Method and device for simulating high-quality reservoir of tight sandstone and computer equipment

Abstract

The invention provides a high-quality reservoir simulation method and device for tight sandstone and computer equipment, and belongs to the technical field of oil-gas exploration. The high-quality reservoir simulation method for the tight sandstone comprises the following steps: establishing a three-dimensional structure stratum framework model of a target oil and gas reservoir according to three-dimensional seismic data; obtaining sand shale discrete data, sedimentary facies discrete data and granularity lithofacies discrete data according to the obtained parameter interpretation data and the three-dimensional structure stratum framework model; performing simulation based on the sand shale discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data to obtain a granularity lithofacies three-dimensional geologic model; and determining a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the granularity lithofacies three-dimensional geologic model, wherein the high-quality reservoir three-dimensional distribution model is used for describing a high-quality reservoir of the target oil and gas reservoir. According to the method, the high-quality reservoir of the tight sandstone is constrained and simulated by establishing the granularity lithofacies three-dimensional geologic model of the tight sandstone, and three-dimensional distribution and accurate simulation of different lithofacies are quantitatively reflected.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oil and gas exploration, in particular to a high-quality reservoir simulation method for tight sandstone, a high-quality reservoir simulation device for tight sandstone, a computer device and a computer storage medium. BACKGROUND

[0002] With the development of natural gas industry, the field and scale of tight sandstone gas reservoir development are rapidly expanding. The natural gas in the tight sandstone gas reservoir is stored in low-porosity and low-permeability reservoir rocks, so it is difficult to effectively develop it by using conventional exploration and development technology. The key to the exploration and development of such gas reservoirs lies in accurately depicting the relatively good physical properties of part of the reservoir under the background of generally low porosity and low permeability (i.e. high-quality reservoir geological modeling). For the tight clastic rock reservoir, quantitatively describing the three-dimensional spatial distribution of high-quality reservoirs is the basis for the efficient development of such oil and gas reservoirs.

[0003] Facies-constrained modeling is to analyze and simulate various reservoir property and oil / gas-bearing property attribute models by taking facies (such as lithofacies, sedimentary microfacies, and petrophysical facies) as the constraint condition. In conventional sandstone reservoirs, the distribution of high-quality reservoirs is usually related to the dominant sedimentary facies belt. In tight sandstone reservoirs, due to the complexity of the reservoir densification process, the original sedimentation often undergoes late diagenetic evolution and becomes dense, and the degree of response of reservoir rock structure and reservoir property (porosity, permeability, etc.) to sedimentary facies is reduced. Therefore, for tight sandstone gas reservoirs, a lithology or lithofacies model that is more sensitive to reservoir property needs to be constructed for high-quality reservoirs as the target.

[0004] Lithofacies is an evaluation of underground geological bodies based on sedimentary structure, particle type or sedimentary structure, and is a sedimentary body with the same porosity and permeability relationship or trend under the same sedimentary environment. The world is rich in tight sandstone oil and gas resources, but the actual producing degree is very low due to the strong heterogeneity of the reservoir. The main reason is that the spatial distribution of high-quality reservoirs has not been accurately depicted. Lithofacies in tight sandstone reservoirs is the fundamental factor controlling diagenesis and reservoir and permeability properties, and high-quality reservoirs are controlled by the development degree of favorable lithofacies. Although the traditional stochastic simulation method can directly construct a three-dimensional lithofacies model from the well point, it is difficult to accurately characterize the interwell distribution of lithofacies during modeling due to the strong heterogeneity of tight sandstone reservoirs, resulting in the final model lacking geological content of high-quality reservoirs and having no predictive significance.

[0005] The important data for evaluating underground oil and gas reservoir resources and development potential are reservoir porosity and permeability and other attribute parameters. To establish these attribute geological models, a lithofacies model that can reliably constrain the quality of the reservoir is needed as a control. SUMMARY

[0006] In order to solve the above technical defects, the application provides a high-quality reservoir simulation method and device for tight sandstone and computer equipment, the high-quality reservoir simulation method for tight sandstone is used for restraining simulation of high-quality reservoirs of tight sandstone by establishing a particle size lithofacies three-dimensional geological model of tight sandstone, and the three-dimensional distribution of different lithofacies is quantitatively reflected and accurately simulated by sequentially constructing three-level models of sand-shale facies, sedimentary facies and particle size lithofacies, so that spatial description of high-quality reservoirs of strong heterogeneous tight sandstone is finally realized, and a more reliable geological model basis is provided for numerical simulation and productivity prediction of tight sandstone oil and gas reservoirs.

[0007] The first aspect of the application provides a high-quality reservoir simulation method for tight sandstone, comprising:

[0008] A three-dimensional structure stratigraphic framework model of the target oil and gas reservoir is established according to three-dimensional seismic data;

[0009] Parameter interpretation data of the target oil and gas reservoir is obtained according to cable logging curve data of the target oil and gas reservoir, and sand-shale discrete data, sedimentary facies discrete data and particle size lithofacies discrete data are obtained according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure stratigraphic framework model;

[0010] A particle size lithofacies three-dimensional geological model is obtained by simulation based on the sand-shale discrete data, the sedimentary facies discrete data and the particle size lithofacies discrete data;

[0011] A physical parameter three-dimensional geological model of the target oil and gas reservoir is obtained by simulation according to the parameter interpretation data of the target oil and gas reservoir, the particle size lithofacies discrete data and the particle size lithofacies three-dimensional geological model;

[0012] A high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is determined according to the physical three-dimensional geological model of the target oil and gas reservoir, and the high-quality reservoir three-dimensional distribution model is used for describing the high-quality reservoir of the target oil and gas reservoir.

[0013] In the embodiment of the application, the three-dimensional structure stratigraphic framework model of the target oil and gas reservoir is established according to three-dimensional seismic data, comprising:

[0014] Fine structure interpretation is performed on the target oil and gas reservoir by using three-dimensional seismic data, and layer and fracture data of the target oil and gas reservoir are obtained;

[0015] The three-dimensional structure stratigraphic framework model is established according to the layer and fracture data of the target oil and gas reservoir.

[0016] In the embodiment of the application, the parameter interpretation data of the target oil and gas reservoir is obtained according to logging curve data of the target oil and gas reservoir, and the sand-shale discrete data, the sedimentary facies discrete data and the particle size lithofacies discrete data are obtained according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure stratigraphic framework model, comprising:

[0017] According to well logging curve data of a target oil and gas reservoir, parameter interpretation data of the target oil and gas reservoir is obtained, wherein the parameter interpretation data comprises physical property parameter interpretation data, lithology parameter interpretation data and mineral parameter interpretation data;

[0018] The obtained parameter interpretation data is discretized into the three-dimensional structure stratigraphic framework model to obtain sand-shale discrete data, sedimentary facies discrete data and grain size lithofacies discrete data.

[0019] In the embodiment of the present application, the simulation based on the sand-shale discrete data, the sedimentary facies discrete data and the grain size lithofacies discrete data obtains a grain size lithofacies three-dimensional geological model, which comprises:

[0020] A sand-shale three-dimensional geological model is obtained according to the sand-shale discrete data;

[0021] A sedimentary facies three-dimensional geological model is obtained according to the sedimentary facies discrete data and the sand-shale three-dimensional geological model;

[0022] A shale content three-dimensional geological model is obtained according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model;

[0023] A grain size lithofacies three-dimensional geological model is obtained according to the grain size lithofacies discrete data, the sedimentary facies three-dimensional geological model and the shale content three-dimensional geological model.

[0024] In the embodiment of the present application, the sand-shale three-dimensional geological model is obtained according to the sand-shale discrete data, which comprises:

[0025] Geostatistical analysis is performed on the sand-shale discrete data to obtain variogram parameter data of the sand-shale discrete data distribution;

[0026] Sequential indicator simulation is performed on the target oil and gas reservoir according to the variogram parameter data of the sand-shale discrete data distribution to obtain the sand-shale three-dimensional geological model.

[0027] In the embodiment of the present application, the sedimentary facies three-dimensional geological model is obtained according to the sedimentary facies discrete data and the sand-shale three-dimensional geological model, which comprises:

[0028] A sedimentary facies planar distribution two-dimensional map of the target oil and gas reservoir is obtained according to the sedimentary facies discrete data;

[0029] A sedimentary facies ratio is obtained according to the sand-shale three-dimensional geological model;

[0030] Multi-point geostatistical simulation is performed on the target oil and gas reservoir according to the sedimentary facies planar distribution two-dimensional map and the sedimentary facies ratio to obtain the sedimentary facies three-dimensional geological model.

[0031] In the embodiment of the present application, the obtaining of the shale content three-dimensional geological model according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model comprises:

[0032] According to the shale content interpretation data in the sedimentary facies discrete data and the parameter interpretation data, the spatial data distribution of the shale content in different sedimentary facies is obtained.

[0033] According to the sedimentary facies three-dimensional geological model and the spatial data distribution of the shale content in the sedimentary facies, the sequential Gaussian random simulation of the target oil and gas reservoir is performed to obtain the shale content three-dimensional geological model.

[0034] In the embodiment of the present application, the obtaining of the shale content three-dimensional geological model according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model comprises:

[0035] The optimal grain size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and a three-dimensional inversion volume model is obtained according to the optimal grain size lithofacies sensitive well logging curve.

[0036] According to the shale content three-dimensional geological model and the three-dimensional inversion volume model, the spatial probability volume of each grain size lithofacies is obtained.

[0037] The geological statistical analysis of the grain size lithofacies discrete data is performed to obtain the variogram parameter data of the grain size lithofacies discrete data distribution.

[0038] According to the variogram parameter data of the grain size lithofacies discrete data distribution, the spatial probability volume of each grain size lithofacies and the sedimentary facies three-dimensional geological model, the sequential indicator random simulation of the target oil and gas reservoir is performed to obtain the grain size lithofacies three-dimensional geological model.

[0039] In the embodiment of the present application, the obtaining of the optimal grain size lithofacies sensitive well logging curve of the target oil and gas reservoir and the obtaining of the three-dimensional inversion volume model according to the optimal grain size lithofacies sensitive well logging curve comprise:

[0040] The grain size lithofacies sensitive well logging curve of the participating well of the target oil and gas reservoir and the shale content interpretation data in the parameter interpretation data are input into the grain size lithofacies classification information neural network model which is pre-trained.

[0041] The grain size lithofacies classification information neural network model outputs the classification information of the grain size lithofacies of the participating well.

[0042] According to the comparison of the classification information of the grain size lithofacies of the non-participating well of the target oil and gas reservoir and the classification information of the grain size lithofacies of the participating well, the comparison result is obtained.

[0043] According to the comparison result, the optimal grain size lithofacies sensitive well logging curve in the grain size lithofacies sensitive well logging curve of the participating well is determined.

[0044] According to the optimal granularity lithofacies sensitive well logging curve, the three-dimensional inversion body model is obtained.

[0045] In the embodiment of the present application, the space probability body of each granularity lithofacies is obtained according to the argillaceous content three-dimensional geological model and the three-dimensional inversion body model, and the method comprises the following steps:

[0046] The argillaceous content three-dimensional geological model and the three-dimensional inversion body model are input into the granularity lithofacies space probability body neural network model which is pre-trained;

[0047] The granularity lithofacies space probability body neural network model outputs the space probability body of each granularity lithofacies.

[0048] In the embodiment of the present application, the physical property parameter three-dimensional geological model of the target oil and gas reservoir is obtained by simulation according to the parameter interpretation data of the target oil and gas reservoir, the granularity lithofacies discrete data and the granularity lithofacies three-dimensional geological model, and the method comprises the following steps:

[0049] The spatial data distribution of the physical property parameters in different granularity lithofacies is obtained according to the granularity lithofacies discrete data and the physical property parameter interpretation data in the parameter interpretation data.

[0050] The physical property parameter three-dimensional geological model is obtained by sequentially simulating the target oil and gas reservoir according to the granularity lithofacies three-dimensional geological model and the spatial distribution of the physical property parameters in different granularity lithofacies.

[0051] In the embodiment of the present application, the high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is determined according to the physical property three-dimensional geological model of the target oil and gas reservoir, and the method comprises the following steps:

[0052] The relationship between the physical property parameters in each granularity lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir is obtained.

[0053] The threshold value of the physical property parameters of different granularity lithofacies to the high-quality reservoir is determined according to the relationship between the physical property parameters in each granularity lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir.

[0054] According to the granularity lithofacies three-dimensional geological model and the physical property parameter three-dimensional geological model, the grid body which meets all the physical property parameter threshold values at the same time is calculated.

[0055] According to the grid body which meets all the physical property parameter threshold values at the same time, the high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is obtained.

[0056] The second aspect of the present application provides a high-quality reservoir simulation device for tight sandstone, which comprises:

[0057] The first model construction module is used for establishing the three-dimensional structure stratigraphic framework model of the target oil and gas reservoir according to the three-dimensional seismic data.

[0058] The data acquisition module acquires parameter interpretation data of the target oil and gas reservoir according to the cable logging curve data of the target oil and gas reservoir, and obtains sandstone and mudstone discrete data, sedimentary facies discrete data and grain size lithofacies discrete data according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure stratigraphic framework model;

[0059] The first simulation module is configured to simulate based on the sandstone and mudstone discrete data, the sedimentary facies discrete data and the grain size lithofacies discrete data, and obtain a grain size lithofacies three-dimensional geological model.

[0060] The second simulation module is configured to simulate according to the parameter interpretation data of the target oil and gas reservoir, the grain size lithofacies discrete data and the grain size lithofacies three-dimensional geological model, and obtain a physical property parameter three-dimensional geological model of the target oil and gas reservoir.

[0061] The second model construction module is configured to determine a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the physical property three-dimensional geological model of the target oil and gas reservoir, and the high-quality reservoir three-dimensional distribution model is used to describe the high-quality reservoir of the target oil and gas reservoir.

[0062] The third aspect of the present application provides a computer device, comprising:

[0063] a memory;

[0064] a processor; and

[0065] a computer program;

[0066] The computer program is stored in the memory and is configured to be executed by the processor to realize the high-quality reservoir simulation method of the tight sandstone as described above.

[0067] The fourth aspect of the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor to realize the high-quality reservoir simulation method of the tight sandstone as described above.

[0068] The high-quality reservoir simulation method of the tight sandstone constrains the simulation of the high-quality reservoir of the tight sandstone by establishing a grain size lithofacies three-dimensional geological model of the tight sandstone, and quantitatively reflects the three-dimensional distribution of different lithofacies and accurately simulates by sequentially constructing three-level models of sandstone and mudstone facies, sedimentary facies and grain size lithofacies, and finally realizes the spatial description of the high-quality reservoir of the strong heterogeneous tight sandstone, and provides a more reliable geological model basis for numerical simulation and productivity prediction of the tight sandstone oil and gas reservoir.

[0069] Other features and advantages of the technical scheme of the present application will be described in detail in the specific implementation manner part below. BRIEF DESCRIPTION OF DRAWINGS

[0070] The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application. In the drawings:

[0071] Figure 1 is a flow chart of the high-quality reservoir simulation method for tight sandstone provided by the embodiments of the application;

[0072] Figure 2 is a structural block diagram of the high-quality reservoir simulation device for tight sandstone provided by the embodiments of the application;

[0073] Figure 3 is a three-dimensional geological model diagram of sandstone and mudstone facies in the research area in the embodiments;

[0074] Figure 4 is a sedimentary microfacies geological model diagram in the research area in the embodiments, wherein Figure 4 a is a three-dimensional model diagram of sedimentary microfacies; Figure 4 b is a corresponding two-dimensional geological model profile diagram in the three-dimensional model;

[0075] Figure 5 is a relationship diagram between the granularity lithofacies and the shale content in the research area in the embodiments, wherein Figure 5 a is a crossplot of the shale content and the natural gamma-ray data obtained by logging statistics; Figure 5 b is a corresponding relationship diagram of the granularity lithofacies and the sensitive curve natural gamma-ray;

[0076] Figure 6 is a three-dimensional geological model diagram of the shale content in the research area in the embodiments;

[0077] Figure 7 is a granularity lithofacies classification diagram in the research area in the embodiments by using the two-dimensional neural network clustering of the shale content;

[0078] Figure 8 is a granularity lithofacies probability diagram in the research area in the embodiments by using the three-dimensional neural network clustering of the shale content, wherein Figure 8 a is a three-dimensional probability body model of coarse sandstone facies; Figure 8 b is a three-dimensional probability body model diagram of medium sandstone facies;8c is a three-dimensional probability body model diagram of fine sandstone facies;8d is a three-dimensional probability body model of mudstone facies;

[0079] Figure 9 is a three-dimensional geological model diagram of the granularity lithofacies in the research area in the embodiments;

[0080] Figure 10 is a three-dimensional geological model diagram of the physical property parameters in the research area in the embodiments, wherein Figure 10 a is a three-dimensional geological model diagram of porosity; Figure 10 b is a three-dimensional geological model diagram of permeability;

[0081] Figure 11 Figure 3 is a three-dimensional distribution model diagram of high-quality reservoirs in the research area in the embodiment;

[0082] Figure 12 Figure 4 is a corresponding two-dimensional geological model profile diagram of the three-dimensional model of granular facies in the research area in the embodiment. DETAILED DESCRIPTION

[0083] In order to make the technical solutions and advantages of the embodiments of the present application clearer, the exemplary embodiments of the present application are further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0084] In the description of the present application, it should be understood that the terms "center", "longitudinal", "transverse", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation of the present application.

[0085] In addition, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise explicitly specified and limited.

[0086] In the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connection", "connection", "fixing" and the like should be understood broadly, for example, can be fixedly connected, or can be detachably connected, or can be integrated; can be mechanically connected, or can be electrically connected or can communicate with each other; can be directly connected, or can be indirectly connected through an intermediate medium; can be the internal connection of two elements or the interaction relationship between two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0087] In the process of implementing the present application, the inventors found that with the development of the natural gas industry, the field and scale of development of tight sandstone gas reservoirs are rapidly expanding. The natural gas in tight sandstone gas reservoirs is hosted in low-porosity, low-permeability reservoir rocks, so it is difficult to effectively develop them using conventional exploration and development techniques. The key to the exploration and development of such gas reservoirs is to accurately characterize the relatively good physical properties of part of the reservoir under the background of generally low porosity and low permeability (i.e., high-quality reservoir geological modeling). For the tight clastic reservoir, quantitatively describing the three-dimensional spatial distribution of high-quality reservoirs is the basis for the efficient development of such oil and gas reservoirs.

[0088] The facies-constrained modeling is to analyze and simulate various reservoir property and oil / gas-bearing attribute models with facies (such as lithofacies, sedimentary microfacies, and petrophysical facies) as the constraint condition. In conventional sandstone reservoirs, the distribution of high-quality reservoirs is usually related to the dominant sedimentary facies belt. In tight sandstone reservoirs, due to the complexity of the reservoir densification process, the original sedimentation often undergoes late diagenetic evolution and becomes tight, and the degree of the controlled response of the reservoir rock structure and the reservoir property (porosity, permeability, etc.) to the sedimentary facies is reduced, so the sedimentary facies alone cannot control the changes in the internal reservoir property. Therefore, for tight sandstone gas reservoirs, a lithology or lithofacies model that is more sensitive to the reservoir property needs to be constructed for the target of high-quality reservoirs.

[0089] The lithofacies is evaluated based on the sedimentary structure, particle type, or sedimentary structure of the underground geological body, and is a sedimentary body with the same porosity and permeability relationship or trend under the same sedimentary environment. The tight sandstone oil and gas resources are abundant in the world, but the actual producing degree is very low due to the strong heterogeneity of the reservoirs, and the main reason is that the spatial distribution of high-quality reservoirs is not accurately characterized. The lithofacies in the tight sandstone reservoir is the fundamental factor controlling the diagenesis and reservoir and permeability properties, and the high-quality reservoirs are controlled by the development degree of the favorable lithofacies. Although the traditional stochastic simulation method can directly construct a three-dimensional lithofacies model from the well points, due to the strong heterogeneity of the tight sandstone reservoir, it is difficult to accurately characterize the interwell distribution of the lithofacies during the modeling process, resulting in the final model lacking the geological content of the high-quality reservoirs and thus having no predictive significance.

[0090] The important data for evaluating the underground oil and gas reservoir resources and development potential are the reservoir porosity and permeability and other attribute parameters, and to establish these attribute geological models, a lithofacies model that can reliably constrain the quality of the reservoir is needed.

[0091] According to the core and logging interpretation results, it is found that the porosity of different particle size lithology / lithofacies in the tight sandstone reservoirs has obvious differences. According to the logging particle size interpretation, several particle size lithofacies types can be classified. Under normal circumstances, the medium-coarse grained sandstone is a favorable lithofacies, and the fine sand and siltstone has obviously poor physical properties.

[0092] Because the planar distribution of different lithofacies is not known, if the facies model is established by relying on the traditional sequential indication method, the randomness is too strong, and many "noises" will appear, and the distribution of the lithofacies is also difficult to guarantee. In this case, to establish the grain lithofacies model, the spatial distribution of the grain size needs to be first determined to be controlled by the sedimentary facies, for example, the grain size of the channel sandstone is coarse at the bottom and fine at the top. However, the qualitative understanding cannot meet the needs of quantitative modeling. For example, in the channel sand body, the grain size presents the characteristics of changing from coarse to fine from bottom to top, but the specific positions of the different grain size sandstone cannot be determined.

[0093] In order to more accurately and quantitatively depict the spatial distribution of the grain lithofacies, a variable needs to be found to connect the spatial relationship between the sedimentary facies and the grain lithofacies. Based on the logging grain size interpretation model, the shale content distribution of different grain size sandstone has significant differences, that is, the amount of shale content can quantitatively determine the sandstone grain size to a certain extent. For example, in the channel sedimentary facies, from the bottom of the channel to the top, the sandstone grain size gradually transitions from coarse sandstone to fine sandstone and siltstone, and correspondingly, from bottom to top, the shale content in the sandstone changes from less to more, and the mouth bar sand body is opposite.

[0094] In view of the above problems, the embodiment of the present application provides a high-quality reservoir simulation method of tight sandstone, comprising the following steps: establishing a three-dimensional structure stratigraphic framework model of a target oil and gas reservoir according to three-dimensional seismic data; obtaining parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtaining sandstone and mudstone discrete data, sedimentary facies discrete data and grain lithofacies discrete data according to the obtained parameter interpretation data and the three-dimensional structure stratigraphic framework model; obtaining a sandstone and mudstone three-dimensional geological model according to the sandstone and mudstone discrete data; obtaining a sedimentary facies three-dimensional geological model according to the sedimentary facies discrete data and the sandstone and mudstone three-dimensional geological model; obtaining a shale content three-dimensional geological model according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model; obtaining a grain lithofacies three-dimensional geological model according to the grain lithofacies discrete data, the sedimentary facies three-dimensional geological model and the shale content three-dimensional geological model; obtaining a physical property parameter three-dimensional geological model of the target oil and gas reservoir according to the grain lithofacies discrete data, the parameter interpretation data and the grain lithofacies three-dimensional geological model; and determining a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the physical property three-dimensional geological model, wherein the high-quality reservoir three-dimensional distribution model is used for describing the high-quality reservoir of the target oil and gas reservoir. The high-quality reservoir simulation method of tight sandstone constrains the simulation of the high-quality reservoir of tight sandstone by establishing a grain lithofacies three-dimensional geological model of tight sandstone, and quantitatively reflects the three-dimensional distribution of different lithofacies and accurate simulation by sequentially constructing three-level models of sandstone and mudstone facies, sedimentary facies and grain lithofacies, so as to finally realize the spatial depiction of the high-quality reservoir of strong heterogeneous tight sandstone, and provide a more reliable geological model basis for numerical simulation and productivity prediction of tight sandstone oil and gas reservoirs.

[0095] Further, the application clusters and discriminates the particle size lithofacies model establishment method by shale content, on the basis of sedimentary facies modeling, the technical process can further simulate the sand body quality (such as good sand, bad sand and the like) of the tight reservoir, and provides a new idea for spatial prediction and quantitative characterization of the high-quality reservoir of the tight sandstone oil and gas reservoir.

[0096] The tight sandstone oil and gas reservoir has strong heterogeneity, and it is difficult to control the change of the reservoir internal storage and percolation and the like only by relying on the sedimentary facies, so it is necessary to explore a three-dimensional geological modeling method which can more accurately characterize the distribution of the high-quality reservoir. In order to overcome the defects of the prior art, the high-quality reservoir simulation method of the tight sandstone provided by the application realizes the spatial characterization and distribution evaluation of the high-quality reservoir of the strong heterogeneity tight sandstone by sequentially constructing the sand-shale facies, the sedimentary facies and the particle size lithofacies three-level models, quantitatively reflecting the three-dimensional distribution of different lithofacies and accurately simulating, and provides a more reliable geological model basis for numerical simulation and productivity prediction of the tight sandstone oil and gas reservoir. Through the neural network clustering and discrimination of the shale content, the quantitative model construction from the sedimentary facies to the particle size lithofacies in the three-dimensional space is realized, and through the three-level constraints of the sand-shale facies, the sedimentary facies and the particle size lithofacies, the model establishment which can characterize the high-quality reservoir and the spatial distribution of the attribute parameters of the tight clastic rock is realized.

[0097] Figure 1 is a flow chart of the high-quality reservoir simulation method of the tight sandstone provided by the embodiment of the application. As shown in Figure 1 the high-quality reservoir simulation method of the tight sandstone provided by the embodiment of the application includes the following steps:

[0098] S1. establishing a three-dimensional structure stratigraphic framework model of a target oil and gas reservoir according to three-dimensional seismic data;

[0099] S2. obtaining parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtaining sand-shale discrete data, sedimentary facies discrete data and particle size lithofacies discrete data according to the obtained parameter interpretation data and the three-dimensional structure stratigraphic framework model;

[0100] S3. simulating based on the sand-shale discrete data, the sedimentary facies discrete data and the particle size lithofacies discrete data to obtain a particle size lithofacies three-dimensional geological model;

[0101] S4. simulating according to the parameter interpretation data of the target oil and gas reservoir, the particle size lithofacies discrete data and the particle size lithofacies three-dimensional geological model to obtain a physical property parameter three-dimensional geological model of the target oil and gas reservoir;

[0102] S5. determining a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the physical property three-dimensional geological model of the target oil and gas reservoir, and the high-quality reservoir three-dimensional distribution model is used for describing the high-quality reservoir of the target oil and gas reservoir.

[0103] In step S1, the three-dimensional structure stratigraphic framework model of the target oil and gas reservoir is established according to the three-dimensional seismic data of the target oil and gas reservoir, including:

[0104] The three-dimensional structure stratigraphic framework model is established by using the three-dimensional seismic data to perform fine structure interpretation on the target oil and gas reservoir, and obtaining the layer and fault data of the target oil and gas reservoir.

[0105] The three-dimensional structure stratigraphic framework model is established according to the layer and fault data of the target oil and gas reservoir.

[0106] Specifically, the three-dimensional structure stratigraphic framework model is established by using the three-dimensional seismic data to perform fine structure interpretation on the target oil and gas reservoir, obtaining the layer and fault data of the target oil and gas reservoir, creating a layer model and a fault model, setting a modeling area and a grid step.

[0107] Further, the layer model and the fault model should be corrected with the drilling stratification data and the breakpoint data respectively, and if the structure is complex, the fault surface shape, the fault cutting relationship, and the fault throw and other information should also be corrected, so as to ensure the accuracy of the three-dimensional structure stratigraphic framework model and the accuracy of the position of the drilling in the model.

[0108] The grid step includes a plane step and a vertical step, the plane grid step should be set according to the area of the modeling area and the development degree of the target oil and gas reservoir, and the vertical grid step should be set according to the fluctuation degree of the structure, the formation thickness of the target oil and gas reservoir, and the reservoir complexity.

[0109] In this step S2, the parameter interpretation data of the target oil and gas reservoir is obtained according to the well logging curve data of the target oil and gas reservoir, and the sand shale discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data are obtained according to the obtained parameter interpretation data and the three-dimensional structure stratigraphic framework model, including:

[0110] The parameter interpretation data of the target oil and gas reservoir is obtained according to the well logging curve data of the target oil and gas reservoir, wherein the parameter interpretation data includes physical property parameter interpretation data, lithology parameter interpretation data and mineral parameter interpretation data.

[0111] The obtained parameter interpretation data is dispersed into the three-dimensional structure stratigraphic framework model to obtain the sand shale discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data.

[0112] Further, the obtained parameter interpretation data is dispersed into the three-dimensional structure stratigraphic framework model by using the arithmetic average method to obtain the sand shale known well data (i.e. the sand shale discrete data), the sedimentary facies known well data (i.e. the sedimentary facies discrete data) and the granularity lithofacies known well data (i.e. the granularity lithofacies discrete data) required for modeling.

[0113] Further, the sand shale discrete data is classified and integrated by the interpreted lithology data, for example, the sand-containing rock is classified as sandstone, and the others are classified as mudstone, that is, the sand shale discrete data is also classified as sandstone discrete data and mudstone discrete data.

[0114] The sedimentary facies discrete data is obtained by the interpreted lithology data and physical property data according to the physical property rhythm of sandstone, the vertical sedimentary sequence of sand shale, the spatial distribution pattern of sedimentary facies and other characteristics.

[0115] The granularity lithofacies discrete data is obtained by the interpreted mineral data and lithology data, and is calculated in combination with the electrical logging curves (such as natural gamma, resistivity, acoustic time difference and the like).

[0116] In step S3, the simulation is performed based on the sand shale discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data to obtain the granularity lithofacies three-dimensional geological model, which includes:

[0117] The sand shale three-dimensional geological model is obtained according to the sand shale discrete data.

[0118] The sedimentary facies three-dimensional geological model is obtained according to the sedimentary facies discrete data and the sand shale three-dimensional geological model.

[0119] The shale content three-dimensional geological model is obtained according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model, and specifically, the parameter interpretation data is the mineral interpretation data.

[0120] The granularity lithofacies three-dimensional geological model is obtained according to the granularity lithofacies discrete data, the sedimentary facies three-dimensional geological model and the shale content three-dimensional geological model.

[0121] Specifically, in step S3, the sand shale three-dimensional geological model is obtained according to the sand shale discrete data, which includes:

[0122] The geological statistical analysis is performed on the sand shale discrete data to obtain the variogram parameter data of the sand shale discrete data distribution;

[0123] The sequential indicator random simulation is performed on the target oil and gas reservoir according to the variogram parameter data of the sand shale discrete data distribution to obtain the sand shale three-dimensional geological model.

[0124] Specifically, in step S3, the geological statistical analysis is performed on the sandstone discrete data in the sand shale discrete data to obtain the variogram of the sandstone discrete data distribution, the pre-optimized seismic attribute is taken as a second variable, and the sequential indicator random simulation is performed using the collocated co-Kriging algorithm to obtain the sand shale three-dimensional geological model.

[0125] Further, the variogram parameters of the discrete sandstone and mudstone data distribution mainly include the primary and secondary variable range directions and variable range data, and the variogram parameter retrieval needs to refer to the knowledge of the target reservoir sedimentary system background, mainly including the sediment source direction, sediment body scale, and sediment body spatial form.

[0126] Further, the pre-selected seismic attribute refers to a seismic attribute body with relatively good matching degree with the sandstone and mudstone data of well interpretation, and is mainly used to indicate the probability of the existence of sandstone and mudstone between wells.

[0127] Further, the specific implementation steps of the sequential indicator stochastic simulation are as follows: a, randomly establishing a simulation path; b, performing the coordinated Kriging calculation in the order of the grid in the random path, searching the sandstone and mudstone data and the seismic attribute data around the grid, combining the variogram parameters, solving the Kriging equation, obtaining the calculation result (sandstone or mudstone) of the grid and taking it as the condition data for calculating the next grid; c, repeating a and b until the entire random path is completely calculated.

[0128] In step S3, the sedimentary facies three-dimensional geological model is obtained according to the discrete sedimentary facies data and the sandstone and mudstone three-dimensional geological model, including:

[0129] obtaining the sedimentary facies plane distribution two-dimensional graph of the target oil and gas reservoir according to the discrete sedimentary facies data;

[0130] obtaining the sedimentary facies ratio according to the sandstone and mudstone three-dimensional geological model;

[0131] performing the multiple-point geostatistical simulation on the target oil and gas reservoir according to the sedimentary facies plane distribution two-dimensional graph and the sedimentary facies ratio, to obtain the sedimentary facies three-dimensional geological model.

[0132] Specifically, according to the discrete data of the sedimentary facies, the sedimentary facies plane distribution two-dimensional graph of the target reservoir of the target oil and gas reservoir is drawn in combination with the knowledge of the regional sedimentary background, so as to be used as a training image to control the development probability of each sedimentary facies at a certain position in space, and the sedimentary facies ratio in the sandstone and mudstone is constrained by the sandstone and mudstone three-dimensional geological model, and the multiple-point geostatistical method is used for random simulation to obtain the sedimentary facies three-dimensional geological model.

[0133] Further, the training image (Training Image, TI) is a prior geological pattern, and in the present application, the training image represents the structure, geometric form and distribution pattern of different sedimentary facies. The specific implementation method is to assign the sedimentary facies plane distribution two-dimensional graph to the grid corresponding to the X and Y coordinates according to the discrete attribute data of the sedimentary facies.

[0134] The specific implementation steps of the multiple-point geostatistical simulation are: a, setting a three-dimensional grid, grid sampling a two-dimensional map of sedimentary facies distribution, and constructing a training image model; b, establishing data events in a random sampling manner from the sedimentary facies data obtained in step S2; c, scanning the data events in the training image, and obtaining a multiple-point probability and a conditional probability distribution function at a point to be simulated; d, selecting multiple-point geostatistical modeling algorithms including but not limited to Snesim, Simpat, Deesse, etc. to perform multiple-point geostatistical simulation; e, adjusting different sedimentary facies proportion parameters to obtain multiple simulation implementations, and selecting an optimal simulation result according to standards such as sedimentary facies spatial continuity and similarity to the training image, to obtain an optimal sedimentary facies three-dimensional geological model.

[0135] In step S3, the mud content three-dimensional geological model is obtained according to the sedimentary facies discrete data, the parameter interpretation data, and the sedimentary facies three-dimensional geological model, including:

[0136] According to the sedimentary facies discrete data and the mud content interpretation data in the parameter interpretation data, the spatial data distribution of the mud content in different sedimentary facies is obtained; wherein the parameter interpretation data is mineral parameter interpretation data.

[0137] According to the sedimentary facies three-dimensional geological model and the spatial data distribution of the mud content in the sedimentary facies, sequential Gaussian random simulation is performed on the target oil and gas reservoir to obtain the mud content three-dimensional geological model.

[0138] Specifically, the spatial data distribution of the mud content in the sedimentary facies includes characteristic values, vertical distribution ratios, and variogram parameters of the mud content in different sedimentary facies.

[0139] The characteristic values of the mud content in different sedimentary facies include maximum value, minimum value, average value, variance, coefficient of variation, etc.

[0140] The specific implementation steps of the sequential Gaussian random simulation are: a, randomly establishing a simulation path; b, performing Kriging calculation according to the order of the grid in the random path, searching for the mud content data around the grid, combining the variogram parameters to solve the Kriging equation, obtaining the calculation result (mud content) of the grid and taking it as the conditional data for calculating the next grid; c, repeating a and b until the entire random path is completely calculated.

[0141] In step S3, the grain size lithofacies three-dimensional geological model is obtained according to the grain size lithofacies discrete data, the sedimentary facies three-dimensional geological model, and the mud content three-dimensional geological model, including:

[0142] An optimal grain size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and a three-dimensional inversion body model is obtained according to the optimal grain size lithofacies sensitive well logging curve;

[0143] According to the shale content three-dimensional geological model and the three-dimensional inversion body model, a spatial probability body of each particle size lithofacies is obtained;

[0144] The particle size lithofacies discrete data is subjected to a geostatistical analysis to obtain variogram parameter data of the particle size lithofacies discrete data distribution;

[0145] According to the variogram parameter data of the particle size lithofacies discrete data distribution, the spatial probability body of each particle size lithofacies and the sedimentary facies three-dimensional geological model, a sequential indicator stochastic simulation is performed on the target oil and gas reservoir to obtain a particle size lithofacies three-dimensional geological model.

[0146] Further, the optimal particle size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and a three-dimensional inversion body model is obtained according to the optimal particle size lithofacies sensitive well logging curve, which comprises:

[0147] The particle size lithofacies sensitive well logging curve and the shale content interpretation data in the parameter interpretation data of the participating well of the target oil and gas reservoir are input into the particle size lithofacies classification information neural network model which is pre-trained;

[0148] The particle size lithofacies classification information neural network model outputs the classification information of the particle size lithofacies of the participating well;

[0149] The classification information of the particle size lithofacies of the non-participating well of the target oil and gas reservoir is compared with the classification information of the particle size lithofacies of the participating well to obtain a comparison result;

[0150] The optimal particle size lithofacies sensitive well logging curve is determined in the particle size lithofacies sensitive well logging curve of the participating well according to the comparison result;

[0151] The three-dimensional inversion body model is obtained according to the optimal particle size lithofacies sensitive well logging curve.

[0152] Specifically, the training method of the particle size lithofacies classification information neural network model is:

[0153] The drilling wells of the target oil and gas reservoir are randomly extracted, and the shale content interpretation data and the particle size lithofacies sensitive electric logging curve of the extracted drilling wells of the target oil and gas reservoir are taken as input samples, and the extracted particle size lithofacies classification information is taken as target output samples, and a neural network algorithm is selected for training and learning to obtain the trained particle size lithofacies classification information neural network model.

[0154] The comparison result is obtained by comparing the classification information of the particle size lithofacies of the non-participating well of the target oil and gas reservoir with the classification information of the particle size lithofacies of the participating well, which comprises:

[0155] The particle size lithofacies clustering result identified by the neural network is compared and evaluated by using the particle size lithofacies classification information of the non-participating well, and the sensitive curve or sensitive curve combination with the highest coincidence degree in identifying the particle size lithofacies by the neural network algorithm is selected as the optimal particle size lithofacies sensitive well logging curve.

[0156] Further, the selected neural network algorithm includes but is not limited to a convolutional neural network algorithm, an adversarial neural network algorithm, etc., to quickly identify (training process) multiple information, realize fuzzy classification and estimation method.

[0157] Further, the comparison of the granularity lithofacies neural network clustering results includes the following specific implementation steps: a, statistics of all the granularity lithofacies intermediate depth information and discrete values of the non-participating well interpreted; b, statistics of the granularity lithofacies discrete values identified by the neural network corresponding to each intermediate depth value; c, comparison of the coincidence degree of the two columns of discrete values.

[0158] In step S3, the spatial probability body of each granularity lithofacies is obtained according to the shale content three-dimensional geological model and the three-dimensional inversion body model, including:

[0159] The shale content three-dimensional geological model and the three-dimensional inversion body model are input into the pre-trained granularity lithofacies spatial probability body neural network model;

[0160] The granularity lithofacies spatial probability body neural network model outputs the spatial probability body of each granularity lithofacies.

[0161] Specifically, the shale content three-dimensional geological model and the three-dimensional inversion body model data obtained based on the optimal granularity lithofacies sensitive well curve are used as samples, and a neural network algorithm is used for training and learning, with the granularity lithofacies data at the single well grid participating in supervised training, and clustering according to the existing granularity lithofacies classification information, to establish the conditional probability distribution of the shale content and the sensitive curve of different granularity lithofacies, and obtain the spatial probability body of each granularity lithofacies.

[0162] The three-dimensional inversion body refers to a process of solving and imaging the underground rock properties reflected by the logging curve using the mapping relationship between the logging curve and the seismic reflection, and using seismic data. Common calculation methods include but are not limited to geostatistics, waveform indication, etc.; the three-dimensional inversion body model refers to a deterministic establishment of the three-dimensional model of the inversion body through grid resampling of the inversion results.

[0163] The spatial probability body refers to the probability that a granularity lithofacies may exist at a certain grid in a three-dimensional space. All grids in the space correspond to the distribution probability of one or more granularity lithofacies, and the sum of the probability values is 1.

[0164] The geological statistics analysis of the granularity lithofacies discrete data obtains the variogram parameter data of the granularity lithofacies discrete data distribution; the sequential indicator stochastic simulation of the target oil and gas reservoir is performed according to the variogram parameter data of the granularity lithofacies discrete data distribution, the spatial probability body of each granularity lithofacies, and the sedimentary facies three-dimensional geological model, to obtain the granularity lithofacies three-dimensional geological model, which specifically includes:

[0165] According to the discrete data of the grain size lithofacies, a geostatistical analysis is performed to obtain variogram parameter data, so as to control the distribution direction and scale of each grain size lithofacies, and the spatial probability body of each grain size lithofacies is used to control the development probability of each grain size lithofacies at a certain position in space, and the grain size lithofacies ratio in each sedimentary facies is constrained by the three-dimensional geological model of the sedimentary facies, sequential indicator random simulation is performed by using the collocated cokriging to obtain the three-dimensional geological model of the grain size lithofacies.

[0166] In the process, the variogram parameter retrieval refers to being performed respectively for each type of grain size lithofacies.

[0167] The sequential indicator random simulation refers to being performed respectively in the three-dimensional space occupied by each type of sedimentary facies, and the specific implementation steps are as follows: a, a simulation path is randomly established; b, the collocated cokriging calculation is performed in the order of the grid in the random path, each type of grain size lithofacies data and the corresponding probability body data around the grid are searched, the cokriging equation is solved by combining the variogram parameters, and the calculation result (grain size lithofacies type) of the grid is obtained and used as the condition data for calculating the next grid; c, steps a and b are repeated until the entire random path is completely calculated.

[0168] In step S4, the three-dimensional geological model of the physical property parameters of the target oil and gas reservoir is obtained according to the discrete data of the grain size lithofacies, the parameter interpretation data and the three-dimensional geological model of the grain size lithofacies, and the three-dimensional geological model of the physical property parameters of the target oil and gas reservoir is obtained.

[0169] The spatial data distribution of the physical property parameters in different grain size lithofacies is obtained according to the discrete data of the grain size lithofacies and the physical property parameter interpretation data in the parameter interpretation data.

[0170] The three-dimensional geological model of the physical property parameters is obtained by sequentially performing Gaussian random simulation on the target oil and gas reservoir according to the three-dimensional geological model of the grain size lithofacies and the spatial data distribution of the physical property parameters in different grain size lithofacies.

[0171] The spatial data distribution of the physical property parameters in different grain size lithofacies includes the characteristic values, correlations and variogram parameters of the physical property data in different grain size lithofacies.

[0172] Specifically, the physical property parameters mainly refer to the porosity reflecting the reservoir storage capacity and the permeability reflecting the reservoir flow capacity.

[0173] The three-dimensional geological model of the physical property parameters includes the three-dimensional geological model of the porosity and the three-dimensional geological model of the permeability.

[0174] If there is a suitable rock physical seismic attribute that can match the porosity data interpreted by the well logging, it can be used as a second variable to perform sequential Gaussian random simulation by using the collocated cokriging to obtain the three-dimensional geological model of the porosity.

[0175] The permeability model is preferably executed after the porosity model is established, the porosity model is taken as a second variable, the porosity-permeability correlation in different particle facies is used, and the sequential Gaussian random simulation or double-variable cloud transformation calculation is performed by using the coordinated Kriging.

[0176] The sequential Gaussian random simulation refers to being respectively implemented in the three-dimensional space occupied by each type of particle facies, and the specific implementation steps are as follows: a, a simulation path is randomly established; b, the Kriging calculation is performed in the order of the grid in the random path, if there is a second variable, the coordinated Kriging calculation is performed, the physical property parameter data around the grid is searched, the Kriging equation is solved in combination with the variation function parameter, the calculation result of the grid is obtained and is taken as the condition data for calculating the next grid; c, a and b are repeated until the whole random path is completely calculated.

[0177] In step S5, the high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is determined according to the physical property three-dimensional geological model of the target oil and gas reservoir, and the high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is determined according to the physical property three-dimensional geological model of the target oil and gas reservoir.

[0178] The relationship between the physical property parameters in each particle facies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir is obtained.

[0179] The threshold value of the physical property parameters of the high-quality reservoir according to different particle facies is determined according to the relationship between the physical property parameters in each particle facies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir.

[0180] The grid body satisfying all the physical property parameter threshold values at the same time is calculated according to the particle facies three-dimensional geological model and the physical property parameter three-dimensional geological model.

[0181] The high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir is obtained according to the grid body satisfying all the physical property parameter threshold values at the same time.

[0182] Figure 2 is the structural block diagram of the high-quality reservoir simulation device for tight sandstone provided by the embodiment of the present application. Figure 2As shown, the high-quality reservoir simulation device for compact sandstone provided by the embodiment comprises: a first model construction module, configured to establish a three-dimensional structure stratigraphic framework model of a target oil and gas reservoir according to three-dimensional seismic data; a data acquisition module, configured to acquire parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtain sandstone and mudstone discrete data, sedimentary facies discrete data and grain size lithofacies discrete data according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure stratigraphic framework model; a first simulation module, configured to simulate based on the sandstone and mudstone discrete data, the sedimentary facies discrete data and the grain size lithofacies discrete data, and obtain a grain size lithofacies three-dimensional geological model; a second simulation module, configured to simulate according to the parameter interpretation data of the target oil and gas reservoir, the grain size lithofacies discrete data and the grain size lithofacies three-dimensional geological model, and obtain a physical property parameter three-dimensional geological model of the target oil and gas reservoir; and a second model construction module, configured to determine a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the physical property three-dimensional geological model of the target oil and gas reservoir, wherein the high-quality reservoir three-dimensional distribution model is used to describe the high-quality reservoir of the target oil and gas reservoir.

[0183] The first model construction module is specifically configured to: perform fine structure interpretation on the target oil and gas reservoir by using the three-dimensional seismic data, and acquire horizon and fault data of the target oil and gas reservoir.

[0184] The three-dimensional structure stratigraphic framework model is established according to the horizon and fault data of the target oil and gas reservoir.

[0185] Specifically, fine structure interpretation is performed on the target oil and gas reservoir by using the three-dimensional seismic data, horizon and fault data of the target oil and gas reservoir are acquired, a horizon model and a fault model are created, a modeling area and a grid step are set, and the three-dimensional structure stratigraphic framework model is established.

[0186] Further, the horizon model and the fault model should be corrected with drilling stratification data and breakpoint data respectively, and if the structure is complex, the fault surface shape, fault cutting relationship and fault throw and the like should also be corrected, so as to ensure the accuracy of the three-dimensional structure stratigraphic framework model and the accuracy of the position of the drilling in the model.

[0187] The grid step comprises a plane step and a vertical step, the plane grid step should be set according to the area of the modeling area and the development degree of the target oil and gas reservoir, and the vertical grid step should be set according to the fluctuation degree of the structure, the formation thickness of the target oil and gas reservoir and the reservoir complexity and the like.

[0188] The data acquisition module is specifically configured to: acquire parameter interpretation data of the target oil and gas reservoir according to logging curve data of the target oil and gas reservoir, wherein the parameter interpretation data comprises physical property parameter interpretation data, lithology parameter interpretation data and mineral parameter interpretation data.

[0189] Disperse the obtained parameter interpretation data into the three-dimensional tectonic stratigraphic framework model to obtain sand shale discrete data, sedimentary facies discrete data and granularity lithofacies discrete data.

[0190] Further, disperse the obtained parameter interpretation data into the three-dimensional tectonic stratigraphic framework model by the method of arithmetic average to obtain sand shale known well data (i.e. sand shale discrete data), sedimentary facies known well data (i.e. sedimentary facies discrete data) and granularity lithofacies known well data (i.e. granularity lithofacies discrete data) required for modeling.

[0191] Further, the sand shale discrete data is obtained by classifying and integrating the interpreted lithology data, for example, sand-containing rocks are classified as sandstone and other rocks are classified as mudstone, i.e. the sand shale discrete data is classified into sandstone discrete data and mudstone discrete data.

[0192] The sedimentary facies discrete data is obtained by the interpreted lithology data and physical property data according to the physical property rhythm of sandstone, the vertical sedimentary sequence of sandstone and mudstone and the spatial distribution pattern of sedimentary facies.

[0193] The granularity lithofacies discrete data is obtained by the interpreted mineral data and lithology data in combination with electrical logging curves (such as natural gamma, resistivity, acoustic time difference, etc.).

[0194] The first simulation module is specifically configured to obtain a sand shale three-dimensional geological model according to the sand shale discrete data.

[0195] Obtain a sedimentary facies three-dimensional geological model according to the sedimentary facies discrete data and the sand shale three-dimensional geological model.

[0196] Obtain a shale content three-dimensional geological model according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model, and specifically, the parameter interpretation data is mineral parameter interpretation data.

[0197] Obtain a granularity lithofacies three-dimensional geological model according to the granularity lithofacies discrete data, the sedimentary facies three-dimensional geological model and the shale content three-dimensional geological model.

[0198] Specifically, the sand shale three-dimensional geological model is obtained according to the sand shale discrete data, and the method comprises the following steps.

[0199] Perform a geostatistical analysis on the sand shale discrete data to obtain variogram parameter data of the sand shale discrete data distribution;

[0200] Perform sequential indicator simulation on a target oil and gas reservoir according to the variogram parameter data of the sand shale discrete data distribution to obtain the sand shale three-dimensional geological model.

[0201] Specifically, the sandstone discrete data in the sand shale discrete data is subjected to a geostatistical analysis to obtain a variogram of the sandstone discrete data distribution, and according to the variogram of the sandstone discrete data distribution and the pre-selected seismic attribute as a second variable, a sequential indicator random simulation is performed using a co-Kriging algorithm to obtain the sand shale three-dimensional geological model.

[0202] Further, the variogram parameters of the sand shale discrete data distribution mainly include the primary and secondary range directions and range data, and the variogram parameter retrieval needs to refer to the knowledge of the target reservoir sedimentary system background, mainly including the sediment source direction, sediment body scale, and sediment body spatial form.

[0203] Further, the pre-selected seismic attribute refers to a seismic attribute body with a relatively good matching degree with the sand shale data interpreted from the well logs, and is mainly used to indicate the probability of the existence of sand shale between wells.

[0204] Further, the specific implementation steps of the sequential indicator random simulation are as follows: a, randomly establishing a simulation path; b, performing co-Kriging calculation in the order of the grid in the random path, searching for the sand shale data and seismic attribute data around the grid, combining the variogram parameters to solve the Kriging equation, obtaining the calculation result (sandstone or mudstone) of the grid and taking it as the condition data for calculating the next grid; c, repeating a and b until the entire random path is completely calculated.

[0205] The sand shale three-dimensional geological model is obtained according to the sedimentary facies discrete data and the sand shale three-dimensional geological model, and includes:

[0206] A sedimentary facies planar distribution two-dimensional map of the target oil and gas reservoir is obtained according to the sedimentary facies discrete data.

[0207] A sedimentary facies ratio is obtained according to the sand shale three-dimensional geological model.

[0208] A sedimentary facies three-dimensional geological model is obtained by performing a multiple-point geostatistical simulation on the target oil and gas reservoir according to the sedimentary facies planar distribution two-dimensional map and the sedimentary facies ratio.

[0209] Specifically, according to the sedimentary facies discrete data and the knowledge of the regional sedimentary background, a sedimentary facies planar distribution two-dimensional map of the target reservoir of the target oil and gas reservoir is drawn as a training image to control the development probability of each sedimentary facies at a certain position in space, and the sedimentary facies ratio in the sandstone and mudstone is constrained by the sand shale three-dimensional geological model, and a multiple-point geostatistical method is used for random simulation to obtain the sedimentary facies three-dimensional geological model.

[0210] Further, the training image (TI) is a priori geological model, in the present application, the training image represents the structure, geometry and distribution pattern of different sedimentary facies. The specific implementation method is to assign the sedimentary facies planar distribution two-dimensional graph to the grid corresponding to the X, Y coordinates according to the sedimentary facies discrete attribute data.

[0211] The specific implementation steps of the multiple-point geostatistical simulation are: a, setting a three-dimensional grid, grid sampling the sedimentary facies planar distribution two-dimensional graph to construct a training image model; b, establishing data events in a random sampling manner according to the sedimentary facies data obtained in step S2; c, scanning the data events in the training image, and obtaining the multiple-point probability and the conditional probability distribution function at the point to be simulated; d, selecting multiple-point geostatistical modeling algorithms including but not limited to Snesim, Simpat, Deesse and the like to perform multiple-point geostatistical simulation; e, adjusting different sedimentary facies proportion parameters to obtain multiple simulation implementations, and selecting the optimal simulation result according to the sedimentary facies spatial continuity, the similarity degree with the training image and the like to obtain the optimal sedimentary facies three-dimensional geological model.

[0212] According to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model, the mud content three-dimensional geological model is obtained, which includes:

[0213] According to the mud content interpretation data in the sedimentary facies discrete data and the parameter interpretation data, the spatial data distribution of the mud content in different sedimentary facies is obtained; specifically, the parameter interpretation data is mineral parameter interpretation data.

[0214] According to the sedimentary facies three-dimensional geological model and the spatial data distribution of the mud content in the sedimentary facies, the target oil and gas reservoir is sequentially and randomly simulated to obtain the mud content three-dimensional geological model.

[0215] Specifically, the spatial data distribution of the mud content in the sedimentary facies includes characteristic values, vertical distribution ratios and variogram parameters of the mud content in different sedimentary facies.

[0216] The characteristic values of the mud content in different sedimentary facies include maximum value, minimum value, average value, variance, coefficient of variation and the like.

[0217] The specific implementation steps of the sequential Gaussian random simulation are: a, randomly establishing a simulation path; b, performing Kriging calculation according to the order of the grid in the random path, searching the mud content data around the grid, combining the variogram parameters to solve the Kriging equation, obtaining the calculation result (mud content) of the grid and taking it as the conditional data for calculating the next grid; c, repeating a and b until the entire random path is completely calculated.

[0218] The particle size lithofacies three-dimensional geological model is obtained according to the particle size lithofacies discrete data, the sedimentary facies three-dimensional geological model and the argillan content three-dimensional geological model, and includes:

[0219] The optimal particle size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and a three-dimensional inversion body model is obtained according to the optimal particle size lithofacies sensitive well logging curve.

[0220] According to the argillan content three-dimensional geological model and the three-dimensional inversion body model, a spatial probability body of each particle size lithofacies is obtained.

[0221] The particle size lithofacies discrete data is subjected to a geostatistical analysis to obtain variogram parameter data of the particle size lithofacies discrete data distribution.

[0222] The target oil and gas reservoir is subjected to sequential indicator simulation according to the variogram parameter data of the particle size lithofacies discrete data distribution, the spatial probability body of each particle size lithofacies and the sedimentary facies three-dimensional geological model, so as to obtain the particle size lithofacies three-dimensional geological model.

[0223] Further, the optimal particle size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and the three-dimensional inversion body model is obtained according to the optimal particle size lithofacies sensitive well logging curve, and includes:

[0224] The particle size lithofacies sensitive well logging curve of the participating well of the target oil and gas reservoir and the argillan content interpretation data in the parameter interpretation data are input into the particle size lithofacies classification information neural network model which is pre-trained.

[0225] The particle size lithofacies classification information neural network model outputs the classification information of the particle size lithofacies of the participating well.

[0226] The classification information of the particle size lithofacies of the non-participating well of the target oil and gas reservoir is compared with the classification information of the particle size lithofacies of the participating well to obtain a comparison result.

[0227] The optimal particle size lithofacies sensitive well logging curve is determined in the particle size lithofacies sensitive well logging curve of the participating well according to the comparison result.

[0228] The three-dimensional inversion body model is obtained according to the optimal particle size lithofacies sensitive well logging curve.

[0229] Specifically, the training method of the particle size lithofacies classification information neural network model is:

[0230] The drilling of the target oil and gas reservoir is randomly extracted, and the argillan content interpretation data and the particle size lithofacies sensitive electric logging curve of the extracted target oil and gas reservoir drilling are taken as input samples, the extracted particle size lithofacies classification information is taken as a target output sample, a neural network algorithm is selected for training and learning, and a trained particle size lithofacies classification information neural network model is obtained.

[0231] The granularity lithofacies classification information of the non-participating well is compared with the granularity lithofacies classification information of the participating well according to the target oil and gas reservoir, and a comparison result is obtained, which comprises:

[0232] The granularity lithofacies clustering result identified by the neural network is compared and evaluated by using the granularity lithofacies classification information of the non-participating well, and the sensitive curve or sensitive curve combination with the highest coincidence degree in identifying the granularity lithofacies by using the neural network algorithm is selected as the optimal granularity lithofacies sensitive logging curve.

[0233] Further, the neural network algorithm selected includes but is not limited to a convolutional neural network algorithm and an adversarial neural network algorithm, which is used to quickly identify (in a training process) multiple information, and realizes a fuzzy classification and estimation method.

[0234] Further, the comparison of the granularity lithofacies neural network clustering result comprises the following steps: a, statistics the intermediate depth information and the discrete value of all the granularity lithofacies explained by the non-participating well; b, statistics the granularity lithofacies discrete value identified by the neural network corresponding to each intermediate depth value; c, compare the coincidence degree of the two columns of discrete values.

[0235] The spatial probability body of each granularity lithofacies is obtained according to the argillan content three-dimensional geological model and the three-dimensional inversion body model, which comprises:

[0236] The argillan content three-dimensional geological model and the three-dimensional inversion body model are input into the pre-trained granularity lithofacies spatial probability body neural network model;

[0237] The granularity lithofacies spatial probability body neural network model outputs the spatial probability body of each granularity lithofacies.

[0238] Specifically, the argillan content three-dimensional geological model and the three-dimensional inversion body model data obtained based on the optimal granularity lithofacies sensitive logging curve are used as samples, and a neural network algorithm is used for training and learning, so that the granularity lithofacies data at a single well grid participates in supervised training, and clustering is performed according to the existing granularity lithofacies classification information, the conditional probability distribution of the argillan content and the sensitive curve of different granularity lithofacies is established, and the spatial probability body of each granularity lithofacies is obtained.

[0239] The three-dimensional inversion body refers to a process of solving and imaging the underground rock properties reflected by the logging curve by using the mapping relationship between the logging curve and the seismic reflection and using seismic data. Common calculation methods include but are not limited to geostatistics and waveform indication; the three-dimensional inversion body model refers to a deterministic establishment of the three-dimensional model of the inversion body by using grid resampling of the inversion result.

[0240] The spatial probability body refers to the probability of the existence of the granular lithofacies at a certain grid in the three-dimensional space, and the distribution probability of one or more granular lithofacies in all grids in the space is simultaneously corresponding, and the sum of the probability values is 1.

[0241] The granular lithofacies discrete data is subjected to the geostatistical analysis to obtain the variogram parameter data of the distribution of the granular lithofacies discrete data, and the target oil and gas reservoir is subjected to the sequential indicator random simulation according to the variogram parameter data of the distribution of the granular lithofacies discrete data, the spatial probability body of each granular lithofacies, and the sedimentary facies three-dimensional geological model to obtain the granular lithofacies three-dimensional geological model, and specifically:

[0242] The granular lithofacies discrete data is subjected to the geostatistical analysis to obtain the variogram parameter data, so as to control the distribution direction and scale of each granular lithofacies, the spatial probability body of each granular lithofacies is used to control the development probability of each granular lithofacies at a certain position in space, and the ratio of the granular lithofacies in each sedimentary facies is constrained by using the sedimentary facies three-dimensional geological model, the sequential indicator random simulation is performed by using the co-Kriging to obtain the granular lithofacies three-dimensional geological model.

[0243] The variogram parameter retrieval refers to the implementation for each type of granular lithofacies.

[0244] The sequential indicator random simulation refers to the implementation in the three-dimensional space occupied by each type of sedimentary facies, and the specific implementation steps are as follows: a, a simulation path is randomly established; b, the co-Kriging calculation is performed in the order of the grid in the random path, each type of granular lithofacies data and the corresponding probability body data around the grid are searched, the Kriging equation is solved by combining the variogram parameter to obtain the calculation result (the type of the granular lithofacies) of the grid and take it as the condition data for calculating the next grid; c, the steps a and b are repeated until the entire random path is completely calculated.

[0245] The second simulation module is specifically used for obtaining the spatial data distribution of the physical property parameters in different granular lithofacies according to the granular lithofacies discrete data and the physical property parameter interpretation data in the parameter interpretation data.

[0246] The target oil and gas reservoir is subjected to the sequential Gaussian random simulation according to the granular lithofacies three-dimensional geological model and the spatial distribution of the physical property parameters in different granular lithofacies to obtain the physical property parameter three-dimensional geological model.

[0247] The spatial data distribution of the physical property parameters in different granular lithofacies includes the characteristic value, correlation and variogram parameter of the physical property data in different granular lithofacies.

[0248] Specifically, the physical property parameters mainly refer to the porosity reflecting the reservoir storage capacity and the permeability reflecting the reservoir flow capacity.

[0249] The physical property parameter three-dimensional geological model comprises a porosity three-dimensional geological model and a permeability three-dimensional geological model.

[0250] If a suitable petrophysical seismic attribute can match the porosity data interpreted from the well logging, the attribute can be used as a second variable to perform sequential Gaussian random simulation by using the co-Kriging to obtain the porosity three-dimensional geological model.

[0251] The permeability model is preferably executed after the porosity model is established, and the porosity model can be used as a second variable to perform sequential Gaussian random simulation by using the co-Kriging or double-variable cloud transformation calculation by using the porosity-permeability correlation in different grain size lithofacies.

[0252] The sequential Gaussian random simulation refers to being respectively implemented in the three-dimensional space occupied by each type of grain size lithofacies, and the specific implementation steps are as follows: a, randomly establishing a simulation path; b, performing Kriging calculation in the order of the grid in the random path, and performing co-Kriging calculation if there is a second variable, searching for the physical property parameter data around the grid, combining the variogram parameters to solve the Kriging equation, obtaining the calculation result of the grid and taking the result as the condition data for calculating the next grid; c, repeating a and b until the entire random path is completely calculated.

[0253] The second model construction module is specifically used for: obtaining the relationship between the physical property parameter in each grain size lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir;

[0254] According to the relationship between the physical property parameter in each grain size lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir, the threshold value of the physical property parameter of the high-quality reservoir for different grain size lithofacies is determined;

[0255] According to the grain size lithofacies three-dimensional geological model and the physical property parameter three-dimensional geological model, the grid body that simultaneously meets all the physical property parameter threshold values of each grain size lithofacies is calculated.

[0256] According to the grid body that simultaneously meets all the physical property parameter threshold values of each grain size lithofacies, the three-dimensional distribution model of the high-quality reservoir of the target oil and gas reservoir is obtained.

[0257] The embodiment specifically provides that the test area is a Xinchang structural belt in the western Sichuan depression of the Sichuan Basin, and the target oil and gas reservoir is a tight sandstone gas reservoir of the Triassic Xujiahe Formation. The western Sichuan depression is located in the western Sichuan Basin, has a northeast-stretching strip shape, is bordered by the Longmen Mountain structural belt in the west and the Kunlun-Qinling structural belt in the northeast, and has an area of more than 50,000 square kilometers. The western Sichuan depression is located at several plate junctions in the Late Triassic period, is a very active tectonic zone, and is a superimposed basin formed since the Late Triassic. The Longmen Mountain orogenic belt and the Micang Mountain-Daba Mountain affect the secondary structural form and sedimentary filling evolution of the western Sichuan depression. The Xinchang structural belt is located in the middle segment of the western Sichuan depression, is formed in the early Indosinian movement, and has a whole near-east-west trending long-fence anticline in structure.

[0258] In the early deposition of the Triassic Xujiahe Formation, with the bidirectional subduction of the Yangtze plate to the Qiangtang plate and the Kunlun plate, the Micang Mountain and the Daba Mountain paleo-land in the northeast of the Sichuan Basin appeared, and the Longmen Mountain in the west also began to thrust and uplift. The north segment of the Longmen Mountain gradually protruded and exposed the water surface, and the western Sichuan foreland basin began to form. During the deposition of the Xujiahe Formation, the tectonic movement in the east-west direction of the Xinchang structural belt further intensified, the western Sichuan area greatly subsided, the depression was formed along the front of the Longmen Mountain, and the depression center was located in the Pengzhou-Anxian area. The Xujiahe Formation in the Xinchang area is a tectonic activity period, the Longmen Mountain and the Micang Mountain-Daba Mountain have water systems input, and a sedimentary system mainly developed by braided river delta is mainly developed, the sand body has large thickness and wide distribution, and is an important reservoir of the Xujiahe gas reservoir in the Xinchang area. The Xujiahe Formation reservoir in the western Sichuan depression is generally gas-bearing, and the previous research shows that the wide development of sand bodies with good porosity and permeability (high-quality reservoirs) is the basis for stable production of oil and gas and formation of gas reservoirs with high economic exploitation value. In order to establish the attribute parameter model of porosity and permeability, the quality of the sand body needs to be measured by a facies model, and the three-dimensional spatial distribution of the sand body needs to be controlled.

[0259] The high-quality reservoir model of the gas reservoir is established by using the invention.

[0260] (1) The depth domain horizon interpretation data of the Xujiahe Formation is obtained by time-depth conversion through the time-depth conversion relationship of multiple seismic interpretation layers of the whole Xujiahe Formation. The horizon data of the sand group top surface interpreted by the seismic is further corrected by using the fine stratigraphic framework built by the drilling and the actual stratification information of the drilling to obtain the horizon data and structure map of each sand group. The longitudinal horizon includes 10 sand groups and 11 layers of TX2-1 sand group-TX2-10 sand group. The fault depth of each fault at different positions, the fault occurrence and the combination relationship between the faults are finely analyzed and described.

[0261] The fault scale of different faults is controlled by fault throw parameter, and the deformation degree of different sand groups is controlled by the layer displacement parameter contained in the original stratigraphic model, and the target oil and gas reservoir present structure-stratigraphic model is created. The obtained three-dimensional structure-stratigraphic model is gridded with a plane grid accuracy of 100m x 100m. In the vertical direction, the 10 sand groups of Xujiahe Formation are divided into 300 grids. Considering that TX2-2 and TX2-4 sand groups in the study area are the main producing layers, higher vertical grid accuracy is given, which is 60 and 80 grids respectively, with an accuracy of about 1m per grid. The grid thickness of other sand groups is given according to the contribution of gas production, which is about 2m per grid on average, to ensure that the reservoir heterogeneity in the vertical direction is reflected to the greatest extent in the three-dimensional grid. The total grid of the model is about 32.11 million, which provides a carrier for the following various attribute models.

[0262] (2) Using cable logging curve data, according to the characteristics of mineral composition in the study area, through the analysis of the adaptability of the evaluation model and the results, various interpretation models are tried, and the optimized method based on the rock physical volume model is selected to carry out the simultaneous solution of mineral composition and porosity, and the accurate and reliable mineral and physical property evaluation results are obtained.

[0263] According to the lithology, sedimentary characteristics and resolution of logging curves in the study area, the natural gamma ray and acoustic time curve logging response sequence is selected for the logging facies analysis of the main layer of Xujiahe Formation in Xinchang. After accurate observation and description of the core, the logging microfacies template is summarized by corresponding the core microfacies and logging curves, and the main sedimentary microfacies types are comprehensively identified from the aspects of logging curve shape, sedimentary lithology and sedimentation, and the logging microfacies template of the study area is obtained. According to the characteristics of petrology, rock combination, biological fossils, phase sequence structure and sand body development position, the Xujiahe Formation reservoir is further divided into 4 kinds of sedimentary microfacies, which are distributary channel, channel margin, interdistributary bay and river mouth bar. Among them, the superimposed channel of braided river delta plain / frontal is the most developed sedimentary microfacies type in the study area.

[0264] The lithology of the second member of Xujiahe Formation in the new field is characterized by interbedding of sandstone and mudstone, and coal seam is occasionally seen. According to the grain size classification limit in the industry standard of grain size analysis method (SY / T 5434-1999), the sandstone of the second member of Xujiahe Formation can be divided into four types of grain size lithofacies, i.e. coarse sandstone (grain size 0.5-1mm), medium sandstone (grain size 0.25-0.5mm), fine sandstone (grain size 0.0625-0.25mm) and siltstone (grain size 0.0039-0.0625mm). The relationship between thin section grain size analysis, core grain size description, logging grain size description and logging data shows that the logging response of the second member of Xujiahe Formation is obviously related to thin section grain size, and with the change of grain size from coarse to fine, the values of natural gamma ray, neutron porosity and density logging increase. According to the statistical results of core and logging interpretation, the average porosity of coarse sandstone is 4.8%, the average porosity of medium sandstone is 4.2%, and the average porosity of fine sandstone is 3.3%. The porosity of different grain size lithology / lithofacies is obviously different. The medium-coarse grain sandstone is the favorable lithofacies, and is the main development area of high quality reservoir, and the fine sandstone and siltstone have obviously poor physical properties.

[0265] The discrete data of acquired lithology, sedimentary microfacies, grain size lithofacies, etc. are discretely put into the already built structure-stratigraphic model grid by using the arithmetic mean method, so as to provide condition data for the modeling of high quality reservoir and attribute parameters.

[0266] (3) Using the preferred impedance seismic attribute, combined with core observation and drilling data, the sedimentary filling characteristics of Xujiahe 2 member in the new field structural belt are analyzed. The sandstone and mudstone obtained in step 2 are taken as the object, which is closely related to the sediment source direction. The sandstone extension direction is consistent with the river-delta progradation direction, and the experimental variogram is established. The internal sedimentary sand body of Xujiahe 2 member is overlaid by multiple stages of river channel in the longitudinal direction, and the river channel is distributed in strips with different scales from west to east in the plane. Combined with the seismic attribute and the drilling sand body thickness, it is known that the delta sand body of Xujiahe 2 lower submember is mainly located in the west, mainly supplied by the short axis source of Longmen Mountain, the sand body thickness is thin in the edge and thick in the east, the main variable range direction is between 130°-165°, and the secondary variable range direction is between 40°-75°. The sand body of Xujiahe 2 middle submember is developed, TX2-6 and TX2-5 sand groups are still mainly supplied by the northwest direction source, but the main variable range direction is between 150°-175°, and the secondary variable range direction is between 60°-85°, both of which are larger. In addition, the river channel scale of TX2-4 sand group is the largest during the sedimentary period, and the drilling reveals that the sand body thickness in the whole region basically reaches more than 60 meters. Because the sand body supply of the northwest and northeast directions is very strong, the variogram is not set to participate in modeling; Xujiahe 2 upper submember is mainly supplied by the long axis source of the northeast direction, the east sand body is more developed, and the west sand body is obviously less than that of the middle and lower submembers, the main variable range direction is between 40°-65°, and the secondary variable range direction is between 130°-155°. For example, the river source of TX2-2 sand group further changes the direction to the east, and is distributed in the northeast-southwest direction. Because the source transport distance is farther, the main variable range of the variogram increases, and the sand body evolution becomes mainly single river channel deposition, the secondary variable range decreases, and the ratio of the main and secondary variable ranges is greater than that of the sand body of Xujiahe 2 lower submember (such as Table 1).

[0267] Table 1 Parameter table of sandstone variogram of each layer in the example research area

[0268]

[0269] The main and secondary variable ranges of each sand group in the plane are spherical models, and the vertical variogram is calculated by using the sand body thickness data obtained by drilling interpretation, still using the spherical model. When fitting the variogram, the nugget effect in the horizontal and vertical directions is 0.005. The variogram parameter used is:

[0270]

[0271] In the formula, γ(h) is the variogram, which reflects the spatial variation degree of the regional variable (such as the sandstone or mudstone simulated herein) with the distance, and has no unit; h is the relative distance (lag) between the two points participating in the calculation, with the unit of m; C represents the value of γ(h) when h is greater than the variable range, which reflects the total variability of the variable in space, and has no unit; a is the variable range, which represents the range of the correlation of the regional variable in space, with the unit of m.

[0272] The subsurface deposit is considered as a whole, two lithofacies types of sandstone and mudstone are divided, well-logging interpreted sand body data is used as hard data, and preferred seismic attributes are used as the second variable to perform sequential indicator modeling by using co-Kriging. The specific simulation process is as follows: ① the well-logging interpreted lithology data (hard data) is coarsened by using the "dominance method" according to the set grid, and the preferred wave impedance seismic attribute volume (second variable) is normalized and coarsened by using the average method according to the set grid; ② the percentage of sandstone and mudstone before and after coarsening is compared to test whether the data coarsening result well preserves the original data characteristics; ③ a random path is established for the space grid without hard data in the first step simulation grid; ④ the co-Kriging calculation is performed according to the order of the grid in the random path, the hard data and the second variable around the grid are searched, the Kriging equation is solved combined with the variation function parameters, the calculation result of the grid is obtained and used as the hard data for calculating the next grid, until the whole random path is completely calculated; ⑤ steps ③ and ④ are repeated to establish multiple simulation implementations; ⑥ the percentage data of sandstone occupied by the simulation implementations are compared to screen the simulation implementations, the most representative result screened out is averaged and attributed to the representative data of sandstone and mudstone, and the simulation result obtained is as shown in FIG. 2. Figure 3

[0273] (4) Based on the sandstone and mudstone facies model obtained in step three, a multi-level facies control is adopted to simulate the sedimentary microfacies as a secondary facies. In order to more reasonably establish the sedimentary facies model, more reliably control the shale content model, and more accurately establish the grain size lithofacies model in the later stage, the multiple-point geostatistical modeling method is selected to perform. The multiple-point statistical geostatistical modeling method uses "training image" to replace the variation function in the two-point statistical modeling method to represent the spatial structure and variation of the geological variable, which can overcome the deficiency of the traditional two-point statistical geostatistical modeling method that cannot well reproduce the spatial geometry of the geological body. At the same time, the multiple-point statistical geostatistical modeling method uses a sequential simulation process based on a pixel, rather than an iterative trial-and-error simulation process based on a target, which is easy to condition well data and other geological information, improves the calculation efficiency, and in addition to the use of "training image", there is no need to provide the geometric parameters of the target body, which overcomes the deficiency of the random simulation method based on the target. As for the Xujiahe gas reservoir in the New Field area in the embodiment, the multiple-point geostatistical method model has the advantages of using training image to replace variation function, the model completely conforms to the interpreted data of single well sedimentary microfacies, and the distribution between different sedimentary microfacies can reflect the overall sedimentary characteristics and evolution law.

[0274] ​Because of the sequence variation of sedimentary microfacies in each sand group of Xu-2 member in Xinchang area is not smooth, the traditional multiple-point geostatistics algorithm (such as Snesim algorithm) will cause the unreasonable contact relationship of simulated sedimentary microfacies in space and the discontinuity of simulation target. In order to solve the problems of training image smoothness, reservoir shape reproduction, multi-scale geological body reproduction, and calculation efficiency improvement, multiple-point geostatistics has gradually developed many new modeling algorithms. In view of the problem of non-smoothness of sedimentary microfacies under the background of delta plain-delta front deposition of Xu-2 member in Xinchang area, this paper applies PVDsim algorithm based on pattern vector distance method to simulation. On the basis of training pattern and data event vector distance calculation, PVDsim algorithm uses secondary matching method to measure the similarity of data event and training pattern through vector distance, and finally determines the training pattern with the highest similarity to data event. This algorithm significantly reduces the uncertainty of training pattern selection, and the built model is more consistent with geological understanding.

[0275] According to the drilling sedimentary microfacies interpretation data, the planar sedimentary microfacies distribution of each sand group is taken as the training image during the simulation of the sand group. The multiple-point geostatistics method based on PVDsim algorithm is used to simulate the quantitative distribution of sedimentary microfacies in three-dimensional space by sand group. The results show that the model not only conforms to the single well microfacies interpretation, but also conforms to the spatial structure between different sedimentary microfacies expressed by training image between wells, and maintains the sedimentary continuity in the vertical direction. In the sand facies model, it basically presents the continuous distribution of superimposed channel-channel edge-interdistributary bay (such as Figure 4 )。

[0276] (5) By analyzing the relationship between drilling shale content interpretation curve and sandstone and mudstone as well as different lithofacies, it is found that from medium-coarse sandstone facies to fine sandstone and then to siltstone facies, the shale content presents an obvious change rule from less to more, and shale content is more obvious for distinguishing different lithofacies types (such as Figure 5 ). The shale content distribution model in three dimensions is constructed by using the shale content interpreted from well logging in step two. The shale content parameter still needs to find a related attribute, i.e. the second variable, to perform spatial constraint during simulation, mainly to increase the control conditions of shale content simulation. Based on the sedimentary microfacies interpretation obtained in step two and the sedimentary microfacies geological model description obtained in step four, the shale content distribution in different sedimentary facies is further statistically described, and the control effect of sedimentary facies on shale content is analyzed from the aspects of plane and vertical, to carry out the shale content distribution model controlled by sedimentary facies.

[0277] The sedimentary facies plane restricts the distribution of the shale content. The distribution range and data pattern of the shale content are controlled by the sedimentary microfacies. For example, in the interdistributary bay facies belt, the average value of the shale content is 0.32, while in the superimposed main channel sedimentary facies belt, the average value of the shale content is 0.09. In the whole delta front sedimentary environment, the shale content gradually decreases from the prodelta to the interdistributary bay to the channel edge to the main channel. The sedimentary facies longitudinally restricts the distribution of the shale content. The distribution trend of the shale content is controlled by the sedimentary microfacies. For example, in the channel facies, the shale content increases from bottom to top due to the positive rhythm characteristics. Based on each sedimentary facies sand body such as the channel or the mouth bar, the three-dimensional geological model of the sedimentary microfacies is used as the target to perform the longitudinal shale content distribution restriction through the corresponding longitudinal shale content variation trend.

[0278] The variogram is the most critical parameter in the geological statistics simulation, and determines the distribution direction and scale of the simulation target. In the shale content data analysis, the variogram fitting is performed from the primary variable range, the secondary variable range and the vertical variable range, which are used as the input conditions for the random simulation. Under the constraint control of the sedimentary facies model, the sequential Gaussian simulation is used to establish the shale content distribution model. During the simulation, the overall proportion of the shale content of each sand group needs to be adjusted to ensure that the spatial distribution of the shale content is consistent with the lithology variation trend reflected by the sedimentary facies (as shown in the accompanying drawings Figure 6 ).

[0279] (6) The concept of the neural network algorithm is to simulate the fast identification (training process) of the human brain to the multiple phenomena and multiple signals to realize the fuzzy classification and estimation, such as the differentiation of multiple objects with similar colors and similar shapes. In the execution step two, the gamma ray data has great reference value for the division of the grain size lithofacies in the logging interpretation process. In addition, the data distribution between wells needs to be established for the data samples participating in the training, so that the neural network clustering can be implemented in the three-dimensional space. According to the present technical method, the shale content data and the gamma ray data are selected to perform the neural network clustering analysis, and the convolution algorithm is used. The participating wells are randomly extracted, the shale content and the gamma ray data of the wells are selected as the samples based on the two-dimensional logging curve data of the wells, the neural network training and learning are performed, and the output result is set to four discrete data. During the training process, the grain size lithofacies classification information (i.e. four lithofacies types: coarse sandstone, medium sandstone, fine sandstone and mudstone) of the participating wells is involved in the supervision.

[0280] The grain size lithofacies data obtained by the single-well neural network clustering are compared with the grain size lithofacies data obtained by the original logging interpretation, and the matching degree between the two is compared by using the median method. For the participating wells and the non-participating wells, the similarity comparison is performed respectively, and the similarity degree of the two grain size lithofacies is tested to be more than 70% (as shown in the accompanying drawings Figure 7The results show that the neural network clustering effect of the grain size lithofacies is better using the shale content and natural gamma data, and the feasibility of the method is also confirmed.

[0281] The single-well-based neural network analysis method is extended to three-dimensional space, and the natural gamma inversion body is obtained through geophysical inversion method and resampled into the structure-stratigraphic model grid obtained in step one. The shale content model and the natural gamma inversion body model obtained in step five are used as input data, and the grain size lithofacies data at the single-well grid obtained in step two are used as supervision. The supervised convolutional neural network learning and training are performed according to the method in step six, and the output results are set to four discrete data corresponding to four lithofacies types (coarse sandstone, medium sandstone, fine sandstone and mudstone), and the spatial distribution data of the probability of each type of discrete phase (such as Table 2).

[0282] Table 2 Grain size lithofacies three-dimensional neural network clustering cross probability data of the example research area

[0283]

[0284]

[0285] On the basis of the three-dimensional shale content model and the three-dimensional natural gamma inversion body model, the corresponding shale content and natural gamma data "window" is given through the distribution of shale content and natural gamma in each lithofacies, and the cross verification of lithofacies interpreted from drilling and shale content and natural gamma is performed. The neural network clustering of the shale content model and the natural gamma inversion body is performed to establish the conditional probability distribution of shale content and natural gamma of different lithofacies, and the spatial probability body distribution of each lithofacies is obtained (Appendix Figure 8 ).

[0286] (8) The conditional data used for the grain size lithofacies modeling is the discrete data of the grain size lithofacies interpreted from the single well obtained in step two. The data analysis of the grain size lithofacies is an important basis for determining the simulation method and controlling the simulation results, and it is statistically analyzed from multiple angles.

[0287] ① The longitudinal ratio distribution of the grain size lithofacies is set. In the entire Xujiahe Formation, the longitudinal ratio distribution of the lithofacies is mainly used to reflect the differences in the sedimentary environment of each sand group. According to the thin section identification data of the Xinchang area, the characteristics of the Xujiahe Formation reservoir are analyzed. It is found that the medium and coarse grains are dominant in the middle and upper submembers, accounting for more than 85% of the sandstone in each submember. At the same time, compared with the upper submember, the coarse grain sandstone accounts for more in the middle submember, accounting for about 12% of the sandstone in the middle submember, while the upper submember accounts for only 2.47% of the total amount of sandstone. The lower submember is mainly medium and fine grain sandstone, of which fine grain sandstone accounts for 55%. Within the sand group, the longitudinal ratio of the grain size lithofacies can reflect the vertical sedimentary rhythm characteristics, and the vertical variation of the shale content can be reflected in the three-dimensional model.

[0288] 2. Set the grain size lithofacies thickness distribution. The grain size lithofacies thickness distribution statistics are used to master the thickness range of different lithofacies types, which are used as input parameters in the simulation process. Different grain size lithofacies types have different thickness distributions. For example, in the delta front deposition, the thickness of the medium-coarse sandstone facies is larger than that of the fine sandstone facies, because the fine sandstone facies is mainly deposited in the underwater distributary channel and the river mouth bar, and thus the thickness is larger. The fine-siltstone facies and the mudstone facies are mainly deposited in the underwater natural levee and the front sheet sand, and thus the thickness is smaller.

[0289] 3. Set the second variable of the grain size lithofacies spatial distribution. The constraint data of the grain size lithofacies distribution mainly reflect the distribution trend of each grain size lithofacies between wells. The spatial probability body data of each grain size lithofacies obtained by the neural network clustering of the shale content and the natural gamma ray can be used as a good constraint variable to guide the distribution trend of each grain size lithofacies in the simulation process.

[0290] 4. Set the grain size lithofacies variogram. Each grain size lithofacies still needs a corresponding variogram to control the continuity of the distribution. Based on the discretized grain size lithofacies data at the drilling wells, the variogram fitting of each lithofacies is performed to obtain the parameters of the main and secondary ranges and the vertical range of each lithofacies in each sand group.

[0291] The sequential indicator simulation method is used, and the spatial development probability body of each grain size lithofacies is used as the second variable constraint. The variogram analysis and the vertical lithofacies distribution proportion data are used as the control condition constraint to simulate, and finally the three-dimensional distribution model of the grain size lithofacies is obtained (as shown in FIG. 6). Figure 9

[0292] (9) Analyze the porosity distribution range of each grain size lithofacies (medium-coarse sandstone facies, fine sandstone facies, and fine-siltstone facies) in each sand group, and perform variogram fitting. In the fitting process, the distribution law of the grain size lithofacies should be fully referenced, and the size and direction of the main range, the secondary range, and the vertical range should be selected according to the analysis results. It is found through the analysis of the relationship between the porosity interpreted from the drilling and the rock physical seismic inversion attribute that there is a good negative correlation between the longitudinal wave velocity and the porosity. Therefore, the longitudinal wave velocity inversion body is used as the second variable for spatial constraint in the porosity simulation.

[0293] The well logging porosity curve obtained in step two is discretized into the model grid using the arithmetic mean method, and this is used as the condition data. According to the geostatistical analysis, the porosity parameter variogram analysis results of each grain size lithofacies in each sand group are combined with the probability of the porosity distribution range in each lithofacies, and the sequential Gaussian random simulation algorithm is used to obtain the grid simulation value and establish the three-dimensional porosity model by using the longitudinal wave velocity-porosity trend body model as the second variable constraint.

[0294] ​The correlation between porosity and permeability in each grain size lithofacies is analyzed, and it is found that the correlation exists in coarse sandstone, medium sandstone and fine sandstone. Based on the grain size lithofacies model obtained in step eight, the porosity model is used as a second variable for collaborative constraint, and the sequential Gaussian simulation method is used to simulate each lithofacies in each lithofacies to establish a three-dimensional permeability distribution model (attached Figure 10 ) In the simulation process, the collaborative variable of porosity in the medium sandstone phase can be appropriately increased compared with the fine sandstone to meet the pore-permeability relationship and differences of different lithofacies.

[0295] (10) According to a large amount of thin section identification data and geological understanding of reservoir sedimentation, combined with the previous understanding of the sweet spot of the reservoir in Xinchang area, it is considered that the overall T2 reservoir is dense, and the reservoir quality mainly depends on the matrix porosity and matrix permeability. Under the overall dense background, the relatively high-quality reservoir (sweet spot) mainly develops in the medium sandstone and coarse sandstone, and the porosity is mostly greater than 4% and the permeability is greater than 0.03 mD; the lower limit threshold value of the physical property of the relatively high-quality reservoir in the fine sandstone phase is relatively low, the porosity is greater than 3% and the permeability is greater than 0.03 mD. Therefore, in the T2 sedimentary body, the high-quality reservoir can be further described in the sandstone phase.

[0296] Based on the established porosity model and permeability model, according to the physical property parameter threshold value of the high-quality reservoir, the porosity and permeability of each lithofacies region in the grain size lithofacies model are calculated, and the high-quality sandstone is obtained when the porosity and permeability are in the value range of the high-quality reservoir, and then a high-quality reservoir model is constructed (attached Figure 11 ).

[0297] Compared with the random distribution of each lithofacies in the phase model established by the traditional method, the grain size lithofacies model established by the method of the present invention includes the sedimentary geological thought. Therefore, the distribution of each grain size lithofacies is in the sandstone phase, and the horizontal and vertical distribution of each grain size lithofacies obeys the sedimentary law. For example, in the channel sand body, coarse sandstone is often deposited in the middle and lower parts with strong hydrodynamic force, near the center of the channel on the plane, and the proportion of medium sandstone and fine sandstone gradually increases to the edge of the channel. By extracting the model thickness distribution of each type of lithofacies in each sand group in the model, the sedimentary thickness and range of coarse sandstone, medium sandstone and fine sandstone are analyzed. The distribution of grain size lithofacies in the model conforms to the sedimentary law, that is, in the sandstone phase, it basically presents a continuous distribution of coarse sandstone-medium sandstone-fine sandstone, and the change process of the lithofacies fully reflects the grain size lithology characteristics of each sedimentary facies.

[0298] The effective reservoir thickness and average porosity (sandstone with a porosity greater than 3% is an effective reservoir) and the high-quality reservoir thickness and average porosity (sandstone with a porosity greater than 4% is a high-quality reservoir) of the TX2-2 and TX2-4 sand groups are extracted from the established reservoir model, and are compared with the actual drilling conditions of the newly drilled well A (attachedFigure 12 The effective reservoir thickness of the TX2-2 sand group in the model of well A is 32.00 m, the average porosity is 4.50%, the high-quality reservoir thickness is 20.00 m, and the average porosity is 4.90%. The actual drilling in the TX2-2 sand group reveals that the effective reservoir thickness is 39.69 m, the average porosity is 4.74%, the high-quality reservoir thickness is 35.29 m, and the average porosity is 4.88%. By comparing with the actual geological data revealed by the later drilled well, it is found that the established geological model is in good agreement, the model has high reliability, and the reservoir prediction is good (as shown in Table 3).

[0299] Table 3 Comparison table of new drilled well granularity lithofacies model and actual drilling parameters in the example research area

[0300]

[0301] It is found through comparison that, based on the sand-shale model, the granularity lithofacies is further controlled in multiple levels, the spatial quantitative distribution of the favorable lithofacies is obtained through the neural network identification of the shale content, and then the reservoir property parameter model is established by controlling the lithofacies, so that the quantitative characterization and prediction of the high-quality reservoir and the attribute parameters are realized, and the model guidance can be provided for the later well site deployment and the optimized development scheme.

[0302] The embodiment of the present application also provides a computer device, which comprises a memory, a processor and a computer program.

[0303] The embodiment of the present application also provides a machine readable storage medium, which stores computer program instructions, and the computer program instructions are executed by the processor to realize the high-quality reservoir simulation method of the dense sandstone.

[0304] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can adopt a completely hardware embodiment, a completely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can adopt a computer program product in the form of one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program codes. The solutions in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming languages such as Java and direct script languages such as JavaScript.

[0305] The embodiments of methods, devices (systems), and computer program products of the application can be described in reference to flowchart illustrations and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart illustrations and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products.

[0306] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products.

[0307] These computer program instructions can also be loaded into a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products. Figure 1 one or more functions specified in the flowchart and / or block diagrams of the flowchart and / or block diagrams of the methods, devices (systems), and computer program products.

[0308] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those of skill in the art once they have the benefit of the present disclosure. Therefore, the appended claims are intended to encompass within their scope all such variations and modifications as are within the scope of the application. It should be apparent that a person of skill in the art can make modifications to the application without departing from the scope thereof. Therefore, the following claims should be construed to include within their scope all equivalent variations and modifications to the embodiments disclosed herein.

[0309] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described herein.

Claims

1. A method of modeling a quality reservoir of a tight sand, characterized in that, The application relates to a method for establishing a three-dimensional distribution model of high-quality reservoirs of a target oil and gas reservoir. The method comprises the following steps: establishing a three-dimensional structure and stratum framework model of the target oil and gas reservoir according to three-dimensional seismic data of the target oil and gas reservoir; obtaining parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtaining sand-mudstone discrete data, sedimentary facies discrete data and granularity lithofacies discrete data according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure and stratum framework model; carrying out simulation based on the sand-mudstone discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data to obtain a granularity lithofacies three-dimensional geological model; carrying out simulation according to the parameter interpretation data of the target oil and gas reservoir, the granularity lithofacies discrete data and the granularity lithofacies three-dimensional geological model to obtain a physical parameter three-dimensional geological model of the target oil and gas reservoir; 2. The method of simulating a quality reservoir of a tight sand according to claim 1, characterized in that, determining a high-quality reservoir three-dimensional distribution model of the target oil and gas reservoir according to the physical three-dimensional geological model of the target oil and gas reservoir, and the high-quality reservoir three-dimensional distribution model is used for describing the high-quality reservoir of the target oil and gas reservoir. The method comprises the following steps: carrying out fine structure interpretation on the target oil and gas reservoir by using three-dimensional seismic data to obtain layer and fracture data of the target oil and gas reservoir; 3. The method of modeling quality reservoirs of tight sandstones according to claim 1, characterized in that, establishing a three-dimensional structure and stratum framework model according to the layer and fracture data of the target oil and gas reservoir. The method comprises the following steps: obtaining parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtaining sand-mudstone discrete data, sedimentary facies discrete data and granularity lithofacies discrete data according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structure and stratum framework model.

4. The method of modeling quality reservoirs of tight sandstones according to claim 1, characterized in that, The method comprises the following steps: obtaining parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, wherein the parameter interpretation data comprises physical parameter interpretation data, lithological parameter interpretation data and mineral parameter interpretation data; discretizing the obtained parameter interpretation data into the three-dimensional structure and stratum framework model to obtain sand-mudstone discrete data, sedimentary facies discrete data and granularity lithofacies discrete data. The method comprises the following steps: obtaining a sand-mudstone three-dimensional geological model according to the sand-mudstone discrete data; 5. The method of modeling quality reservoirs of tight sandstones according to claim 4, characterized in that, obtaining a sedimentary facies three-dimensional geological model according to the sedimentary facies discrete data and the sand-mudstone three-dimensional geological model; obtaining a shale content three-dimensional geological model according to the sedimentary facies discrete data, the parameter interpretation data and the sedimentary facies three-dimensional geological model; obtaining a granularity lithofacies three-dimensional geological model according to the granularity lithofacies discrete data, the sedimentary facies three-dimensional geological model and the shale content three-dimensional geological model.

6. The method of simulating a quality reservoir of a tight sand of claim 4, wherein, The method comprises the following steps: carrying out geological statistical analysis on the sand-mudstone discrete data to obtain variogram parameter data of the sand-mudstone discrete data distribution; carrying out sequential indicator simulation on the target oil and gas reservoir according to the variogram parameter data of the sand-mudstone discrete data distribution to obtain a sand-mudstone three-dimensional geological model. The method comprises the following steps: obtaining a sedimentary facies plane distribution two-dimensional graph of the target oil and gas reservoir according to the sedimentary facies discrete data; obtaining a sedimentary facies ratio according to the sand-mudstone three-dimensional geological model; According to the two-dimensional diagram of the sedimentary facies planar distribution and the sedimentary facies ratio, multi-point geostatistical simulation is performed on the target oil and gas reservoir to obtain a three-dimensional geological model of the sedimentary facies.

7. The method of modeling quality reservoirs of tight sandstones according to claim 4, characterized in that, The three-dimensional geological model of the shale content is obtained according to the discrete data of the sedimentary facies, the parameter interpretation data and the three-dimensional geological model of the sedimentary facies, and includes the following steps: According to the discrete data of the sedimentary facies and the shale content interpretation data in the parameter interpretation data, the spatial data distribution of the shale content in different sedimentary facies is obtained. According to the three-dimensional geological model of the sedimentary facies and the spatial data distribution of the shale content in the sedimentary facies, sequential Gaussian random simulation is performed on the target oil and gas reservoir to obtain the three-dimensional geological model of the shale content.

8. The method of modeling quality reservoirs of tight sandstones according to claim 4, characterized in that, The three-dimensional geological model of the grain size lithofacies is obtained according to the discrete data of the grain size lithofacies, the three-dimensional geological model of the sedimentary facies and the three-dimensional geological model of the shale content, and includes the following steps: An optimal grain size lithofacies sensitive well logging curve of the target oil and gas reservoir is obtained, and a three-dimensional inversion body model is obtained according to the optimal grain size lithofacies sensitive well logging curve; According to the three-dimensional geological model of the shale content and the three-dimensional inversion body model, the spatial probability body of each grain size lithofacies is obtained. Geostatistical analysis is performed on the discrete data of the grain size lithofacies to obtain variogram parameter data of the discrete data distribution of the grain size lithofacies. According to the variogram parameter data of the discrete data distribution of the grain size lithofacies, the spatial probability body of each grain size lithofacies and the three-dimensional geological model of the sedimentary facies, sequential indicator random simulation is performed on the target oil and gas reservoir to obtain the three-dimensional geological model of the grain size lithofacies.

9. The method of modeling quality reservoirs of tight sandstones according to claim 8, characterized in that, The three-dimensional inversion body model is obtained according to the optimal grain size lithofacies sensitive well logging curve, and includes the following steps: The grain size lithofacies sensitive well logging curve of the participating well of the target oil and gas reservoir and the shale content interpretation data in the parameter interpretation data are input into a pre-trained grain size lithofacies classification information neural network model; The grain size lithofacies classification information neural network model outputs the classification information of the grain size lithofacies of the participating well; The classification information of the grain size lithofacies of the non-participating well of the target oil and gas reservoir is compared with the classification information of the grain size lithofacies of the participating well to obtain a comparison result; The optimal grain size lithofacies sensitive well logging curve is determined in the grain size lithofacies sensitive well logging curve of the participating well according to the comparison result; The three-dimensional inversion body model is obtained according to the optimal grain size lithofacies sensitive well logging curve.

10. The method of modeling quality reservoirs of tight sandstones according to claim 8, characterized in that, The spatial probability body of each grain size lithofacies is obtained according to the three-dimensional geological model of the shale content and the three-dimensional inversion body model, and includes the following steps: The three-dimensional geological model of the shale content and the three-dimensional inversion body model are input into a pre-trained grain size lithofacies spatial probability body neural network model; The grain size lithofacies spatial probability body neural network model outputs the spatial probability body of each grain size lithofacies.

11. The method of simulating a quality reservoir of a tight sand of claim 1, wherein, The three-dimensional geological model of the physical property parameters of the target oil and gas reservoir is obtained by simulation according to the parameter interpretation data of the target oil and gas reservoir, the discrete data of the grain size lithofacies and the three-dimensional geological model of the grain size lithofacies, and includes the following steps: According to the discrete data of the grain size lithofacies and the physical property parameter interpretation data in the parameter interpretation data, the spatial data distribution of the physical property parameters in different grain size lithofacies is obtained. According to the three-dimensional geological model of the grain size lithofacies and the spatial distribution of the physical property parameters in different grain size lithofacies, sequential Gaussian random simulation is performed on the target oil and gas reservoir to obtain the three-dimensional geological model of the physical property parameters.

12. The method of modeling quality reservoirs of tight sandstones according to claim 1, characterized in that, The method comprises the following steps: obtaining the relationship between the physical parameters in each granularity lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir; determining the threshold value of the physical parameters of the high-quality reservoir for different granularity lithofacies according to the relationship between the physical parameters in each granularity lithofacies and the oil and gas content and the oil and gas production capacity of the target oil and gas reservoir; calculating the grid body that meets all the threshold values of the physical parameters at the same time according to the three-dimensional geological model of the granularity lithofacies and the three-dimensional geological model of the physical parameters; obtaining the three-dimensional distribution model of the high-quality reservoir of the target oil and gas reservoir according to the grid body that meets all the threshold values of the physical parameters at the same time.

13. A quality reservoir simulation apparatus for tight sandstones, characterized in that, The method comprises the following steps: a first model construction module is configured to establish a three-dimensional structural stratigraphic framework model of a target oil and gas reservoir according to three-dimensional seismic data; a data acquisition module is configured to acquire parameter interpretation data of the target oil and gas reservoir according to cable logging curve data of the target oil and gas reservoir, and obtain sand-shale discrete data, sedimentary facies discrete data and granularity lithofacies discrete data according to the parameter interpretation data of the target oil and gas reservoir and the three-dimensional structural stratigraphic framework model; a first simulation module is configured to simulate based on the sand-shale discrete data, the sedimentary facies discrete data and the granularity lithofacies discrete data, and obtain a three-dimensional geological model of the granularity lithofacies; a second simulation module is configured to simulate according to the parameter interpretation data of the target oil and gas reservoir, the granularity lithofacies discrete data and the three-dimensional geological model of the granularity lithofacies, and obtain a three-dimensional geological model of the physical parameters of the target oil and gas reservoir; a second model construction module is configured to determine a three-dimensional distribution model of the high-quality reservoir of the target oil and gas reservoir according to the three-dimensional geological model of the physical parameters of the target oil and gas reservoir, and the three-dimensional distribution model of the high-quality reservoir is used to describe the high-quality reservoir of the target oil and gas reservoir.

14. A computer device, comprising: The method comprises the following steps: a memory; a processor; and a computer program; wherein the computer program is stored in the memory and is configured to be executed by the processor to implement the method for simulating the high-quality reservoir of the tight sandstone according to any one of claims 1 to 12.

15. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method for simulating the high-quality reservoir of the tight sandstone according to any one of claims 1 to 12.

Images