A method for constructing a low-frequency model of a tight heterogeneous reservoir in a fluvial facies

Abstract

The application relates to a method for constructing a low-frequency model of a fluvial facies dense heterogeneous reservoir, and relates to the technical field of geophysical reservoir prediction and seismic inversion, and comprises the following steps: dividing a fluvial facies dense heterogeneous reservoir into lithofacies types and counting the proportion of each lithofacies; obtaining a P-wave impedance, a S-wave impedance and a density parameter inversion three-dimensional data body after first-time inversion; obtaining a P-wave impedance, a S-wave impedance and a density parameter low-frequency model after last-time iteration, and a P-wave impedance, a S-wave impedance and a density parameter three-dimensional data body after prestack inversion of last-time iteration; calculating a P-S wave velocity ratio data body to obtain a fluvial facies dense sandstone reservoir distribution rule. The method for constructing a low-frequency model of a fluvial facies dense heterogeneous reservoir disclosed by the application provides an accurate low-frequency model for reservoir prediction, and the prestack inversion result can better depict channel distribution characteristics.

Description

Technical Field

[0001] This invention relates to the field of geophysical reservoir prediction and seismic inversion technology, specifically to a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs. Background Technology

[0002] The Permian Shihezi Formation in the Ordos Basin is characterized by fluvial-facies tight sandstone reservoirs with rapid lateral variations and strong heterogeneity. Rock physical analysis reveals that the sandstone exhibits a low P-wave / S-wave velocity ratio, while the mudstone displays a high P-wave / S-wave velocity ratio. Pre-stack inversion calculations of the P-wave / S-wave velocity ratio parameters can effectively identify tight sandstone reservoirs. However, since the bandwidth of the pre-stack inversion results is limited, typically ranging from tens of hertz to hundreds of hertz, supplementary low-frequency information is needed to obtain absolute impedance values, enabling the characterization of the vertical and lateral boundaries of subsurface geological bodies. Therefore, establishing a low-frequency model that reflects the variation patterns of fluvial-facies heterogeneous reservoirs is crucial for pre-stack inversion and tight sandstone reservoir prediction.

[0003] Conventional low-frequency modeling methods mainly fall into two categories: one uses well logging curve interpolation and extrapolation between wells. Low-frequency models built using this method are prone to the "bull's-eye" problem (circling around well points), which does not conform to the geological characteristics of heterogeneous reservoirs in fluvial facies. The other category uses regional stratigraphic compaction trends to build low-frequency models. This method typically only considers the compaction trend of mudstone, failing to account for the differences in compaction trends between different lithofacies (sandstone, mudstone, etc.), and can only supplement very low-frequency information (less than 2Hz), resulting in low vertical and horizontal resolution and poor spatial identification ability of geological bodies. Ideally, low-frequency modeling should consider the compaction trend patterns of different lithofacies, establishing low-frequency models separately for each lithofacies.

[0004] In summary, conventional low-frequency modeling methods for fluvial facies tight heterogeneous reservoirs suffer from problems such as low resolution, poor geological body identification ability, and "bull's-eye" errors. Summary of the Invention

[0005] To address the aforementioned problems, the purpose of this invention is to provide a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs, which solves the problems of low model resolution, poor geological body identification ability, and "bull's-eye" errors in conventional methods, and provides an accurate low-frequency model for the prediction of tight sandstone reservoirs.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention discloses a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs, including classifying fluvial facies tight heterogeneous reservoirs into lithofacies types and calculating the proportion of each lithofacies.

[0008] Compaction trend curves were analyzed based on the lithofacies types of fluvial tight heterogeneous reservoirs to obtain three-dimensional compaction trend volumes of P-wave impedance, S-wave impedance, and density parameters for each lithofacies. These results were then weighted proportionally with the proportion of the corresponding lithofacies type to obtain initial low-frequency models of P-wave impedance, S-wave impedance, and density parameters. Pre-stack inversion was then performed on these models to obtain three-dimensional inversion data volumes of P-wave impedance, S-wave impedance, and density parameters after the first inversion.

[0009] Based on the P-wave impedance and S-wave impedance obtained from the first inversion, cross-plot analysis and Bayesian discriminant analysis are performed to obtain the first three-dimensional volume of lithofacies probability. Based on the first three-dimensional volume of lithofacies probability, the low-frequency models of the initial P-wave impedance, S-wave impedance, and density parameters are updated. Using the updated low-frequency models of P-wave impedance, S-wave impedance, and density parameters as initial values, iterative calculations of pre-stack inversion, cross-plot analysis, and Bayesian discriminant analysis are performed until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached, and then the calculation terminates. The low-frequency models of P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the last iteration are obtained.

[0010] Based on the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the last iteration, the P-wave and S-wave velocity ratio data volume is calculated to obtain the distribution pattern of fluvial facies tight sandstone reservoirs.

[0011] Preferably, the step of classifying fluvial facies tight heterogeneous reservoirs into lithofacies types and calculating the proportion of each lithofacies includes the following steps:

[0012] Based on the mud content and porosity curves of the drilled wells in the study area, the mud content and porosity threshold values ​​of the fluvial facies tight heterogeneous reservoirs were obtained. Based on the mud content and porosity threshold values, the fluvial facies tight heterogeneous reservoirs were divided into mudstone facies, tight sandstone facies and high-porosity sandstone facies.

[0013] The proportions of mudstone facies, tight sandstone facies, and high-porosity sandstone facies in the drilled formations were calculated separately.

[0014] Preferably, obtaining the three-dimensional data volume of the longitudinal wave impedance, transverse wave impedance, and density parameters after the first inversion includes the following steps:

[0015] Based on the described lithofacies type, the compaction trend curves of the P-wave impedance, S-wave impedance and density parameters of each lithofacies are determined. Under the constraint of the stratigraphic position, a three-dimensional body of the compaction trend of the P-wave impedance, S-wave impedance and density parameters of each lithofacies is established.

[0016] Based on the proportion of each rock facies and the compaction trend of the longitudinal wave impedance, transverse wave impedance and density parameters of each rock facies in a three-dimensional volume, a low-frequency model of the initial longitudinal wave impedance, transverse wave impedance and density parameters is obtained by proportional weighting calculation.

[0017] The initial low-frequency model of longitudinal wave impedance, transverse wave impedance and density parameters is inverted before stacking to obtain the three-dimensional data volume of longitudinal wave impedance, transverse wave impedance and density parameters after the first inversion.

[0018] Preferably, the initial longitudinal wave impedance low-frequency resistance model is as follows:

[0019] Zp0=p1×Zp f1 +p2×Zp f2 +p3×Zp f3 (Equation 1)

[0020] In the formula, Zp0 represents the initial low-frequency model of the longitudinal wave impedance.

[0021] Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies.

[0022] Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies.

[0023] Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies.

[0024] p1 represents the proportion of mudstone facies.

[0025] p2 represents the proportion of tight sandstone facies.

[0026] p3 represents the proportion of high-porosity sandstone facies;

[0027] The initial low-frequency model of the transverse wave impedance:

[0028] Zs0=p1×Zs f1 +p2×Zs f2 +p3×Zs f3 (Equation 2)

[0029] In the formula, Zs0 represents the initial low-frequency model of transverse wave impedance.

[0030] Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies.

[0031] Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies.

[0032] Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies.

[0033] p1 represents the proportion of mudstone facies.

[0034] p2 represents the proportion of tight sandstone facies.

[0035] p3 represents the proportion of high-porosity sandstone facies;

[0036] The low-frequency model of the initial density parameters:

[0037] ρ0=p1×ρ f1 +p2×ρ f2 +p3×ρ f3 (Equation 3)

[0038] In the formula, ρ0 represents the initial density parameter low-frequency model.

[0039] ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters.

[0040] ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies.

[0041] ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters.

[0042] p1 represents the proportion of mudstone facies.

[0043] p2 represents the proportion of tight sandstone facies.

[0044] p3 represents the proportion of high-porosity sandstone facies.

[0045] Preferably, obtaining the low-frequency model of the P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters obtained by pre-stack inversion, includes the following steps:

[0046] Step C1: Calculate the P-wave and S-wave velocity ratio based on the P-wave impedance and S-wave impedance after the first inversion. Perform cross-intersection analysis based on the P-wave and S-wave velocity ratio and the P-wave impedance after the first inversion to obtain an interaction diagram. Perform Bayesian discriminant analysis on the obtained interaction diagram to calculate the lithofacies probability value at each sampling point location and obtain the first three-dimensional lithofacies probability volume.

[0047] Step C2: Based on the first probabilistic three-dimensional volume of lithofacies, update the low-frequency models of the initial P-wave impedance, S-wave impedance and density parameters to obtain the updated low-frequency models of P-wave impedance, S-wave impedance and density parameters.

[0048] Step C3: Using the updated low-frequency model of P-wave impedance, S-wave impedance, and density parameters as initial values, repeat the pre-stack inversion work to obtain the second inverted three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters; repeat the cross-plot analysis and Bayesian discriminant analysis to obtain the second lithofacies probability three-dimensional volume, completing one iteration;

[0049] Step C4: Repeat step C3 until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached, and then terminate the process to obtain the low-frequency model of the P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters obtained by pre-stack inversion.

[0050] Preferably, obtaining the updated low-frequency model of longitudinal wave impedance, transverse wave impedance, and density parameters includes the following steps:

[0051] The P-wave and S-wave velocity ratios are calculated based on the P-wave impedance and S-wave impedance after the first inversion. Cross-plot analysis is then performed based on the P-wave and S-wave velocity ratios and the P-wave impedance after the first inversion to obtain a cross-plot diagram.

[0052] Bayesian discriminant analysis was performed on the intersection graph to calculate the lithofacies probability value at each sampling point location, thus obtaining the first three-dimensional lithofacies probability volume.

[0053] The Bayesian discriminant analysis method involves calculating the lithofacies category f using Bayes' theorem. i The posterior probability p(f|d) is calculated using the following formula:

[0054]

[0055] In the formula, p(f|d) represents the lithofacies category f. i The posterior probability;

[0056] p(d|f) indicates that the sample point is of lithofacies f. i The prior probability of d corresponding to time;

[0057] p(f) represents the lithofacies type f i The probability of lithofacies type f is determined by statistical analysis of well logging data. i The prior probability of that lithofacies type, i.e., the proportion of that lithofacies type to all lithofacies types;

[0058] p(d) represents the scaling factor, which is a constant value in Bayesian discriminant analysis;

[0059] Among them, f i (i = 1, ..., N) represents N different lithofacies categories;

[0060] d represents the single-parameter or multi-parameter sample value observed from seismic attributes or well logging curve sample values.

[0061] Preferably, the updated low-frequency model of the longitudinal wave impedance is:

[0062] Zp=π(f1)×Zp f1 +π(f2)×Zp f2 +π(f3)×Zp f3 (Equation 5)

[0063] In the formula, Zp represents the updated low-frequency model of the longitudinal wave impedance.

[0064] Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies.

[0065] Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies.

[0066] Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies.

[0067] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0068] π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis.

[0069] π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria;

[0070] The updated low-frequency model of the shear wave impedance:

[0071] Zs=π(f1)×Zs f1 +π(f2)×Zs f2 +π(f3)×Zs f3 (Equation 6)

[0072] In the formula, Zs represents the updated low-frequency model of the transverse wave impedance.

[0073] Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies.

[0074] Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies.

[0075] Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies;

[0076] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0077] π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis.

[0078] π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria;

[0079] The low-frequency model of the updated density parameters:

[0080] ρ=π(f1)×ρ f1 +π(f2)×ρ f2 +π(f3)×ρ f3 (Equation 7)

[0081] In the formula, ρ represents the low-frequency model of the updated density parameters.

[0082] ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters.

[0083] ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies.

[0084] ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters.

[0085] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0086] π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis.

[0087] π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria.

[0088] Secondly, this invention discloses an apparatus for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir, comprising:

[0089] The first processing unit is used to classify fluvial facies tight heterogeneous reservoirs into lithofacies types and to calculate the proportion of each lithofacies.

[0090] The second processing unit is used to analyze the compaction trend curve based on the lithofacies type of the fluvial tight heterogeneous reservoir, obtain the three-dimensional compaction trend volume of the P-wave impedance, S-wave impedance and density parameters of each lithofacies, and calculate the initial low-frequency model of the P-wave impedance, S-wave impedance and density parameters by weighting the result with the proportion of the corresponding lithofacies according to the ratio. Then, it performs pre-stack inversion on the model to obtain the three-dimensional data volume of the inverted P-wave impedance, S-wave impedance and density parameters after the first inversion.

[0091] The third processing unit is used to perform cross-plot analysis and Bayesian discriminant analysis based on the P-wave impedance and S-wave impedance after the first inversion to obtain the first three-dimensional volume of lithofacies probability; based on the first three-dimensional volume of lithofacies probability, update the low-frequency model of the initial P-wave impedance, S-wave impedance and density parameters; using the updated low-frequency model of P-wave impedance, S-wave impedance and density parameters as the initial value, perform iterative calculations of pre-stack inversion, cross-plot analysis and Bayesian discriminant analysis until the P-wave impedance, S-wave impedance and density parameters involved in the iterative calculation match the logging curve or reach the set maximum number of iterations, and then terminate to obtain the low-frequency model of P-wave impedance, S-wave impedance and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters after the last iteration of pre-stack inversion;

[0092] The fourth processing unit is used to calculate the P-wave and S-wave velocity ratio data volume based on the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters obtained from the pre-stack inversion after the last iteration, and to obtain the distribution law of fluvial facies tight sandstone reservoirs.

[0093] Thirdly, the present invention also discloses a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the above-described method.

[0094] Fourthly, the present invention also discloses a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the steps of the above-described method.

[0095] Compared with the prior art, the beneficial effects of the present invention are:

[0096] This invention proposes a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs. First, based on the porosity and clay content curves of drilled wells in the study area, lithofacies are sequentially classified and their proportions are statistically analyzed. Then, compaction trend curve analysis, proportional weighted calculations, and pre-stack inversion are performed to obtain the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters after the first inversion. Next, iterative calculations using cross-plot analysis and Bayesian discriminant analysis are performed to obtain the low-frequency model of P-wave impedance, S-wave impedance, and density parameters after the final iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the final iteration. Finally, based on the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the final iteration, the P-wave and S-wave velocity ratio data volume is calculated to obtain the distribution pattern of fluvial facies tight sandstone reservoirs. This invention discloses a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs, providing an accurate low-frequency model for reservoir prediction. Its pre-stack inversion results can better characterize channel distribution features, and it has the following advantages:

[0097] (1) Based on conventional single mudstone compaction trend modeling, this invention considers the compaction trend differences between different rock facies, establishes low-frequency models for different rock facies, eliminates the influence of compaction trend differences on low-frequency modeling, and can effectively avoid the "bull's eye" phenomenon of well interpolation low-frequency modeling.

[0098] (2) This invention achieves an iterative cycle of pre-stack inversion and low-frequency modeling of compaction trends by introducing Bayesian discriminant analysis and lithofacies probability volume. The iterative low-frequency model can supplement the low-frequency components (2-10Hz) missing in conventional compaction trend modeling, thereby improving the vertical and horizontal resolution of the model;

[0099] (3) The low-frequency model constructed in this invention can reflect the sedimentary characteristics of fluvial reservoirs, improve the spatial identification of geological bodies, and improve the pre-stack inversion and characterization effect of fluvial heterogeneous reservoirs. Attached Figure Description

[0100] Figure 1 This is a flowchart illustrating the method for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir provided in Embodiment 1 of the present invention.

[0101] Figure 2 This is a schematic diagram of the classification results of mudstone facies, dense sandstone facies, and high-porosity sandstone facies provided in Embodiment 1 of the present invention;

[0102] Figure 3 This is a schematic diagram of the compaction trend curves of the longitudinal wave impedance, transverse wave impedance, and density parameters for each rock facies provided in Embodiment 1 of the present invention, wherein, Figure 3 (a) is mudstone facies. Figure 3 (b) is a dense sandstone facies. Figure 3 (c) is a high-porosity sandstone facies;

[0103] Figure 4 This is a schematic diagram illustrating the iterative effect of the low-frequency longitudinal wave impedance model provided in Embodiment 1 of the present invention, wherein... Figure 4 (a) is the initial longitudinal wave impedance low-frequency model. Figure 4 (b) is the low-frequency model of longitudinal wave impedance after the first iteration. Figure 4 (c) is the low-frequency model of longitudinal wave impedance after the second iteration. Figure 4 (d) is the final iteratively optimized low-frequency model of longitudinal wave impedance;

[0104] Figure 5 This is a planar schematic diagram of the P-wave and S-wave velocity ratio inversion results obtained using low-frequency modeling with well interpolation in a conventional method.

[0105] Figure 6 This is a planar schematic diagram of the inversion results of the P-wave and S-wave velocity ratio using the low-frequency modeling method disclosed in this invention. Detailed Implementation

[0106] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the invention and to fully convey the scope of the invention to those skilled in the art.

[0107] To address the problems of low model resolution, poor geological body identification, and "bull's-eye" errors in conventional methods, this invention discloses a method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs. First, based on the porosity and clay content curves of drilled wells in the study area, lithofacies are sequentially classified and their proportions are statistically analyzed. Then, compaction trend curve analysis, proportional weighted calculations, and pre-stack inversion are performed to obtain the first inversion's P-wave impedance, S-wave impedance, and density parameter inversion three-dimensional data volumes. Next, iterative calculations using cross-plot analysis and Bayesian discriminant analysis are performed to obtain the final iteration's P-wave impedance, S-wave impedance, and density parameter low-frequency model, as well as the final iteration's pre-stack inversion's P-wave impedance, S-wave impedance, and density parameter three-dimensional data volumes. Finally, based on the final iteration's pre-stack inversion's P-wave impedance, S-wave impedance, and density parameter three-dimensional data volumes, the P-wave and S-wave velocity ratio data volumes are calculated to obtain the distribution patterns of fluvial facies tight sandstone reservoirs, providing an accurate low-frequency model for tight sandstone reservoir prediction.

[0108] Example 1: A method for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir

[0109] Embodiment 1 of this invention provides a method for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir, taking a small-well exploration area on the eastern margin of the Ordos Basin as an example. Figure 1 As shown, the construction method includes the following steps:

[0110] Step A: Based on the porosity and clay content curves of the drilled wells in the study area, the fluvial facies tight heterogeneous reservoirs are classified into lithofacies types, and the proportion of each lithofacies is calculated. The lithofacies types include mudstone facies, tight sandstone facies, and high-porosity sandstone facies, including the following specific steps:

[0111] Step A1: Based on the mud content and porosity curves of the drilled wells in the study area, obtain the mud content and porosity threshold values ​​of the fluvial facies tight heterogeneous reservoirs. Based on the mud content and porosity threshold values, the fluvial facies tight heterogeneous reservoirs are divided into mudstone facies, tight sandstone facies, and high-porosity sandstone facies.

[0112] Specifically, a threshold value of mud content is defined as 'a'. Strata with mud content above the threshold value 'a' are classified as mudstone facies f1, and strata with mud content below the threshold value 'a' are classified as sandstone. In sandstone, a threshold value of porosity is defined as 'b'. Strata with porosity below the threshold value 'b' are classified as dense sandstone facies f2, and strata with porosity above the threshold value 'b' are classified as high-porosity sandstone facies f3.

[0113] Specifically, in this embodiment, the mud content threshold value 'a' is 0.5, and the porosity threshold value 'b' is 0.06, dividing the rock into three different lithofacies: mudstone facies, dense sandstone facies, and high-porosity sandstone facies. Figure 2 As shown.

[0114] Step A2: Calculate the proportions of mudstone facies, tight sandstone facies, and high-porosity sandstone facies in the drilled formations.

[0115] Where p1 represents the proportion of mudstone facies, p2 represents the proportion of tight sandstone facies, and p3 represents the proportion of high-porosity sandstone facies.

[0116] Specifically, in this embodiment, the proportion of mudstone facies is p1 = 0.67, the proportion of dense sandstone facies is p2 = 0.18, and the proportion of high-porosity sandstone facies is p3 = 0.15.

[0117] Step B: Analyze the compaction trend curve based on the lithofacies type of the fluvial facies tight heterogeneous reservoir, such as... Figure 3 As shown, a three-dimensional compaction trend model of the P-wave impedance, S-wave impedance, and density parameters for each lithofacies was obtained. This result was then weighted proportionally with the proportion of the corresponding lithofacies type to obtain an initial low-frequency model of the P-wave impedance, S-wave impedance, and density parameters, as shown. Figure 4 As shown in (a), pre-stack inversion is performed on each of them to obtain the three-dimensional data volume of the inverted P-wave impedance, S-wave impedance and density parameters after the first inversion, including the following specific steps:

[0118] Step B1: Based on the lithofacies type, determine the compaction trend curves of the P-wave impedance, S-wave impedance and density parameters for each lithofacies, and establish a three-dimensional body of the compaction trend of the P-wave impedance, S-wave impedance and density parameters for each lithofacies under the stratigraphic constraint.

[0119] The method for establishing the compaction trend curve is as follows:

[0120] Using the well logging curves from drilled wells, cross-sectional analyses of P-wave impedance versus time, S-wave impedance versus time, and density parameters versus time were performed on each rock facies to obtain the corresponding compaction trend curves based on exponential functions.

[0121] Step B2: Based on the proportion of each rock facies and the compaction trend of the three-dimensional volume of the longitudinal wave impedance, transverse wave impedance and density parameters of each rock facies, calculate the initial low-frequency model of the longitudinal wave impedance, transverse wave impedance and density parameters by proportional weighting.

[0122] Specifically, the initial low-frequency model of the longitudinal wave impedance:

[0123] Zp0=p1×Zp f1 +p2×Zp f2 +p3×Zp f3 (Equation 1)

[0124] In the formula, Zp0 represents the initial low-frequency model of the longitudinal wave impedance.

[0125] Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies.

[0126] Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies.

[0127] Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies.

[0128] p1 represents the proportion of mudstone facies.

[0129] p2 represents the proportion of tight sandstone facies.

[0130] p3 represents the proportion of high-porosity sandstone facies.

[0131] Specifically, the low-frequency model of the initial transverse wave impedance:

[0132] Zs0=p1×Zs f1 +p2×Zs f2 +p3×Zs f3 (Equation 2)

[0133] In the formula, Zs0 represents the initial low-frequency model of transverse wave impedance.

[0134] Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies.

[0135] Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies.

[0136] Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies.

[0137] p1 represents the proportion of mudstone facies.

[0138] p2 represents the proportion of tight sandstone facies.

[0139] p3 represents the proportion of high-porosity sandstone facies.

[0140] Specifically, the low-frequency model of the initial density parameters:

[0141] ρ0=p1×ρ f1 +p2×ρ f2 +p3×ρ f3 (Equation 3)

[0142] In the formula, ρ0 represents the initial density parameter low-frequency model.

[0143] ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters.

[0144] ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies.

[0145] ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters.

[0146] p1 represents the proportion of mudstone facies.

[0147] p2 represents the proportion of tight sandstone facies.

[0148] p3 represents the proportion of high-porosity sandstone facies.

[0149] Step B3: Perform pre-stack inversion on the initial low-frequency model of P-wave impedance, S-wave impedance and density parameters to obtain the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters after the first inversion.

[0150] Specifically, the pre-stack inversion uses the Fatti approximation equation based on P-wave impedance, S-wave impedance, and density parameters. Industry software such as Jason and HRS both have pre-stack inversion modules; this embodiment uses the Strata module of HRS software. For detailed implementation procedures, please refer to the HRS software Strata module user manual.

[0151] Step C: Based on the P-wave impedance and S-wave impedance obtained from the first inversion, perform cross-plot analysis and Bayesian discriminant analysis to obtain the first three-dimensional volume of lithofacies probability. Based on the first three-dimensional volume of lithofacies probability, update the low-frequency models of the initial P-wave impedance, S-wave impedance, and density parameters. Using the updated low-frequency models of P-wave impedance, S-wave impedance, and density parameters as initial values, perform iterative calculations of pre-stack inversion, cross-plot analysis, and Bayesian discriminant analysis until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached. This process terminates when the low-frequency models of P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the last iteration of pre-stack inversion, including the following specific steps:

[0152] Step C1: Calculate the P-wave and S-wave velocity ratios based on the P-wave impedance and S-wave impedance obtained from the first inversion. Perform cross-plot analysis on the P-wave velocity ratios and the P-wave impedance obtained from the first inversion to obtain an interaction plot. Perform Bayesian discriminant analysis on the obtained interaction plot to calculate the lithofacies probability value at each sampling point location, thereby obtaining the first three-dimensional lithofacies probability volume. This includes the following specific steps:

[0153] Step C11: Calculate the P-wave and S-wave velocity ratios based on the P-wave impedance and S-wave impedance after the first inversion, and perform cross-plot analysis based on the P-wave and S-wave velocity ratios and the P-wave impedance after the first inversion to obtain a cross-plot diagram.

[0154] Step C12: Perform Bayesian discriminant analysis on the intersection graph, calculate the lithofacies probability value at each sampling point location, and obtain the first three-dimensional lithofacies probability volume;

[0155] Bayesian discriminant analysis is a tool that uses statistical rock physics to conduct quantitative seismic interpretation, making full use of prior probabilistic information from drilled wells. The lithofacies probability value refers to the likelihood or probability of quantitatively interpreting a formation as a certain lithofacies based on seismic inversion data using Bayesian discriminant analysis.

[0156] The Bayesian discriminant analysis method is as follows: The lithofacies category f is calculated using the Bayesian formula. i The posterior probability p(f|d) is calculated using the following formula:

[0157]

[0158] In the formula, p(f|d) represents the lithofacies category f. i The posterior probability;

[0159] p(d|f) indicates that the sample point is of lithofacies f. i The prior probability of d corresponding to time;

[0160] p(f) represents the lithofacies type f i The probability of lithofacies type f is determined by statistical analysis of well logging data. i The prior probability of that lithofacies type, i.e., the proportion of that lithofacies type to all lithofacies types;

[0161] p(d) represents the scaling factor, which is a constant value in Bayesian discriminant analysis;

[0162] Among them, f i (i = 1, ..., N) represents N different lithofacies categories;

[0163] d represents the single-parameter or multi-parameter sample value observed from seismic attributes or well logging curve sample values.

[0164] Step C2: Based on the first probabilistic three-dimensional volume of lithofacies, update the low-frequency models of the initial P-wave impedance, S-wave impedance and density parameters to obtain the updated low-frequency models of P-wave impedance, S-wave impedance and density parameters.

[0165] Specifically, the updated low-frequency model of the longitudinal wave impedance:

[0166] Zp=π(f1)×Zp f1 +π(f2)×Zp f2 +π(f3)×Zp f3 (Equation 5)

[0167] In the formula, Zp represents the updated low-frequency model of the longitudinal wave impedance.

[0168] Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies.

[0169] Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies.

[0170] Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies.

[0171] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0172] π(f2) represents the probability volume of tight sandstone facies.

[0173] π(f3) represents the probability volume of high-porosity sandstone facies.

[0174] Specifically, the updated low-frequency model of the transverse wave impedance:

[0175] Zs=π(f1)×Zs f1 +π(f2)×Zs f2 +π(f3)×Zs f3 (Equation 6)

[0176] In the formula, Zs represents the updated low-frequency model of the transverse wave impedance.

[0177] Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies.

[0178] Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies.

[0179] Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies;

[0180] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0181] π(f2) represents the probability volume of tight sandstone facies.

[0182] π(f3) represents the probability volume of high-porosity sandstone facies.

[0183] Specifically, the low-frequency model of the updated density parameters:

[0184] ρ=π(f1)×ρ f1 +π(f2)×ρ f2 +π(f3)×ρ f3 (Equation 7)

[0185] In the formula, ρ represents the low-frequency model of the updated density parameters.

[0186] ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters.

[0187] ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies.

[0188] ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters.

[0189] π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis.

[0190] π(f2) represents the probability volume of tight sandstone facies.

[0191] π(f3) represents the probability volume of high-porosity sandstone facies.

[0192] Step C3: Using the updated low-frequency model of P-wave impedance, S-wave impedance, and density parameters as initial values, repeat the pre-stack inversion work of step B3 to obtain the inverted three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters after the second inversion; repeat the cross-intersection analysis and Bayesian discriminant analysis of step C1 to obtain the second lithofacies probability three-dimensional volume, and complete one iteration.

[0193] Step C4: Repeat step C3 until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached, and then terminate the process to obtain the low-frequency model of the P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters obtained by pre-stack inversion.

[0194] Specifically, the maximum number of iterations should be set to 5 to 10, depending on the computational cost and time consumption of pre-stack inversion.

[0195] Step D: Based on the low-frequency model of P-wave impedance, S-wave impedance and density parameters after the last iteration, and the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters obtained by pre-stack inversion, calculate the P-wave and S-wave velocity ratio data volume to obtain the distribution law of fluvial facies tight sandstone reservoirs.

[0196] contrast Figure 4 (b) Figure 4 (c) and Figure 4 (d) It can be seen that through multiple iterations of inversion, the vertical and horizontal resolution of the low-frequency model is improved, the bandwidth is widened, and the detailed information is richer. It can effectively compensate for the low-frequency information in the 2-10Hz part and improve the accuracy of the inversion results.

[0197] The method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs disclosed in this invention is compared with existing technologies, such as... Figure 5 and Figure 6 As shown, by Figure 5 It is known that the P-wave and S-wave velocity ratio inversion using the conventional well interpolation low-frequency model exhibits a significant "bull's-eye" phenomenon at the well point, and suffers from low lateral resolution and unclear geological regularities, failing to meet the requirements for characterizing fluvial reservoirs; Figure 6 It can be seen that the low-frequency modeling and P-wave velocity ratio inversion method disclosed in this invention effectively overcomes the "bull's-eye" problem caused by well interpolation, highlights the lateral variation law of fluvial facies heterogeneous reservoirs, enhances the spatial identification ability of geological bodies, conforms to the understanding of geological and sedimentary laws, and improves the characterization effect of fluvial facies reservoirs. Therefore, the method for constructing a low-frequency model of fluvial facies tight heterogeneous reservoirs disclosed in this invention has good results.

[0198] Example 2: A device for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir

[0199] Example 2 provides an apparatus for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir, including...

[0200] The first processing unit is used to classify fluvial facies tight heterogeneous reservoirs into lithofacies types and count the proportion of each lithofacies based on the porosity and clay content curves of the wells already drilled in the study area.

[0201] The second processing unit is used to analyze the compaction trend curve based on the lithofacies type of the fluvial tight heterogeneous reservoir, obtain the three-dimensional compaction trend volume of the P-wave impedance, S-wave impedance and density parameters of each lithofacies, and calculate the initial low-frequency model of the P-wave impedance, S-wave impedance and density parameters by weighting the result with the proportion of the corresponding lithofacies according to the ratio. Then, it performs pre-stack inversion on the model to obtain the three-dimensional data volume of the inverted P-wave impedance, S-wave impedance and density parameters after the first inversion.

[0202] The third processing unit is used to perform cross-plot analysis and Bayesian discriminant analysis based on the P-wave impedance and S-wave impedance after the first inversion to obtain the first three-dimensional volume of lithofacies probability; based on the first three-dimensional volume of lithofacies probability, update the low-frequency model of the initial P-wave impedance, S-wave impedance and density parameters; using the updated low-frequency model of P-wave impedance, S-wave impedance and density parameters as the initial value, perform iterative calculations of pre-stack inversion, cross-plot analysis and Bayesian discriminant analysis until the P-wave impedance, S-wave impedance and density parameters involved in the iterative calculation match the logging curve or reach the set maximum number of iterations, and then terminate to obtain the low-frequency model of P-wave impedance, S-wave impedance and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters after the last iteration of pre-stack inversion;

[0203] The fourth processing unit is used to calculate the P-wave and S-wave velocity ratio data volume based on the low-frequency model of P-wave impedance, S-wave impedance and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters obtained by pre-stack inversion, and to obtain the distribution law of fluvial facies tight sandstone reservoirs.

[0204] Example 3: A computer-readable storage medium

[0205] Example 3 provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in Example 1.

[0206] Example 4: A computer device

[0207] Example 4 provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the method described in Example 1.

[0208] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir, characterized in that, include The fluvial facies tight heterogeneous reservoirs were classified into lithofacies types and the proportion of each lithofacies was statistically analyzed. Compaction trend curves were analyzed based on the lithofacies types of fluvial tight heterogeneous reservoirs to obtain three-dimensional compaction trend volumes of P-wave impedance, S-wave impedance, and density parameters for each lithofacies. These results were then weighted proportionally with the proportion of the corresponding lithofacies type to obtain initial low-frequency models of P-wave impedance, S-wave impedance, and density parameters. Pre-stack inversion was then performed on these models to obtain three-dimensional inversion data volumes of P-wave impedance, S-wave impedance, and density parameters after the first inversion. Based on the P-wave impedance and S-wave impedance obtained from the first inversion, cross-plot analysis and Bayesian discriminant analysis are performed to obtain the first three-dimensional volume of lithofacies probability. Based on the first three-dimensional volume of lithofacies probability, the low-frequency models of the initial P-wave impedance, S-wave impedance, and density parameters are updated. Using the updated low-frequency models of P-wave impedance, S-wave impedance, and density parameters as initial values, iterative calculations of pre-stack inversion, cross-plot analysis, and Bayesian discriminant analysis are performed until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached, and then the calculation terminates. The low-frequency models of P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the last iteration are obtained. Based on the three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion after the last iteration, the P-wave and S-wave velocity ratio data volume is calculated to obtain the distribution pattern of fluvial facies tight sandstone reservoirs.

2. The construction method according to claim 1, characterized in that, The process of classifying fluvial facies tight heterogeneous reservoirs into lithofacies types and calculating the proportion of each lithofacies includes the following steps: Based on the mud content and porosity curves of the drilled wells in the study area, the mud content and porosity threshold values ​​of the fluvial facies tight heterogeneous reservoirs were obtained. Based on the mud content and porosity threshold values, the fluvial facies tight heterogeneous reservoirs were divided into mudstone facies, tight sandstone facies and high-porosity sandstone facies. The proportions of mudstone facies, tight sandstone facies, and high-porosity sandstone facies in the drilled formations were calculated separately.

3. The construction method according to claim 1, characterized in that, Obtaining the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters after the first inversion includes the following steps: Based on the described lithofacies type, the compaction trend curves of the P-wave impedance, S-wave impedance and density parameters of each lithofacies are determined. Under the constraint of the stratigraphic position, a three-dimensional body of the compaction trend of the P-wave impedance, S-wave impedance and density parameters of each lithofacies is established. Based on the proportion of each rock facies and the compaction trend of the longitudinal wave impedance, transverse wave impedance and density parameters of each rock facies in a three-dimensional volume, a low-frequency model of the initial longitudinal wave impedance, transverse wave impedance and density parameters is obtained by proportional weighting calculation. The initial low-frequency model of longitudinal wave impedance, transverse wave impedance and density parameters is inverted before stacking to obtain the three-dimensional data volume of longitudinal wave impedance, transverse wave impedance and density parameters after the first inversion.

4. The construction method according to claim 3, characterized in that, The initial low-frequency resistance model for longitudinal wave impedance: Zp0 = p1 × Zp f1 +p2×Zp f2 +p3×Zp f3 (Formula 1) In the formula, Zp0 represents the initial low-frequency model of the longitudinal wave impedance. Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies. Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies. Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies. p1 represents the proportion of mudstone facies. p2 represents the proportion of tight sandstone facies. p3 represents the proportion of high-porosity sandstone facies; The initial low-frequency model of the transverse wave impedance: Zs0=p1×Zs f1 +p2×Zs f2 +p3×Zs f3 (Equation 2) In the formula, Zs0 represents the initial low-frequency model of the transverse wave impedance. Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies. Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies. Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies. p1 represents the proportion of mudstone facies. p2 represents the proportion of tight sandstone facies. p3 represents the proportion of high-porosity sandstone facies; The low-frequency model of the initial density parameters: ρ0=p1×ρ f1 +p2×ρ f2 +p3×ρ f3 (Equation 3) In the formula, ρ0 represents the initial density parameter low-frequency model. ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters. ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies. ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters. p1 represents the proportion of mudstone facies. p2 represents the proportion of tight sandstone facies. p3 represents the proportion of high-porosity sandstone facies.

5. The construction method according to claim 3, characterized in that, Obtaining the low-frequency model of the P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters obtained from the pre-stack inversion, includes the following steps: Step C1: Calculate the P-wave and S-wave velocity ratio based on the P-wave impedance and S-wave impedance after the first inversion. Perform cross-intersection analysis based on the P-wave and S-wave velocity ratio and the P-wave impedance after the first inversion to obtain an interaction diagram. Perform Bayesian discriminant analysis on the obtained interaction diagram to calculate the lithofacies probability value at each sampling point location and obtain the first three-dimensional lithofacies probability volume. Step C2: Based on the first probabilistic three-dimensional volume of lithofacies, update the low-frequency models of the initial P-wave impedance, S-wave impedance and density parameters to obtain the updated low-frequency models of P-wave impedance, S-wave impedance and density parameters. Step C3: Using the updated low-frequency model of P-wave impedance, S-wave impedance, and density parameters as initial values, repeat the pre-stack inversion work to obtain the second inverted three-dimensional data volume of P-wave impedance, S-wave impedance, and density parameters; repeat the cross-plot analysis and Bayesian discriminant analysis to obtain the second lithofacies probability three-dimensional volume, completing one iteration; Step C4: Repeat step C3 until the P-wave impedance, S-wave impedance, and density parameters involved in the iterative calculation match the logging curve or the set maximum number of iterations is reached, and then terminate the process to obtain the low-frequency model of the P-wave impedance, S-wave impedance, and density parameters after the last iteration, as well as the three-dimensional data volume of the P-wave impedance, S-wave impedance, and density parameters obtained by pre-stack inversion.

6. The construction method according to claim 5, characterized in that, Obtaining the updated low-frequency model for the P-wave impedance, S-wave impedance, and density parameters includes the following steps: The P-wave and S-wave velocity ratios are calculated based on the P-wave impedance and S-wave impedance after the first inversion. Cross-plot analysis is then performed based on the P-wave and S-wave velocity ratios and the P-wave impedance after the first inversion to obtain a cross-plot diagram. Bayesian discriminant analysis was performed on the intersection graph to calculate the lithofacies probability value at each sampling point, thus obtaining the first three-dimensional lithofacies probability volume. The Bayesian discriminant analysis method involves calculating the lithofacies category f using Bayes' theorem. i The posterior probability p(f|d) is calculated using the following formula: In the formula, p(f|d) represents the lithofacies category f. i The posterior probability; p(d|f) indicates that the sample point is of lithofacies f. i The prior probability of d corresponding to time; p(f) represents the lithofacies type f i The probability of lithofacies type f is determined by statistical analysis of well logging data. i The prior probability of that lithofacies type, i.e., the proportion of that lithofacies type to all lithofacies types; p(d) represents the scaling factor, which is a constant value in Bayesian discriminant analysis; Among them, f i (i = 1, ..., N) represents N different lithofacies categories; d represents the single-parameter or multi-parameter sample value observed from seismic attributes or well logging curve sample values.

7. The construction method according to claim 5, characterized in that, The updated low-frequency model of the longitudinal wave impedance: Zp=π(f1)×Zp f1 +π(f2)×Zp f2 +π(f3)×Zp f3 (formula 5) In the formula, Zp represents the updated low-frequency model of the longitudinal wave impedance. Zp f1 A three-dimensional volume representing the longitudinal wave impedance compaction trend of mudstone facies. Zp f2 A three-dimensional volume representing the compaction trend of longitudinal wave impedance in tight sandstone facies. Zp f3 A three-dimensional volume representing the longitudinal wave impedance compaction trend of high-porosity sandstone facies. π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis. π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis. π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria; The updated low-frequency model of the shear wave impedance: Zs=π(f1)×Zs f1 +π(f2)×Zs f2 +π(f3)×Zs f3 (formula 6) In the formula, Zs represents the updated low-frequency model of the transverse wave impedance. Zs f1 A three-dimensional volume representing the shear wave impedance compaction trend of mudstone facies. Zs f2 A three-dimensional volume representing the shear wave impedance compaction trend of tight sandstone facies. Zs f3 A three-dimensional volume representing the shear wave impedance compaction trend of high-porosity sandstone facies; π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis. π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis. π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria; The low-frequency model of the updated density parameters: ρ=π(f1)×ρ f1 +π(f2)×ρ f2 +π(f3)×ρ f3 (formula 7) In the formula, ρ represents the low-frequency model of the updated density parameters. ρ f1 A three-dimensional volume representing the compaction trend of mudstone facies density parameters. ρ f2 A three-dimensional volume representing the compaction trend of density parameters in tight sandstone facies. ρ f3 A three-dimensional volume representing the compaction trend of high-porosity sandstone facies density parameters. π(f1) represents the probability volume of mudstone facies determined by Bayesian analysis. π(f2) represents the probability volume of the tight sandstone facies as determined by Bayesian analysis. π(f3) represents the probability volume of high-porosity sandstone facies determined by Bayesian criteria.

8. A device for constructing a low-frequency model of a fluvial facies tight heterogeneous reservoir, characterized in that, include The first processing unit is used to classify fluvial facies tight heterogeneous reservoirs into lithofacies types and to calculate the proportion of each lithofacies. The second processing unit is used to analyze the compaction trend curve based on the lithofacies type of the fluvial tight heterogeneous reservoir, obtain the three-dimensional compaction trend volume of the P-wave impedance, S-wave impedance and density parameters of each lithofacies, and calculate the initial low-frequency model of the P-wave impedance, S-wave impedance and density parameters by weighting the result with the proportion of the corresponding lithofacies according to the ratio. Then, it performs pre-stack inversion on the model to obtain the three-dimensional data volume of the P-wave impedance, S-wave impedance and density parameters after the first inversion. The third processing unit is used to perform cross-plot analysis and Bayesian discriminant analysis based on the P-wave impedance and S-wave impedance after the first inversion to obtain the first three-dimensional volume of lithofacies probability; based on the first three-dimensional volume of lithofacies probability, update the low-frequency model of the initial P-wave impedance, S-wave impedance and density parameters; using the updated low-frequency model of P-wave impedance, S-wave impedance and density parameters as the initial value, perform iterative calculations of pre-stack inversion, cross-plot analysis and Bayesian discriminant analysis until the P-wave impedance, S-wave impedance and density parameters involved in the iterative calculation match the logging curve or reach the set maximum number of iterations, and then terminate to obtain the low-frequency model of P-wave impedance, S-wave impedance and density parameters after the last iteration, as well as the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters after the last iteration of pre-stack inversion; The fourth processing unit is used to calculate the P-wave and S-wave velocity ratio data volume based on the three-dimensional data volume of P-wave impedance, S-wave impedance and density parameters obtained from the pre-stack inversion after the last iteration, and to obtain the distribution law of fluvial facies tight sandstone reservoirs.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method described in any one of claims 1-7.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1-7.

Images