A method for predicting sensitive elastic parameters of a carbonate reservoir

Abstract

The application discloses a kind of prediction methods of carbonate reservoir sensitive elastic parameters, the prediction method described, comprising the following steps: S1, using K-nearest neighbor method, calculate lithology mineral content curve;S2, using neural network method, estimate total porosity curve;S3, optimize Xu-Payne modeling method, predict shear wave curve and S4, sensitive elastic parameter calculation optimization.This application realizes the effective prediction of shear wave curve and the reasonable optimization of elastic parameter under the condition of the absence of lithology content curve and porosity curve in deep carbonate formation, with low cost and strong practicability.Meanwhile, the application of the present application can find out the sensitive elastic parameters of deep carbonate formation reservoir, and provide reliable basis for seismic prestack elastic parameter inversion and quantitative interpretation.

Description

Technical Field

[0001] This invention belongs to the field of oil and gas exploration technology in ultra-deep carbonate rock exploration areas. Specifically, it relates to a method for predicting sensitive elastic parameters of carbonate rock reservoirs. Background Technology

[0002] Deep to ultra-deep marine carbonate rocks hold abundant oil and gas resources, making them an important area for oil and gas exploration. Different rocks, and even the same rock containing different fluids, exhibit varying elastic properties. Generally, parameters reflecting the elastic properties of rocks, such as P-wave and S-wave velocities and impedances, Lamé parameters, Young's modulus, and their combinations, can be referred to as elastic parameters. Oil and gas explorers can use the characteristics of these elastic parameter values ​​to predict the distribution of favorable reservoirs, deploy drilling sites, and estimate reserve size.

[0003] Currently, commonly used reservoir exploration technologies include well logging reservoir evaluation technology and shear wave estimation technology. Well logging reservoir evaluation uses conventional well logging curves to conduct a comprehensive and quantitative assessment of the reservoir, evaluating aspects such as diagenesis, reservoir heterogeneity, interlayers, and reservoir microstructure. Among these, the prediction of reservoir lithology and porosity is a crucial component of well logging reservoir evaluation.

[0004] Currently, lithological mineral content curve prediction falls into two categories. The first is the theoretical (empirical) formula method, which uses existing conventional logging, formation element logging, and relevant theoretical (empirical) formulas to calculate mineral content. Commonly used methods include predicting clay content using natural gamma ray and calculating mineral content using formation element logging. The second is the data-driven method, which uses methods such as multiple regression, neural networks, and machine learning to establish a complex nonlinear relationship between the predicted and measured curves, thereby optimizing the prediction of multiple mineral contents. This type of method is suitable for more complex carbonate rock formations.

[0005] In practice, well logging reservoir evaluation is not linked to the selection of sensitive elastic parameters or seismic inversion, which means that formation lithology content and porosity curves cannot be directly applied to reservoir prediction.

[0006] Shear wave logging data is primarily used for rock physics analysis, sensitive elastic parameter optimization, AVO (Active Velocity Detection), and pre-stack inversion. Combining P-wave and S-wave velocity information can effectively improve reservoir prediction and fluid detection accuracy. However, due to the differences between dipole acoustic logging instruments and conventional logging instruments, additional costs are incurred, resulting in many wells lacking S-wave data or having only moderately high-quality S-wave logging data, which hinders sensitive elastic parameter optimization and pre-stack inversion. S-wave estimation aims to obtain more and more accurate S-wave data, improving the reliability of sensitive elastic parameter optimization and the accuracy of pre-stack reservoir prediction and hydrocarbon detection.

[0007] Generally, shear wave prediction methods can be categorized into empirical formula methods, multivariate fitting methods, and rock physical modeling methods. Empirical formula methods predict shear waves by fitting the relationship between P-wave and S-wave velocities, such as the Greenberg-Castagna multivariate relationship method for pure and mixed lithologies. Multivariate fitting methods are essentially extensions of empirical formulas; for example, multivariate linear fitting incorporates other logging curves such as density, gamma, and resistivity during the fitting process. Rock physical modeling methods calculate shear wave velocities by constructing a rock skeleton model and inputting fluid parameters into the rock physical model. In the ultra-deep carbonate rock field, the rock properties are completely different from those of conventional clastic formations, and the shape, combination, and connectivity of pore spaces are more complex than in conventional clastic formations.

[0008] In the scientific research field, there is considerable discussion on the theory and optimization of carbonate rock physical modeling, with the Xu-Payne carbonate rock physical model being a classic example. However, in the production field, the use of rock physical models often presents certain problems. For instance, the complex lithology and porosity of actual strata often cannot be simplified into equivalent pure lithologies or combinations of two or three simple lithologies. In such cases, the shear wave estimation using rock physical modeling often differs significantly from the measured values. Furthermore, for ultra-deep strata, the influence of the formation pressure system on rock properties cannot be accurately characterized by a fixed model. Summary of the Invention

[0009] To address the above technical problems, this invention provides a method for predicting sensitive elastic parameters of carbonate reservoirs. Based on the prediction of lithological mineral content curves and porosity curves, this method uses a carbonate rock physics model to predict shear wave curves, thereby realizing the prediction of sensitive elastic parameters of ultra-deep carbonate reservoirs and providing technical support for oil and gas exploration in ultra-deep carbonate exploration areas.

[0010] To achieve the above objectives, the present invention adopts the following technical solution:

[0011] A method for predicting sensitive elastic parameters of carbonate reservoirs includes the following steps:

[0012] S1. Based on the conventional logging data of the designated work area, the lithological mineral content curve is calculated using the K-nearest neighbor method.

[0013] S2. Based on the well logging porosity interpretation results table within the designated work area, the total porosity curve is estimated using the neural network method.

[0014] S3. Based on the lithological mineral content curve from step S1 and the total porosity curve from step S2, the optimized Xu-Payne modeling method is used to predict and obtain the shear wave velocity curve, the optimized longitudinal wave velocity curve, and the density curve.

[0015] S4. Using the transverse wave velocity curve, optimized longitudinal wave velocity curve and density curve obtained in step S3, the sensitive elastic parameters are calculated and optimized.

[0016] Preferably, the method for calculating the lithological mineral content curve in step S1 includes the following steps:

[0017] S11. Estimated mudstone content: Based on the natural gamma logging curve (GR curve), preset the dry mudstone density, wet mudstone density, 0% mudstone GR value and 100% mudstone GR value to obtain the initial mudstone content curve;

[0018] S12. Using the initial clay content curve, conduct K-nearest neighbor machine learning to establish learning experience;

[0019] S13. Based on learning experience, make predictions to obtain the initial limestone mineral content curve and the initial dolomite mineral content curve.

[0020] S14. Normalize the initial mudstone content curve obtained in step S11, the initial limestone mineral content curve predicted in step S13, and the initial dolomite mineral content curve to establish the lithological mineral content curve.

[0021] More preferably, the specific method of K-nearest neighbor machine learning in step S12 is as follows:

[0022] Six curves—sonic, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron—with strata containing argillaceous and dolomitic minerals from the initial argillaceous content curve were selected as learning materials. K-nearest neighbor machine learning was performed on the argillaceous mineral content curve and the dolomite mineral content curve with strata containing argillaceous and dolomite minerals, respectively, to predict the argillaceous and dolomite mineral content curves. This aimed to achieve a high degree of agreement between the learning results and the lithological curves interpreted from well logging, thereby establishing learning experience.

[0023] Preferably, the method for estimating the total porosity curve in step S2 includes the following steps:

[0024] S21. Based on the well logging interpretation results table, calibrate the calculated total porosity value of the interpretation well section;

[0025] S22. Using a neural network calculation method, six curves—sound wave, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron—within the target layer are used as input attributes. The total porosity within the layer is used as the objective function to establish a nonlinear neural network attribute training set and obtain the optimized total porosity curve.

[0026] Preferably, the method for predicting the shear wave curve in step S3 includes the following steps:

[0027] S31, Water saturation setting;

[0028] S32. Establish a model using lithological mineral content curves, total porosity curves, and water saturation values;

[0029] S33. Given the initial porosity of each lithological mineral, and taking the measured longitudinal wave curve as the target, use the Xu-Payne modeling method to iteratively optimize the lithological mineral model and the initial porosity of each lithological mineral, predict the transverse wave velocity curve, and simultaneously optimize the longitudinal wave velocity curve and density curve.

[0030] Preferably, the model described in step S2 is a rock skeleton model and a fluid model.

[0031] More preferably, the method for establishing the equivalent rock and mineral model is as follows: using the lithological mineral content curve, given the bulk modulus, shear modulus and density of the rock and mineral, an equivalent rock and mineral model is established according to the VHR method.

[0032] More preferably, the method for establishing the fluid model is as follows: using an equivalent rock and mineral model and a total porosity curve, the fluid model is established according to the Gassmann method.

[0033] Preferably, the method for calculating the sensitive elastic parameter in step S4 includes the following steps:

[0034] S41. Using the optimized longitudinal wave velocity curve, density curve, and estimated transverse wave velocity curve, calculate the curves of each elastic parameter.

[0035] S42. For seismic reservoir prediction, optimize well logging interpretation conclusions based on porosity curves;

[0036] S43. Use the two-phase cross-section analysis method to determine the sensitive elastic parameters.

[0037] This invention also provides the application of the above-mentioned prediction method in the inversion and / or quantitative interpretation of pre-stack elastic parameters of earthquakes.

[0038] The beneficial effects of this invention are as follows:

[0039] This invention enables effective prediction of shear waves and rational optimization of elastic parameters in deep carbonate formations when lithology and porosity curves are missing. It is low-cost and highly practical. Furthermore, this invention can identify sensitive elastic parameters of deep carbonate reservoirs, providing a reliable basis for the inversion and quantitative interpretation of pre-stack elastic parameters during seismic events. Attached Figure Description

[0040] Figure 1 A flowchart illustrating a method for predicting sensitive elastic parameters of carbonate reservoirs provided for the implementation of this invention.

[0041] Figure 2 This is a comparison chart of the initial gray matter content and dolomite content curves predicted by the K-nearest neighbor method and the neural network method in the learning section of well A.

[0042] Figure 3 The graph shows the lithological mineral content curves for Well A, Q group, calculated using machine learning.

[0043] Figure 4 The Xu-Payne modeling method was optimized for predicting shear wave curves, as well as optimizing P-wave curves and velocity curves for well A, Q group.

[0044] Figure 5 This is a two-phase cross-section analysis diagram of the elastic parameters of well A. Detailed Implementation

[0045] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention.

[0046] Before further describing specific embodiments of the present invention, it should be understood that the scope of protection of the present invention is not limited to the specific embodiments described below; it should also be understood that the terminology used in the embodiments of the present invention is for describing specific embodiments and not for limiting the scope of protection of the present invention.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0048] A method for predicting sensitive elastic parameters of carbonate reservoirs includes the following steps:

[0049] S1. Based on the conventional well logging data of the designated work area, calculate the lithological mineral content curve using the K-nearest neighbor method:

[0050] The method for calculating the lithological mineral content curve includes the following steps:

[0051] S11. Estimated mudstone content: Based on the natural gamma logging curve (GR curve), preset the dry mudstone density, wet mudstone density, 0% mudstone GR value and 100% mudstone GR value to obtain the initial mudstone content curve;

[0052] S12. Using the initial clay content curve, conduct K-nearest neighbor machine learning to establish learning experience;

[0053] The specific method of the K-nearest neighbor machine learning is as follows:

[0054] Six curves—sonic, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron—with strata containing argillaceous and dolomitic minerals from the initial argillaceous content curve were selected as learning materials. K-nearest neighbor machine learning was performed on the argillaceous mineral content curve and the dolomite mineral content curve with strata containing argillaceous and dolomite minerals, respectively, to predict the argillaceous and dolomite mineral content curves. This aimed to achieve a high degree of agreement between the learning results and the lithological curves interpreted from well logging, thereby establishing learning experience.

[0055] S13. Based on learning experience, predict the initial limestone mineral content curve and the initial dolomite mineral content curve;

[0056] S14. Normalize the initial mudstone content curve obtained in step S11, the initial limestone mineral content curve predicted in step S13, and the initial dolomite mineral content curve to establish the lithological mineral content curve.

[0057] S2. Based on the well logging porosity interpretation results table within the designated work area, the total porosity curve is estimated using the neural network method.

[0058] The method for estimating the total porosity curve includes the following steps:

[0059] S21. Based on the well logging porosity interpretation results table, calibrate the calculated porosity values ​​of the interpretation well section;

[0060] S22. Using a neural network calculation method, six curves—sound wave, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron—within the target layer are used as input attributes. The total porosity within the layer is used as the objective function to establish a nonlinear neural network attribute training set and obtain the optimized total porosity curve result.

[0061] S3. Based on the lithological mineral content curve from step S1 and the total porosity curve from step S2, the optimized Xu-Payne modeling method is used to predict and obtain the shear wave velocity curve, the optimized longitudinal wave velocity curve, and the density curve.

[0062] The prediction method includes the following steps:

[0063] S31, Water saturation setting;

[0064] S32. Using the lithological mineral content curve, total porosity curve, and water saturation value, establish an equivalent rock and mineral model and a fluid model;

[0065] The method for establishing the equivalent rock and mineral model is as follows: using the lithological mineral content curve, given the bulk modulus, shear modulus and density of the rock and mineral, the equivalent rock and mineral model is established according to the VHR method.

[0066] The method for establishing the fluid model is as follows: using an equivalent rock and mineral model and a total porosity curve, the fluid model is established according to the Gassmann method.

[0067] S33. Given the initial porosity of each lithological mineral, and taking the measured longitudinal wave curve as the target, use the Xu-Payne modeling method to iteratively optimize the lithological mineral model and the initial porosity of each lithological mineral, predict the transverse wave velocity curve, and simultaneously optimize the longitudinal wave velocity curve and density curve.

[0068] S4. Using the transverse wave velocity curve, optimized longitudinal wave velocity curve and density curve obtained in step S3, the sensitive elastic parameters are calculated and optimized.

[0069] The preferred method for calculating the sensitive elastic parameter includes the following steps:

[0070] S41. Using the optimized longitudinal wave velocity curve, density curve, and estimated transverse wave velocity curve, calculate the curves of each elastic parameter.

[0071] S42. For seismic reservoir prediction, optimize well logging interpretation conclusions based on total porosity curve;

[0072] S43. Use the two-phase cross-section analysis method to determine the sensitive elastic parameters.

[0073] This invention also provides the application of the above-mentioned prediction method in the inversion and / or quantitative interpretation of pre-stack elastic parameters of earthquakes.

[0074] This invention enables effective prediction of shear waves and rational optimization of elastic parameters in deep carbonate formations when lithological and mineral content curves and porosity curves are missing. It is low-cost and highly practical. Furthermore, this invention can identify sensitive elastic parameters of deep carbonate reservoirs, providing a reliable basis for the inversion and quantitative interpretation of pre-stack elastic parameters during seismic events.

[0075] To better understand the structure and function of this invention, the following description, in conjunction with the accompanying drawings, takes a drilling well in the Tarim Basin that encountered the target formation Q of the deep Sinian system (hereinafter referred to as Well A) as an example to further illustrate the method for predicting sensitive elastic parameters of carbonate reservoirs according to this invention.

[0076] Well A revealed a porous reservoir (gas layer) in the upper part of the Q group of the Sinian system, containing marl-bearing dolomite. After collecting the logging curves and interpretation results from Well A, three major problems were found: First, the Class II / III reservoirs and the uninterpreted sections were not distinguishable in terms of velocity, density, and impedance on the original logging curves; second, the lithological content curves for the Q group were missing; and third, the porosity curves for the Q group were missing, with only porosity values ​​from the interpreted logging sections available. These three problems hinder the analysis of reservoir-sensitive parameters and quantitative inversion prediction and interpretation. Based on these findings, this invention is provided.

[0077] Figure 1 The main flowchart of the method for predicting sensitive elastic parameters of carbonate reservoirs provided for the implementation of the present invention is shown. The method includes the following steps:

[0078] Step S1: Calculate the lithological mineral content curve using the K-nearest neighbor method.

[0079] First, the mudstone content was estimated using the density method. Using the GR curve, the dry mudstone density was set to 2.68 g / cm³. 3 The density of wet mudstone is 2.35 g / cm³. 3 The initial mudstone content curves were obtained by setting the GR value of 0% mudstone to 25 API and the GR value of 100% mudstone to 350 API. These curves were then overlaid with the existing mudstone content curves of the overlying Y group of the Q group, showing good agreement. Next, four segments from the S, W, R, and Y groups (overlying the Q group and with similar sedimentary environments) with relatively small curve fluctuations, stable value ranges, and curves containing calcareous minerals (calcium carbonate) and dolomitic minerals (calcium magnesium carbonate) were selected as learning segments. The sonic, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron content curves of these four segments were used as learning materials. The limestone and dolomite mineral content curves of these segments were used as learning targets. K-nearest neighbor machine learning calculations were performed to predict the limestone and dolomite mineral content curves, aiming to achieve a high degree of agreement between the learning results and the lithological curves interpreted from well logging, thus establishing learning experience. Figure 2 As shown.

[0080] Depend on Figure 2 It is evident that, compared to the neural network method, the dolomite and argillaceous content curves calculated by the K-nearest neighbor method correspond more consistently with the target curves. Based on learning experience, the initial limestone mineral content curves and initial dolomite mineral content curves for group Q are predicted. Simultaneously, the three lithological content curves—initial argillaceous content, initial limestone mineral content, and initial dolomite mineral content—are normalized to establish the lithological mineral content curve for the target stratigraphic interval, as shown below. Figure 3 As shown.

[0081] Step S2: Estimate the total porosity curve using a neural network method

[0082] First, well logging interpretation results were collected, and the calculated porosity values ​​for the interpretation well sections were calibrated. From the interpretation results of Well A, the porosity values ​​of reservoir sections Y1 and Q4 were extracted. The porosity of the mudstone section in Y1 was 18.2%, and the porosity of the reservoir section in Q ranged from 1.4% to 4.1%, which were comprehensively interpreted as poorly porosity gas-bearing layers (Class II) and Class III gas-bearing layers. Second, using six curves—sonic wave, density, deep lateral resistivity, shallow lateral resistivity, natural gamma ray, and neutron—as input attributes, and with the porosity value within the interpretation sections as the objective function, a nonlinear neural network attribute training set was established to obtain the optimized porosity curve results. The results show that the total porosity curve agrees well with the interpreted porosity values ​​in the four interpretation reservoir sections.

[0083] Step S3: Optimize the Xu-Payne modeling method to predict shear wave curves

[0084] First, the gas saturation of this reservoir section is set to 100% (equivalent fluid model). Second, using the lithological mineral content curve, given the bulk modulus, shear modulus, and density of the rock minerals, an equivalent rock mineral model is established using the VHR method. Using the rock mineral model and the total porosity curve, a rock skeleton model is established using the Gassmann method. Third, initial porosity ratios are set (0.3 for dry clay, 0.1 for calcareous material, and 0.15 for dolomitic material). Using the measured P-wave curve as the target, the mineral content curve and the rock mineral porosity ratio are iteratively updated using the Xu-Payne method to obtain optimized mineral content curves and rock mineral porosity ratios. Simultaneously, the shear wave velocity curve, optimized P-wave velocity curve, and density curve are calculated. Figure 4 As shown.

[0085] Step S4: Calculation and Optimization of Sensitive Elastic Parameters

[0086] First, using the optimized P-wave velocity and density curves, as well as the estimated S-wave velocity curve, the curves for each elastic parameter are calculated. Second, for seismic reservoir prediction, the well logging interpretation conclusions are optimized based on the porosity curve. Overall, the target interval of Well A has poor reservoir properties and poorly developed matrix pores, but the drilled and exposed continuous reservoir thickness is large, with a wide and stable lateral distribution. The original reservoir interpretation scheme relied on well logging curves, but for seismic reservoir prediction, the well logging interpretation conclusions need to be optimized and determined based on porosity and elastic parameters.

[0087] The porosity of Class I reservoirs is ≥4.5%, Class II reservoirs are 2.5-4.5%, Class III reservoirs are 1.5-2.5%, and Class IV reservoirs are ≤1.5%. Finally, two-phase cross-plot analysis is used, such as... Figure 5As shown, sensitive elastic parameters were determined. The results show that Class I and II reservoirs are more sensitive to velocity, density, and P-wave / S-wave velocity ratio compared to Class III and non-reservoir reservoirs, and exhibit characteristics of low velocity, low density, low impedance, and low velocity ratio.

[0088] In summary, this invention achieves effective prediction of shear waves and reasonable optimization of elastic parameters in deep carbonate formations when lithology and porosity curves are missing, with low cost and high practicality. Furthermore, this invention can identify sensitive elastic parameters of deep carbonate reservoirs, providing a reliable basis for the inversion and quantitative interpretation of pre-stack elastic parameters during seismic events.

[0089] The above description, in conjunction with specific embodiments, further illustrates the present invention. However, these embodiments are merely exemplary and do not constitute any limitation on the scope of the present invention. Those skilled in the art should understand that modifications or substitutions to the details and form of the technical solutions of the present invention can be made without departing from the spirit and scope of the invention, and all such modifications and substitutions fall within the protection scope of the present invention.

Claims

1. A method for predicting sensitive elastic parameters of carbonate reservoirs, characterized in that, Includes the following steps: S1. Based on the conventional logging data of the designated work area, the lithological mineral content curve is calculated using the K-nearest neighbor method. S2. Based on the well logging porosity interpretation results table within the designated work area, the total porosity curve is estimated using the neural network method. S3. Based on the lithological mineral content curve from step S1 and the total porosity curve from step S2, the optimized Xu-Payne modeling method is used to predict and obtain the shear wave velocity curve, the optimized longitudinal wave velocity curve, and the density curve. S4. Using the transverse wave velocity curve, optimized longitudinal wave velocity curve and density curve obtained in step S3, the sensitive elastic parameters are calculated and optimized. The calculation method for the lithological mineral content curve in step S1 includes the following steps: S11. Estimated mudstone content: Based on the natural gamma logging curve, preset the dry mudstone density, wet mudstone density, 0% mudstone GR value and 100% mudstone GR value to obtain the initial mudstone content curve; S12. Using the initial clay content curve, conduct K-nearest neighbor machine learning to establish learning experience; The specific method of the K-nearest neighbor machine learning is as follows: select six curves on the initial mudstone content curve containing calcareous and dolomitic mineral curve segments, namely, sonic, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron curves, as learning materials. Take the calcareous mineral content curve and the dolomite mineral content curve containing calcareous and dolomite mineral curve segments as learning targets, respectively, and perform K-nearest neighbor machine learning operation to predict the calcareous mineral content curve and the dolomite mineral content curve, so as to achieve a high degree of consistency between the learning results and the lithology curves interpreted by well logging, and establish learning experience. S13. Based on learning experience, make predictions to obtain the initial limestone mineral content curve and the initial dolomite mineral content curve. S14. Normalize the initial mudstone content curve obtained in step S11, the initial limestone mineral content curve obtained in step S13, and the initial dolomite mineral content curve to establish the lithological mineral content curve.

2. The prediction method according to claim 1, characterized in that, The method for estimating the total porosity curve in step S2 includes the following steps: S21. Based on the well logging porosity interpretation results table, calibrate the calculated total porosity value of the interpretation well section; S22. Using a neural network calculation method, six curves—sound wave, density, deep lateral resistivity, shallow lateral resistivity, natural gamma, and neutron—within the target layer are used as input attributes. The total porosity within the layer is used as the objective function to establish a nonlinear neural network attribute training set and obtain the optimized total porosity curve.

3. The prediction method according to claim 1, characterized in that, The method for predicting the shear wave curve in step S3 includes the following steps: S31, Water saturation setting; S32. Establish a model using lithological mineral content curves, total porosity curves, and water saturation values; S33. Given the initial porosity of each lithological mineral, and taking the measured longitudinal wave curve as the target, use the Xu-Payne modeling method to iteratively optimize the lithological mineral model and the initial porosity of each lithological mineral, predict the transverse wave velocity curve, and simultaneously optimize the longitudinal wave velocity curve and density curve.

4. The prediction method according to claim 3, characterized in that, The model mentioned in step S2 is an equivalent rock and mineral model and a fluid model.

5. The prediction method according to claim 4, characterized in that, The method for establishing the equivalent rock and mineral model is as follows: using the lithological mineral content curve, given the bulk modulus, shear modulus and density of the rock and mineral, the equivalent rock and mineral model is established according to the VHR method.

6. The prediction method according to claim 4, characterized in that, The method for establishing the fluid model is as follows: using an equivalent rock and mineral model and a total porosity curve, the fluid model is established according to the Gassmann method.

7. The prediction method according to claim 1, characterized in that, The preferred method for calculating the sensitive elastic parameter in step S4 includes the following steps: S41. Using the optimized longitudinal wave velocity curve, density curve, and estimated transverse wave velocity curve, calculate the curves of each elastic parameter. S42. For seismic reservoir prediction, optimize well logging interpretation conclusions based on total porosity curve; S43. Use the two-phase cross-section analysis method to determine the sensitive elastic parameters.

8. The application of the prediction method according to any one of claims 1-7 in the inversion and / or quantitative interpretation of pre-stack elastic parameters of earthquakes.

Images