression analysis is one of linear regression analysis, logarithmic regression analysis, exponential regression analysis, or quadratic regression analysis.

[0029] Furthermore, the reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability also includes: in the joint inversion, adjusting the weighting factor W of the meshing grid according to the distance of the seismic survey lines to the cross-sections in the seismic exploration. t,i Assignment, its expression is:

[0030]

[0031] Where J represents the number of meshes; dn j Let represent the distance from the i-th mesh to the j-th mesh marked as 1, and Δx j Δy represents the number of grids that differ in the x-direction from the j-th grid (marked as 1) within the plane containing the i-th grid; j The number of grids that differ in the y-direction between the i-th grid and the j-th grid (marked as 1) within the plane containing the i-th grid. Here, x and y are the two directions of the Cartesian coordinate system, defined according to the actual measurement point location.

[0032] Furthermore, the first physical property model is modified using the second physical property model, including the following steps:

[0033] The structural boundary of the first physical model is obtained by using the second physical property model and K-means clustering analysis;

[0034] Regression analysis was used to adjust the predicted values ​​of resistivity or seismic wave velocity in the first physical model;

[0035] Substitute the values ​​into the joint inversion process and determine if the termination condition is met. If not, continue with K-means clustering and regression analysis until the condition is met. This termination condition is a threshold value set for the objective function; values ​​below this threshold indicate that the termination condition has been met.

[0036] Furthermore, seismic and electromagnetic detection data of the area to be detected and identified are acquired, and a cross-gradient function is constructed, the expression of which is:

[0037]

[0038] In the formula, τ(x,y,z) is the cross gradient function. The gradient of the resistivity model, This represents the gradient of the velocity model.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] (1) This invention integrates and processes data from seismic exploration, electromagnetic exploration, resistivity logging, and polarizability logging. By acquiring seismic data, information on the elastic properties of underground rocks can be obtained; by acquiring resistivity data, different types of fluids can be identified; and by acquiring polarizability data, this invention can further confirm the presence and distribution of fluids. This invention interprets these three types of data together, allowing them to corroborate each other and improve the accuracy and reliability of reservoir fluid detection.

[0041] (2) This invention extracts and analyzes the elastic properties of underground rocks using acquired seismic wave propagation data. This invention employs a seismic inversion algorithm to infer elastic properties such as density and Poisson's ratio of reservoir rocks using information on the velocity and amplitude of seismic waves propagating underground. Furthermore, this invention extracts and analyzes the resistivity characteristics of underground rocks using acquired resistivity measurement data. This invention employs a resistivity imaging algorithm and an electromagnetic exploration inversion resistivity method to reconstruct the resistivity distribution of underground rocks based on resistivity measurement data.

[0042] (3) This invention addresses the polarizability characteristics of reservoir fluids by designing a polarizability measurement experiment. This invention extracts the polarizability information of reservoir fluids by measuring the polarizability of underground rocks. This invention also extracts polarizability information from electromagnetic exploration signals using an inversion algorithm.

[0043] In summary, this invention has the advantages of simple logic and high accuracy and reliability, and has high practical and promotional value in the field of geophysical integrated exploration technology. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope of protection. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a logic flowchart of the present invention. Detailed Implementation

[0046] To make the objectives, technical solutions, and advantages of this application clearer, the present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0047] In this embodiment, the term "and / or" is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone.

[0048] The terms "first" and "second," etc., used in the specification and claims of this embodiment are used to distinguish different objects, not to describe a specific order of objects. For example, "first target object" and "second target object," etc., are used to distinguish different target objects, not to describe a specific order of target objects.

[0049] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0050] In the description of the embodiments in this application, unless otherwise stated, "multiple" means two or more. For example, multiple processing units means two or more processing units; multiple systems means two or more systems.

[0051] like Figure 1 As shown, this embodiment provides a reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability, which includes the following steps:

[0052] First, seismic and electromagnetic detection data of the area to be detected and identified are acquired, and a cross-gradient joint inversion objective function is constructed to obtain the first physical property model. In this embodiment, the expression of the cross-gradient joint inversion objective function is:

[0053] Given the complexity of rock physical properties, cross-gradient analysis primarily seeks structural similarities across models with different physical parameters, making fewer assumptions about the relationships between physical property parameters. This allows for the inversion of models originating from different data but exhibiting structural similarities. The cross-gradient function for joint inversion of electromagnetic and seismic data is as follows:

[0054]

[0055] In the formula, τ(x,y,z) is the cross gradient function. The gradient of the resistivity model or polarizability model. This represents the gradient of the velocity model.

[0056] The second step involves collecting well logging data containing resistivity, polarizability, and seismic wave velocity, and combining this with rock physics experimental analysis. K-means clustering and regression analysis are then used to obtain the second physical property model.

[0057] In this embodiment, the specific steps of the K-means clustering analysis are as follows:

[0058] (1) Obtain well logging data containing resistivity, polarizability, and seismic wave velocity, and form a dataset X = {x1, x2, ..., x} to be classified. n}

[0059] (2) Initialize the dataset to be classified X = {x1, x2, ..., x} n The cluster centers c1, c2, ..., c of} k .

[0060] (3) The dataset to be classified is X = {x1, x2, ..., x...} n The expression for class allocation is as follows:

[0061]

[0062] Among them, C j Represents the j-th family; c j This indicates the updated class center.

[0063] (4) Calculate the average value of the data points contained in any family, and use this average value as the updated family center c. j Its expression is:

[0064]

[0065] Where, Nj This represents the number of data points within the j-th class.

[0066] (5) Construct the clustering criterion function, the expression of which is:

[0067]

[0068] In this embodiment, if the given conditions are met, the clustering result is output; otherwise, proceed to (2). Since this clustering algorithm only operates on the dataset (velocity, resistivity, etc.), the K-means clustering algorithm can be extended to image segmentation of three-dimensional resistivity or velocity models to obtain the corresponding strata or anomalous geological bodies.

[0069] In this embodiment, the K-means clustering algorithm requires a pre-defined number of clusters and initial cluster centers. The reliability of the given number of clusters and initial cluster centers is crucial for obtaining accurate clustering results. The method for determining the number of clusters and initial cluster centers is given below:

[0070] Number of clusters: For a given study area, the number of clusters can be determined by combining a single electromagnetic or seismic inversion result with existing prior information such as geology, outcrops, drilling, and well logging.

[0071] Initial cluster centers: Since the K-means clustering method is applied during the inversion iteration process, the initial cluster centers given by the K-means clustering method are generally different each time. First, the joint inversion results of N iterations (generally N≤5) (without adaptive correction) are used as the initial cluster centers for the first application of K-means clustering. Then, the initial cluster centers for subsequent applications of K-means clustering are determined based on the statistics of the model update amount in subsequent inversions.

[0072] In this embodiment, after clustering the resistivity and velocity models using K-means clustering, regression analysis is employed to adjust the resistivity or velocity values. In this embodiment, the regression analysis uses one of the following: linear regression, logarithmic regression, exponential regression, or quadratic regression.

[0073] (1) Linear Regression Analysis

[0074] In the joint electromagnetic and seismic inversion, linear regression analysis specifically involves: a series of attribute value points (v) corresponding to a geological body. i ,ρ i (i = 1, 2, ..., n), v i ρ i Given the velocity and conductivity of the seismic wave at point i, find a straight line:

[0075] ρ = k·v + b.

[0076] Here, the goal is to ensure that as many attribute values ​​as possible fall on or are as close as possible to the regression line. In the least squares sense, the specific forms of k and b are as follows:

[0077]

[0078]

[0079] (2) Log-regression analysis:

[0080] Logarithmic regression analysis is a non-linear regression analysis technique that can transform logarithmic regression analysis into linear regression analysis. The specific form of logarithmic regression analysis is:

[0081] ρ=a ln v+b

[0082] Where ρ and v are resistivity and velocity, respectively, and a and b are coefficients. Let but:

[0083]

[0084] The above equation is a linear regression analysis, and then the solution is obtained according to the linear regression analysis introduced earlier.

[0085] (3) Exponential regression analysis:

[0086] Exponential regression analysis is a non-linear regression analysis technique, and it can also be transformed into linear regression analysis. The specific form of exponential regression analysis is as follows:

[0087] ρ=ae bv

[0088] Where ρ and v are resistivity and seismic wave velocity, respectively, and a and b are coefficients.

[0089] Taking the logarithm of both sides of the above equation, we get:

[0090] lnρ=ln a+bv

[0091] make but

[0092]

[0093] The above equation is a linear regression analysis, and then the solution is obtained according to the linear regression analysis introduced earlier.

[0094] (4) Quadratic regression analysis:

[0095] The specific form of quadratic regression analysis is as follows:

[0096] ρ = av 2 +bv+c

[0097] Where ρ and v represent resistivity and seismic wave velocity, respectively, and a, b, and c are coefficients. (v...) This corresponds to a series of attribute values ​​(v...) for a specific geological body. i ,ρ i (i = 1, 2, ..., n), v i ρ i Let the velocity and conductivity at point i be denoted as i and i', respectively. Then we have:

[0098]

[0099] Solving the system of linear equations yields the coefficients a, b, and c.

[0100] The third step is to use the second property model to modify the first property model and perform a joint inversion.

[0101] The correction process is as follows:

[0102] (1) The structural boundary of the first physical model is obtained by using the second physical property model and K-means clustering analysis;

[0103] (2) Regression analysis was used to adjust the predicted values ​​of resistivity or seismic wave velocity in the first physical model;

[0104] (3) Substitute into the joint inversion process and determine whether the termination condition is met. If not, continue to perform K-means clustering analysis and regression analysis until the condition is met.

[0105] During the joint inversion process, a strong coupling can be achieved between the 2D seismic line and the resistivity and velocity within a certain range of the seismic line. The strength of the coupling is determined by the cross-gradient weighting matrix W. t Controlled. The cross-gradient weighting matrix W is determined based on the inverse distance weighting concept. t The different planar distances between the subdivided grid and the seismic line section are used to assign corresponding weight factors to the subdivided grid. If the subdivided grid is closer to the seismic line section, it is assigned a larger weight factor. If the subdivided grid is farther from the seismic line section, it is assigned a smaller weight factor.

[0106] In this embodiment, the weighting matrix W t The specific process is as follows:

[0107] First, the meshes crossed by the seismic survey line (including meshes at different depths below the seismic survey line) are marked as 1, and the remaining meshes are marked as 0; then, the maximum number of meshes N on the plane at the distance from the seismic survey line is given. max If the i-th subdivision grid is taken as the center, and N is the center, max If there are meshes marked as 1 within a plane of radius , then the i-th mesh is assigned a weight factor W.t,i Its expression is:

[0108]

[0109] Where J represents the number of meshes, J≥1 means that there may be multiple meshes marked as 1 within this plane; dn j Let represent the distance from the i-th mesh to the j-th mesh marked as 1, and Δx j Δy represents the number of grids that differ in the x-direction from the j-th grid (marked as 1) within the plane containing the i-th grid; j The number of grids that differ in the y-direction between the i-th grid and the j-th grid (marked as 1) within the plane containing the i-th grid.

[0110] The fourth step is to obtain the model parameters containing resistivity, polarizability, and seismic wave velocity after inversion, and to predict reservoir fluid based on the model parameters.

[0111] In this embodiment, appropriate seismic exploration techniques can be selected based on geological conditions to obtain seismic elastic property data. Then, electromagnetic exploration data, resistivity logging, and polarizability logging techniques are used to obtain resistivity and polarizability data. Finally, these three types of data are jointly analyzed to obtain the reservoir fluid detection results.

[0112] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any changes made based on the design principles of the present invention, or any non-creative modifications made thereon, shall fall within the scope of protection of the present invention.

Claims

1. A reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability, characterized in that, Includes the following steps: Seismic and electromagnetic detection data of the area to be detected and identified are acquired, and resistivity, polarizability and seismic wave velocity are extracted to obtain the first physical property model. Based on well logging data containing resistivity, polarizability, and seismic wave velocity, and combined with rock physics experimental analysis, a second physical property model was obtained using K-means clustering analysis and regression analysis. The first physical property model is modified using the second physical property model, and a joint inversion is performed. The model parameters containing resistivity, polarizability, and velocity are obtained after inversion, and reservoir fluid prediction is performed based on the model parameters.

2. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability as described in claim 1, characterized in that, The K-means clustering analysis includes the following steps: Acquire well logging data containing resistivity, polarizability, and seismic wave velocity, and construct a dataset X = {x1, x2, ..., x} to be classified. n }; where n∈k; where k represents the number of clusters; Initialize the dataset to be classified as X = {x1, x2, ..., x...} n The cluster centers c1, c2, ..., c of} k ; The dataset to be classified is X = {x1, x2, ..., x...} n The expression for class allocation is as follows: Where x represents the dataset to be classified, X = {x1, x2, ..., x} n Data within}; C j Represents the j-th family; c j Indicates the updated class center; Calculate the average value of the data points contained in any family, and use this average value as the updated family center c. j Its expression is: Where, N j This represents the number of data points within the j-th class; The clustering criterion function is constructed as follows:

3. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability as described in claim 1 or 2, characterized in that, K-means clustering analysis was used to cluster resistivity and seismic wave velocity, and regression analysis was used to adjust the resistivity or velocity model.

4. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability as described in claim 3, characterized in that, The regression analysis used is one of the following: linear regression analysis, logarithmic regression analysis, exponential regression analysis, or quadratic regression analysis.

5. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability according to claim 4, characterized in that, Also includes: In the joint inversion, the weighting factor W of the mesh is applied based on the distance of the seismic survey lines from the seismic exploration to the cross section. t,i Assignment, its expression is: Where J represents the number of meshes; dn j Let represent the distance from the i-th mesh to the j-th mesh marked as 1, and Δx j Δy represents the number of grids that differ in the x-direction from the j-th grid (marked as 1) within the plane containing the i-th grid; j The number of grids that differ in the y-direction between the i-th grid and the j-th grid (marked as 1) within the plane containing the i-th grid.

6. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability according to claim 1, characterized in that, The modification of the first physical property model using the second physical property model includes the following steps: The structural boundary of the first physical model is obtained using the second physical property model and K-means cluster analysis; Regression analysis was used to adjust the predicted values ​​of resistivity or seismic wave velocity in the first physical model; Substitute the data into the joint inversion process and determine whether the termination condition is met. If not, continue with K-means clustering analysis and regression analysis until the condition is met.

7. The reservoir fluid identification method based on the combined seismic elastic properties, resistivity, and polarizability according to claim 1, characterized in that, Seismic and electromagnetic detection data of the area to be detected and identified are obtained, and a cross-gradient function is constructed, the expression of which is: In the formula, τ(x,y,z) is the cross gradient function. The gradient of the resistivity model. This is the gradient of the velocity model.

Images