A large-size vertical sand filling model of a heterogeneous reservoir based on real core slice statistics and a construction method thereof

Abstract

The application provides a large-size vertical sand filling model of a heterogeneous reservoir based on real core slice statistics and a construction method thereof, and belongs to the field of oil and gas field development experiment simulation. In order to solve the obvious limitations of the existing sand filling model preparation technology in the aspects of the fineness of geological reduction, the vertical heterogeneity simulation capability and the internal dynamic monitoring level, the application comprises a model main body, a displacement liquid conveying part, an electrode data acquisition part, an output liquid metering part, a plurality of injection ports and a plurality of output ports. Through systematic analysis of the microstructure characteristics of a real core, a quantitative characterization method of the vertical heterogeneity of the reservoir is established, and the quantitative parameters are converted into the construction basis of the physical model, so that the model can highly restore the vertical heterogeneity of the target reservoir. The interlayer difference is controllable, traceable and reproducible, and the flexibility is better. The interlayer heterogeneity can be directly converted into observable layered breakthrough characteristics, and the spatial distribution of oil saturation can be monitored.

Description

Technical Field

[0001] This invention relates to the field of experimental simulation technology for oil and gas field development, and more specifically, to a large-scale vertical sand-filling model of heterogeneous reservoirs based on statistical analysis of real core thin sections and its construction method. Background Technology

[0002] In the field of oil and gas field development, especially for tapping the remaining oil potential in high water-cut oilfields, physical simulation experiments are a key technical means to study the fluid seepage law within the reservoir and evaluate the oil displacement effect. As a commonly used physical simulation device, the core objective of sand-filled models is to reproduce the physical properties of actual reservoirs as realistically as possible, especially their heterogeneity, thereby ensuring the scientific validity of the experimental results and their guiding value for field practice.

[0003] However, existing sand-filled model preparation techniques have the following shortcomings, which limit their application effectiveness:

[0004] 1. The model parameters lack geological basis and are out of touch with actual reservoir characteristics:

[0005] Existing technologies for designing sand-filled models typically rely on macroscopic, averaged geological parameters (such as average permeability). When simulating heterogeneous reservoirs, the quartz sand particle size distribution used for different layers is often based on experience or simplified theoretical models, lacking direct support from quantitative characterization of the microstructure of real reservoir cores. This results in models that cannot accurately reproduce the vertical rhythms and heterogeneous structures (e.g., positive rhythms, vertical variations in median grain size) controlled by sedimentation in actual reservoirs, making it impossible for models to realistically simulate key seepage behaviors such as interlayer disturbances and dominant flow.

[0006] 2. The model structure is simple, but its ability to simulate vertical heterogeneity and monitor stratified conditions is insufficient.

[0007] Conventional sand-filled models are mostly homogeneous or simply layered models, with injection and production ends typically designed as single channels. This structure cannot achieve independent injection and production and real-time monitoring for different permeability layers, making it difficult to accurately identify the dominant seepage channels under multi-layer combined production conditions, and also unable to effectively study the flow patterns of fluids between different layers, which are precisely the key factors affecting the oil displacement efficiency of heterogeneous reservoirs.

[0008] 3. Lack of internal state monitoring methods in the model, resulting in low visualization and quantification of the experimental process:

[0009] Traditional physical simulation experiments mainly rely on macroscopic data such as pressure, flow rate, and recovery rate at the model ends (inlet and outlet) to infer internal processes, which is a "black box" or "grey box" operation. The lack of technical means to monitor the spatial distribution of fluid saturation in real time and in situ within the model limits researchers' intuitive and quantitative understanding of the formation and distribution of residual oil during displacement, making it difficult to reveal the dynamic evolution law of microscopic residual oil.

[0010] In summary, existing sand-filled model preparation techniques have significant limitations in terms of the precision of geological reconstruction, the ability to simulate vertical heterogeneity, and the level of dynamic monitoring within the model. This invention aims to overcome these shortcomings by providing a novel sand-filled model preparation method based on quantitative statistics of real core microscopic characteristics, capable of highly reproducing reservoir vertical heterogeneity, and integrating internal monitoring functions. Summary of the Invention

[0011] The technical problem to be solved by this invention is:

[0012] To address the significant limitations of existing sand-filled model preparation technologies in terms of the precision of geological reconstruction, the ability to simulate vertical heterogeneity, and the level of dynamic monitoring within the model.

[0013] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows:

[0014] This invention provides a large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections. The model includes a main body, a displacement fluid delivery component, an electrode data acquisition component, a produced fluid metering component, multiple injection ports, and multiple production outlets.

[0015] The main body of the model is filled with multiple layers of sand.

[0016] Electrode data acquisition components are embedded in the multi-layer sand filling, including multiple uniformly arranged electrodes to form an electrode array. The electrodes in each layer are arranged at equal intervals in the plane, and at equal intervals in the vertical direction between layers.

[0017] The model body has multiple injection ports at its inlet end, each of which is connected to a displacement fluid delivery component. The displacement fluid delivery component includes a horizontal flow pump, a pressure sensor, a displacement fluid storage device, and an injection pipe. The displacement fluid storage device is connected to multiple injection pipes, and each injection pipe is connected to a corresponding injection port. The displacement fluid storage device pumps the stored displacement fluid into the model through the horizontal flow pump. The pressure sensor is installed on the injection pipe.

[0018] The main body of the model has multiple production outlets at its outlet end. Each production outlet is connected to a production liquid metering component. The production liquid metering component includes a production pipe, a flow sensor, a shut-off valve or a three-way sampling port, and a collection container or an oil-water separation unit. The collection container or oil-water separation unit is connected to multiple production pipes, and each production pipe is connected to a corresponding production outlet. The production pipe is equipped with a flow sensor and a shut-off valve or a three-way sampling port.

[0019] Furthermore, each layer of sand filling corresponds one-to-one with the sand filling parameters obtained from the parameter extraction of the full-diameter large-size thin section. That is, the thickness, sand formula and compaction process of each layer of sand filling are obtained by inversion based on the real full-diameter large-size thin section, which is used to simulate the heterogeneous vertical environment of real oil and gas fields.

[0020] Furthermore, the layers of sand are solidified by epoxy resin pouring.

[0021] Furthermore, the number of electrodes is 216, and they are evenly distributed in multiple layers of sand in a 6-row, 6-column, 6-layer manner.

[0022] Furthermore, there are 36 injection ports and 36 output ports. The injection pipes are all located on one side of the model body and are evenly arranged in a 6×6 pattern. The injection pipes of each layer are located above the sand filling of the corresponding layer. The output pipes are all located on the side of the model body opposite to the injection pipes and are evenly arranged in a 6×6 pattern. The output pipes of each layer are located above the sand filling of the corresponding layer.

[0023] A method for constructing a large-scale vertical sand-filling model of heterogeneous reservoirs based on statistical analysis of real core thin sections includes the following steps:

[0024] S100, Acquisition and Layer Division of Large-Size Full-Diameter Sections, including,

[0025] S110. Select full-diameter cores from different stratigraphic positions of the target heterogeneous reservoir and segment them according to the vertical heterogeneity characteristics.

[0026] S120. Prepare full-diameter large-size thin sections for each layer and retain the correspondence between the thin sections and the core depth;

[0027] S130. Perform full-frame scanning and multi-field acquisition on each full-diameter large-size thin slice to establish an image library;

[0028] S200, particle size / diameter identification and thin-section statistical parameter extraction, including,

[0029] S210. Preprocess the thin-slice image obtained in step S130, including scale calibration, illumination correction, noise suppression and contrast normalization.

[0030] S220. Perform particle and pore segmentation, particle size identification and statistics to obtain particle size structure information;

[0031] S230. Perform pore and throat structure statistics and output microstructure parameters;

[0032] S240. Statistically analyze the average porosity and dispersion of samples within the same layer, and calculate the porosity difference between layers to define the interlayer heterogeneity intensity.

[0033] S300, inversion and selection of stratigraphic parameters into filler particle size and formulation, including,

[0034] S310. Based on the particle size distribution, sorting coefficient, pore throat ratio and throat width distribution of each layer, determine the quartz sand particle size distribution corresponding to that layer.

[0035] S320. Establish a construction list with a three-part table of layer-formula-process for each layer;

[0036] S400, large-scale vertical heterogeneous sand-filled model construction, including

[0037] S410. Scale down the model into layers according to the vertical thickness ratio of the target reservoir, and determine the layer thickness, sand formula and compaction process of each layer.

[0038] S420. Layered sand filling process: According to the matched layer-formulation-process triple table, layer by layer, material is added in stages, and uniform vibration / compaction control is maintained to avoid layer mixing.

[0039] S430. After completing the overall filling, perform basic physical property verification on the model to confirm that the interlayer contrast and hierarchical structure meet the design goals.

[0040] S500, multi-port injection / multi-port output structure installation and dominant layer identification, including,

[0041] S510. Set 36 injection ports vertically at the model inlet end and 36 production outlets vertically at the outlet end.

[0042] S520. When conducting displacement tests, select: overall injection, stratified injection, or combined injection; identify the dominant seepage zones and main channels by monitoring the order of water emergence and production characteristics of each production outlet.

[0043] S600, epoxy resin casting and grid electrode array saturation monitoring, specifically including...

[0044] S610. The model is solidified by epoxy resin casting to form a stable structure and reduce structural disturbances during the flow process.

[0045] S620: An electrode array is embedded inside the model, and the electrode leads are led out to the data acquisition system; the changes in oil saturation in different regions are monitored through electrical response to achieve zonal tracking of the saturation field and identification of residual oil.

[0046] S630, combining multi-port production and electrode monitoring results, forms a comprehensive identification of advantageous seepage location, saturation evolution, and remaining oil enrichment zone, providing a basis for profile control, water shut-off, or chemical flooding schemes.

[0047] Further, in step S110, the heterogeneous features include high-permeability strips, low-permeability interlayers, argillaceous strips, and cemented strips.

[0048] Furthermore, in step S220, the particle size structure information includes rock particle size and content, sorting coefficient, and dense clustering intervals of particle long axis.

[0049] Further, in step S230, the microstructure parameters include porosity, porosity, average pore size, throat width distribution, average pore-throat ratio, average coordination number, and homogeneity coefficient.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] (1) The comprehensive reservoir data restores the reservoir heterogeneity, and the reducibility is significantly improved.

[0052] This invention uses large-diameter thin sections from different strata as statistical objects to extract and output: porosity, porosity, average pore size, throat width distribution, average pore-throat ratio, sorting coefficient, homogeneity coefficient, average coordination number, rock grain size and content, dense clustering intervals of grain long axes, and the proportion of grains of different lithologies. A mapping of "stratum-structure-grain size" is established from these parameters, so that the grain size selection of each layer in the sand-filled model no longer relies on subjective experience, but is determined by inversion from the statistical results of the thin sections, thus making the vertical heterogeneity of the model closer to that of real reservoirs.

[0053] (2) Interlayer differences are controllable, traceable, and reproducible

[0054] This invention quantifies the porosity differences within the same layer and between layers, clarifying "how much difference should there be in each layer and what parameters cause it." The "difference values" are then converted into specific construction control quantities (sand particle size distribution, fine particle content, compaction process parameters, layer thickness ratio), enabling model manufacturing to have a traceable chain of process parameters.

[0055] (3) Identifiable dominant seepage zones: multi-port injection / multi-port output structure

[0056] This invention arranges multiple injection ports and production ports vertically at the inlet and outlet, enabling: switching between single-layer and multi-layer injection; monitoring of multiple production ports to determine "which layer sees water first / breaks through first"; identification of dominant seepage zones; and analysis of interlayer mobility differences. Compared to a single-port model, this invention can directly transform "interlayer heterogeneity" into observable stratified breakthrough characteristics.

[0057] (4) Spatial distribution of oil saturation can be monitored: resin casting + grid electrode array

[0058] A stable structure is formed by epoxy resin casting, and an electrode array is embedded in the model in a high-density grid pattern. The oil saturation changes in different regions are inverted using resistivity / conductivity response, enabling visualized and quantitative tracking of "seepage front advancement, bypass, and residual oil enrichment zones." This overcomes the limitation of traditional sand-filled models that "can only show the total production and not the internal saturation field."

[0059] (5) It has strong applicability and can be extended to various studies on improving oil recovery.

[0060] In addition to basic oil-water two-phase flow, this invention can also be used for research on enhanced oil recovery, such as stratified injection and production optimization, water shut-off and profile control, polymer / surfactant / composite flooding, multi-media interaction and dominant channel management, and has good method scalability.

[0061] This invention effectively solves the above-mentioned defects through a closed-loop route of "statistical analysis of large-size thin sections of real full-diameter cores - stratigraphic parameterization - grain size inversion - layered sand filling - multi-port injection and production - grid electrode saturation monitoring", and realizes the construction of a high-fidelity experimental model and identification of dominant seepage channels for typical vertically heterogeneous reservoirs. Attached Figure Description

[0062] Figure 1 This is a flowchart illustrating a method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, as described in an embodiment of the present invention.

[0063] Figure 2 This is a schematic diagram of the structure of a large-size vertical sand-filling model of a heterogeneous reservoir based on statistics of real core thin sections in an embodiment of the present invention;

[0064] Figure 3 This is an image analysis diagram of the cast thin sheet in an embodiment of the present invention;

[0065] Figure 4 This is a summary diagram of particle size parameters in the embodiments of the present invention; Figure 5 This is a particle size analysis diagram from an embodiment of the present invention. Detailed Implementation

[0066] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0067] Specific Implementation Plan 1: Combining Figure 2 As shown, this invention provides a large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections. The model includes a main body, a displacement fluid delivery component, an electrode data acquisition component, a produced fluid metering component, multiple injection ports, and multiple production outlets.

[0068] The main body of the model is a box structure with a cover plate on top and multiple sand-filling windows on the cover plate, and a bottom plate at the bottom. The internal cavity contains multiple layers of sand. Each layer of sand corresponds one-to-one with the sand-filling parameters obtained by extracting parameters from the full-diameter large-size thin section. That is, the thickness, sand formula and compaction process of each layer of sand are obtained by inversion based on the real full-diameter large-size thin section, which is used to simulate the heterogeneous vertical environment of the real oil and gas field.

[0069] The multi-layered sand filling is solidified by epoxy resin pouring to form a stable structure and reduce structural disturbance during the flow process. It should be noted that the epoxy resin pouring and solidification only involves adding a small amount of epoxy resin for bonding purposes and will not affect subsequent experiments.

[0070] Electrode data acquisition components are embedded in the multi-layer sand filling. The components include multiple uniformly arranged electrodes to form an electrode array. Each layer of electrodes is arranged at equal intervals in the plane and at equal intervals along the vertical direction between layers. The electrode coordinates are established in correspondence with the layer positions of the model to invert the changes in oil saturation or conductivity at different depths / regions. This invention uses 216 electrodes, which are evenly distributed in 6 horizontal rows, 6 columns, and 6 layers in the multi-layer sand filling of the cavity.

[0071] The model body has 36 injection ports at its inlet, each connected to a displacement fluid delivery component. This component includes a horizontal flow pump, a pressure sensor, a displacement fluid storage tank, and injection pipes. The displacement fluid storage tank is connected to each of the 36 injection pipes, and each injection pipe is connected to a corresponding injection port. The displacement fluid storage tank is pumped into the model through the injection pipes by the horizontal flow pump. The pressure sensor is installed on the injection pipes to monitor pressure differentials and for safety protection. Each injection pipe is equipped with a valve. Figure 2 As shown, the injection pipes are all located on one side of the main body of the model and are evenly arranged in a 6×6 pattern. The injection pipes of each layer are set above the sand filling of the corresponding layer.

[0072] The main body of the model has 36 production outlets at its outlet end, each connected to a production liquid metering component. The production liquid metering component includes a production pipe, a flow sensor, a shut-off valve or tee sampling port, and a collection container or oil-water separation unit. The collection container or oil-water separation unit is connected to each of the 36 production pipes, and each production pipe is connected to its corresponding production outlet. Each production pipe is equipped with a flow sensor and a shut-off valve or tee sampling port, used to monitor the production volume of each outlet in real time through the 36 flow sensors, enabling comparison of inter-layer (or zone) production contributions; combined with... Figure 2 As shown, the output pipes are all located on the opposite side of the injection pipe of the model body and are evenly arranged in a 6×6 pattern. The output pipes of each layer are set below the sand filling of the corresponding layer.

[0073] Specific Implementation Plan Two: Combining Figures 1 to 4 As shown, this invention provides a method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, comprising the following steps:

[0074] S100, Acquisition and Layer Division of Large-Size Full-Diameter Sections, including,

[0075] S110. Select full-diameter cores from different strata of the target heterogeneous reservoir and segment them according to the vertical heterogeneous characteristics (high-permeability bands, low-permeability interlayers, argillaceous bands, cemented bands).

[0076] S120. Prepare full-diameter large-size thin sections for each layer (using vacuum resin impregnation and indigo / fluorescence enhancement for pore identification), and retain the correspondence between the thin sections and the core depth;

[0077] S130. Perform full-frame scanning and multi-field acquisition on each full-diameter large-size thin section to establish an image library (containing metadata of depth, layer, field coordinates, magnification, and resolution).

[0078] S200, particle size / diameter identification and thin-section statistical parameter extraction, including,

[0079] S210. Preprocess the thin-slice image obtained in step S130, including scale calibration, illumination correction, noise suppression and contrast normalization.

[0080] S220. Perform particle and pore segmentation, particle size identification and statistics to obtain particle size structure information such as "rock particle size and content", "sorting coefficient" and "dense clustering interval of particle long axis".

[0081] S230. Perform pore and throat structure statistics and output the microstructure parameters of "porosity, porosity, average pore diameter, throat width distribution, average pore-throat ratio, average coordination number, and homogeneity coefficient".

[0082] S240. Statistically analyze the average porosity and dispersion of samples within the same layer, and calculate the porosity difference between layers to define the interlayer heterogeneity intensity.

[0083] S300, inversion and selection of stratigraphic parameters into filler particle size and formulation, including,

[0084] S310. Based on the particle size distribution, sorting coefficient, pore throat ratio and throat width distribution of each layer, determine the quartz sand particle size distribution corresponding to that layer.

[0085] S320. Establish a construction list of "layer-formula-process triple table" for each layer; if there are differences in the proportion of various lithological particles, the layer-by-layer micro-layer stacking method can be adopted (different sand formulas represent different proportions of lithological components); the triple table is based on the rock physical property information of each layer, and the same sand filling material is configured accordingly. The filling material is then filled into the model using an appropriate sand filling process.

[0086] S400, large-scale vertical heterogeneous sand-filled model construction, including

[0087] S410. Scale down the model into layers according to the vertical thickness ratio of the target reservoir, and determine the layer thickness, sand formula and compaction process of each layer.

[0088] S420. Layered sand filling process: According to the matched layer-formulation-process triple table (as shown in Table 1 below), layer by layer, material is added in stages, and uniform vibration / compaction control is maintained to avoid layer mixing; for muddy interlayers or low-permeability interlayers, high fine-particle formulation or independent interlayer material is used to form a continuous barrier.

[0089]

[0090]

[0091] S430. After the overall filling is completed, the basic physical properties of the model are verified (overall and segmental pressure difference / permeability response) to confirm that the interlayer contrast and sequence structure meet the design objectives.

[0092] S500, multi-port injection / multi-port output structure installation and dominant layer identification, including,

[0093] S510. Set 36 injection ports vertically at the model inlet end and 36 production outlets vertically at the outlet end.

[0094] S520. When conducting displacement tests, the following can be selected: overall injection, stratified injection or combined injection; by monitoring the order of water emergence and production characteristics of each production outlet, the dominant seepage zone and the main channel can be identified.

[0095] S600, epoxy resin casting and grid electrode array saturation monitoring, specifically including...

[0096] S610. The model is solidified by epoxy resin casting to form a stable structure and reduce structural disturbances during the flow process.

[0097] S620: An electrode array is embedded inside the model in a high-density grid (i.e., 216 electrodes are evenly distributed in the cavity in a 6-row, 6-column, 6-layer manner), and the electrode leads are led out to the data acquisition system; the changes in oil saturation in different regions are monitored by electrical response to achieve zonal tracking of the saturation field and identification of residual oil.

[0098] S630, combined with multi-port production and electrode monitoring results, forms a comprehensive identification of "dominant seepage location - saturation evolution - remaining oil enrichment zone", providing a basis for profile control, water shut-off or chemical flooding schemes.

[0099] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.

Claims

1. A large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections, characterized in that: It includes the model body, displacement fluid delivery components, electrode data acquisition components, product fluid metering components, multiple injection ports, and multiple product outlets. The model body contains multiple layers of sand, and its preparation method includes: S100, Acquisition and Layer Division of Large-Size Full-Diameter Sections, including, S110. Select full-diameter cores from different stratigraphic positions of the target heterogeneous reservoir and segment them according to the vertical heterogeneity characteristics. S120. Prepare full-diameter large-size thin sections for each layer and retain the correspondence between the thin sections and the core depth; S130. Perform full-frame scanning and multi-field acquisition on each full-diameter large-size thin slice to establish an image library; S200, particle size / diameter identification and thin-section statistical parameter extraction, including, S210. Preprocess the thin-slice image obtained in step S130, including scale calibration, illumination correction, noise suppression and contrast normalization. S220. Perform particle and pore segmentation, particle size identification and statistics to obtain particle size structure information; S230. Perform pore and throat structure statistics and output microstructure parameters; S240. Statistically analyze the average porosity and dispersion of samples within the same layer, and calculate the porosity difference between layers to define the interlayer heterogeneity intensity. S300, inversion and selection of stratigraphic parameters into filler particle size and formulation, including, S310. Based on the particle size distribution, sorting coefficient, pore throat ratio and throat width distribution of each layer, determine the quartz sand particle size distribution corresponding to that layer. S320. Establish a construction list with a three-part table of layer-formula-process for each layer; S400, large-scale vertical heterogeneous sand-filled model construction, including S410. Scale down the model into layers according to the vertical thickness ratio of the target reservoir, and determine the layer thickness, sand formula and compaction process of each layer. S420. Layered sand filling process: According to the matched layer-formulation-process triple table, layer by layer, material is added in stages, and uniform vibration / compaction control is maintained to avoid layer mixing. S430. After completing the overall filling, perform basic physical property verification on the model to confirm that the interlayer contrast and hierarchical structure meet the design goals. Electrode data acquisition components are embedded in the multi-layer sand filling, including multiple uniformly arranged electrodes to form an electrode array. The electrodes in each layer are arranged at equal intervals in the plane, and at equal intervals in the vertical direction between layers. The model body has multiple injection ports at its inlet end, each of which is connected to a displacement fluid delivery component. The displacement fluid delivery component includes a horizontal flow pump, a pressure sensor, a displacement fluid storage device, and an injection pipe. The displacement fluid storage device is connected to multiple injection pipes, and each injection pipe is connected to a corresponding injection port. The displacement fluid storage device pumps the stored displacement fluid into the model through the horizontal flow pump. The pressure sensor is installed on the injection pipe. The main body of the model has multiple production outlets at its outlet end. Each production outlet is connected to a production liquid metering component. The production liquid metering component includes a production pipe, a flow sensor, a shut-off valve or a three-way sampling port, and a collection container or an oil-water separation unit. The collection container or oil-water separation unit is connected to multiple production pipes, and each production pipe is connected to a corresponding production outlet. The production pipe is equipped with a flow sensor and a shut-off valve or a three-way sampling port.

2. The large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections according to claim 1, characterized in that: Each layer of sand filling corresponds one-to-one with the sand filling parameters obtained from the parameter extraction of the full-diameter large-size thin section. That is, the thickness, sand formula and compaction process of each layer of sand filling are obtained by inversion based on the real full-diameter large-size thin section, which is used to simulate the heterogeneous vertical environment of real oil and gas fields.

3. The large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections according to claim 2, characterized in that: The layers of sand are solidified by epoxy resin pouring.

4. The large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections according to claim 3, characterized in that: The number of electrodes is 216, and they are evenly distributed in multiple layers of sand in a 6-row, 6-column, 6-layer manner.

5. A large-scale vertical sand-filling model for heterogeneous reservoirs based on statistical analysis of real core thin sections, as described in claim 4, is characterized in that: The number of injection ports and output ports is 36 each. The injection pipes are all located on one side of the model body and are evenly arranged in a 6×6 pattern. The injection pipes of each layer are located above the sand filling of the corresponding layer. The output pipes are all located on the side of the model body opposite to the injection pipes and are evenly arranged in a 6×6 pattern. The output pipes of each layer are located above the sand filling of the corresponding layer.

6. A method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, as described in any one of claims 1-5, characterized in that, Includes the following steps: S100, Acquisition and Layer Division of Large-Size Full-Diameter Sections, including, S110. Select full-diameter cores from different stratigraphic positions of the target heterogeneous reservoir and segment them according to the vertical heterogeneity characteristics. S120. Prepare full-diameter large-size thin sections for each layer and retain the correspondence between the thin sections and the core depth; S130. Perform full-frame scanning and multi-field acquisition on each full-diameter large-size thin slice to establish an image library; S200, particle size / diameter identification and thin-section statistical parameter extraction, including, S210. Preprocess the thin-slice image obtained in step S130, including scale calibration, illumination correction, noise suppression and contrast normalization. S220. Perform particle and pore segmentation, particle size identification and statistics to obtain particle size structure information; S230. Perform pore and throat structure statistics and output microstructure parameters; S240. Statistically analyze the average porosity and dispersion of samples within the same layer, and calculate the porosity difference between layers to define the interlayer heterogeneity intensity. S300, inversion and selection of stratigraphic parameters into filler particle size and formulation, including, S310. Based on the particle size distribution, sorting coefficient, pore throat ratio and throat width distribution of each layer, determine the quartz sand particle size distribution corresponding to that layer. S320. Establish a construction list with a three-part table of layer-formula-process for each layer; S400, large-scale vertical heterogeneous sand-filled model construction, including S410. Scale down the model into layers according to the vertical thickness ratio of the target reservoir, and determine the layer thickness, sand formula and compaction process of each layer. S420. Layered sand filling process: According to the matched layer-formulation-process triple table, layer by layer, material is added in stages, and uniform vibration / compaction control is maintained to avoid layer mixing. S430. After completing the overall filling, perform basic physical property verification on the model to confirm that the interlayer contrast and hierarchical structure meet the design goals. S500, multi-port injection / multi-port output structure installation and dominant layer identification include, S510. Set 36 injection ports vertically at the model inlet end and 36 production outlets vertically at the outlet end. S520. When conducting displacement tests, select: overall injection, stratified injection, or combined injection; identify the dominant seepage zones and main channels by monitoring the order of water emergence and production characteristics of each production outlet. S600, epoxy resin casting and grid electrode array saturation monitoring, specifically including... S610. The model is solidified by epoxy resin casting to form a stable structure and reduce structural disturbances during the flow process. S620: An electrode array is embedded inside the model, and the electrode leads are led out to the data acquisition system; the changes in oil saturation in different regions are monitored through electrical response to achieve zonal tracking of the saturation field and identification of residual oil. S630, combining multi-port production and electrode monitoring results, forms a comprehensive identification of advantageous seepage location, saturation evolution, and remaining oil enrichment zone, providing a basis for profile control, water shut-off, or chemical flooding schemes.

7. The method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, as described in claim 6, is characterized in that: In step S110, the heterogeneous features include high-permeability strips, low-permeability interlayers, argillaceous strips, and cemented strips.

8. The method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, as described in claim 7, is characterized in that: In step S220, the particle size structure information includes rock particle size and content, sorting coefficient, and densely clustered regions of particle long axis.

9. The method for constructing a large-scale vertical sand-filling model of a heterogeneous reservoir based on statistical analysis of real core thin sections, as described in claim 8, is characterized in that: In step S230, the microstructure parameters include porosity, porosity, average pore size, throat width distribution, average pore-throat ratio, average coordination number, and homogeneity coefficient.

Images