A method for simulating carbon dioxide flooding based on digital core model

Abstract

The application discloses a kind of carbon dioxide oil displacement simulation methods based on digital core model, it is related to oil and gas exploration and development field, solve the problem that existing technology adopts real core model has long test cycle, observation measurement parameter exists deviation insufficient etc..The application creates digital core model using real core pore structure characteristics, combines machine learning and artificial intelligence algorithm, fully couples different physical processes, according to different crude oil properties (density, dynamic viscosity), different CO2 injection rate and other parameters, to simulate the dynamic distribution characteristics of remaining oil in the process of CO2 displacement in underground reservoir with single variable.The application method is suitable for low-permeability-low-permeability reservoirs, medium-high permeability sandstone heavy oil reservoirs (CO2 assisted huff and puff) and fractured oil and gas reservoirs for CO2 displacement, and has important significance for evaluating the development effect of CO2 injection in oilfield.The application method avoids the tedious process of repeated oil washing of real core, and improves work efficiency.

Description

Technical Field

[0001] This invention relates to the field of oil and gas exploration and development, and specifically to a carbon dioxide flooding simulation method based on a digital core model. Background Technology

[0002] In recent years, low-permeability oil reservoirs have been widely distributed in my country and are a major target for oilfield development. Low-permeability reservoirs suffer from drawbacks such as high water injection start-up pressure, low water absorption capacity, rapid rise in water injection pressure, rapid decline in water injection volume, slow pressure recovery in production wells, rapid production decline, and very low water injection volume, oil production, extraction rate, and recovery rate. Typically, the average recovery rate of primary oil recovery (depletion-type oil recovery) and secondary oil recovery (water injection oil recovery) is only 10%-15%. Relying solely on water injection to enhance oil recovery in low-permeability reservoirs is becoming increasingly uneconomical and unsustainable. Therefore, finding alternative technologies for water injection development has become imperative. CO2 gas is generally considered to have advantages such as strong injection capacity, large expansion coefficient, and good miscibility with crude oil. It can improve oil recovery rate (EOR) by approximately 5-10 percentage points based on water injection development in oilfields, demonstrating significant potential. Production practice has proven that CO2-driven enhanced oil recovery technology is an excellent alternative to water-driven enhanced oil recovery technology.

[0003] The mechanisms of CO2-driven enhanced oil recovery mainly include reducing interfacial tension and decreasing residual oil in pore throats; expanding the volume of crude oil after dissolution and increasing the driving energy; reducing crude oil viscosity and improving the mobility ratio; loosening the surface oil film on the rock skeleton, allowing it to disperse and drive migration; and CO2 entering the blind end along the central axis, squeezing out or displacing the crude oil and reducing residual oil in the blind end.

[0004] Long core physical simulation is an important tool for studying oil displacement efficiency. Typically, long core assemblies are 1m in length, with core diameters of 25, 50, and 75mm. Long core gas injection physical simulation experiments can study both huff-and-puff tests and the oil displacement efficiency of different types of gases (water, foam, hydrocarbons, non-hydrocarbons, air, slug gas drive, etc.). This experiment can also investigate the effects of different slug sizes, gas-water ratios, and injection pressures on oil recovery in alternating water-gas injection, as well as the timing of gas injection and the gas injection efficiency in reservoirs with different dip angles.

[0005] Digital cores are used to image and digitize the pore space and mineral matrix (solid skeleton) of natural rocks, providing a detailed description of the microscopic pore structure. Various physicochemical processes within this digital object are then numerically simulated, such as the movement of fluids within the pore space, to obtain macroscopic rock properties like porosity and permeability. Digital cores offer several advantages: compared to conventional methods like mercury intrusion porosimetry, sonic logging, and triaxial experiments, they effectively preserve and reuse cores; they form the basis for numerical simulations of digital core physics, enabling the simulation of physical and chemical processes such as heat, fluidity, force, and diagenesis; and with advancements in high-resolution 3D imaging capabilities and software and hardware computing power, directly calculating rock physical properties based on digital cores is becoming increasingly easier. As experimental equipment and computer hardware and software performance improve, there are two main methods for constructing digital cores: one is based on physical measurements of rock samples (e.g., CT scans), and the other is based on mathematical statistics or computer simulation.

[0006] Indoor tests and production practice have proven that compared with using a real core model, which has disadvantages such as long test cycles and deviations in observation and measurement parameters, this invention proposes a carbon dioxide flooding simulation method based on a digital core model with real core pore structure, which can quickly solve these drawbacks. Summary of the Invention

[0007] To address the problems of long experimental cycles and insufficient deviations in observation and measurement parameters when using real core models in existing technologies, this invention provides a carbon dioxide flooding simulation method based on digital core models.

[0008] A simulation method for carbon dioxide flooding based on a digital core model, comprising the following steps:

[0009] Step 1: Construct a simulated digital core model based on the pore structure characteristics of the actual core.

[0010] Step 2: Using the digital core model described in Step 1, combined with geomechanics, single-well fracture analysis and attribute model, output pore structure parameters such as porosity, permeability, tortuosity and specific surface area through machine deep learning and artificial intelligence algorithms;

[0011] Step 3: Use multi-parameter single-variable simulation to simulate the CO2 displacement process in underground oil reservoirs and obtain the oil displacement efficiency, flow characteristics and residual oil distribution characteristics in the pore structure space.

[0012] The beneficial effects of this invention are:

[0013] This invention utilizes the pore structure characteristics of real core samples to create a digital core model. Combining machine learning and artificial intelligence algorithms, it comprehensively couples different physical processes and simulates the dynamic distribution characteristics of residual oil in underground reservoirs during CO2 displacement, based on parameters such as different crude oil properties (density, dynamic viscosity) and different CO2 injection rates, using a single variable. This method is applicable to low-permeability to ultra-low-permeability reservoirs, medium-to-high permeability sandstone heavy oil reservoirs (with CO2-assisted huff and puff), and fractured oil and gas reservoirs using CO2 displacement, and is of great significance for evaluating the effectiveness of CO2 injection development in oil fields.

[0014] In the method described in this invention, the digital core technology has a high degree of fit with the pore structure of real cores and can repeatedly simulate displacement tests, avoiding tedious processes such as repeated oil washing of real cores and improving work efficiency. Attached Figure Description

[0015] Figure 1 This is a flowchart of a carbon dioxide flooding simulation method based on a digital core model, as described in this invention.

[0016] Figure 2 A rendering of the mesh design for a 3D digital core model (2×2×6cm);

[0017] Figure 3 A rendering of a 3D digital core model (2×2×6cm);

[0018] Figure 4 This is a graph showing the concentration distribution at T=2s;

[0019] Figure 5 This is a diagram showing the velocity distribution at T=2s;

[0020] Figure 6 The pressure distribution diagram is shown at T = 0.69s.

[0021] Figure 7 This is a schematic diagram illustrating the change in displacement rate over time.

[0022] Figure 8 This is a diagram showing the volume fraction distribution of the fluid at T = 0.5 s. Detailed Implementation

[0023] Specific Implementation Method 1: Combination Figures 1 to 8 This embodiment describes a carbon dioxide flooding simulation method based on a digital core model, which is implemented through the following steps:

[0024] Step 1. Construct a simulated digital core model based on CT images of the actual core pore structure; the specific process is as follows:

[0025] Step 1-1: Perform high-resolution CT digital imaging on small rock fragments to obtain important process and variable information such as the three-dimensional spatial structure of the opaque porous medium and the internal fluid flow and transport.

[0026] Steps 1-2: Process the original image (grayscale or intensity image) using methods such as noise reduction, smoothing, and segmentation to separate the pores from the rock mineral matrix components and obtain a segmented image;

[0027] Steps 1-3: Based on the images, information about the pore structure of the porous medium (particle size distribution, pore radius distribution, microcrack distribution, etc.) can be obtained.

[0028] Step 2. Based on geomechanics, single-well fracture analysis and property models, combined with machine deep learning (GraphNeural Networks) and artificial intelligence algorithms (Support Vector Machines), comprehensively couple different physical processes.

[0029] The attribute model includes rock type, rock physical properties (porosity, permeability, etc.), and sandstone-mudstone superposition relationship, and its relationship with the digital core is to assign attribute model information to the former.

[0030] In this embodiment, digital cores are the basis for numerical studies of fluid multiphase flow processes in pore space at the pore scale. Under the conditions of miscible, near-miscible, and immiscible displacement of CO2 in digital cores, many factors need to be considered in the flow simulation.

[0031] The application of digital cores in multiphase flow simulation can be divided into indirect and direct methods. Indirect methods obtain pore structure parameters such as porosity, tortuosity, and specific surface area by calculating and processing digital cores, and then use the Kozeny-Carman formula or pore network model to calculate the seepage properties of porous media.

[0032] Direct methods extract pore structures that characterize the pore shape and wall wettability of porous media. These include direct numerical simulation methods such as the finite difference method, finite element method, finite volume method, smoothed particle hydrodynamics, and lattice Boltzmann method, providing a foundation for multiphase flow calculations such as the fluid volume method, level set method, and phase field method. For CT image data, MATLAB software can be used to indirectly calculate pore structure parameters such as porosity, permeability, tortuosity, and specific surface area.

[0033] Step 3. Multi-parameter single-variable simulation of CO2 displacement in underground oil reservoirs;

[0034] In this embodiment, the complex displacement process of CO2 on crude oil at the micropore scale is simulated, including fluid flow, diffusion mass transfer, and phase change. These processes are closely coupled with heterogeneous pore structures, and the variations in the distribution of CO2 and crude oil between large and small pores greatly determine the upscaling process. Pore scale studies can significantly enhance our understanding of the influence of these coupled processes and micropore structure on the flow behavior of CO2 on crude oil and macroscopic scavenging efficiency.

[0035] Step 3-1: Using a three-dimensional digital core model, the number of vertices, triangles, edge units, vertex units, etc. of the model are related to the complexity of the pore structure and affect the speed of subsequent calculations. The more complex the pore structure, the slower the simulation calculation. To improve the calculation speed, the number of vertices and other parameters can be appropriately simplified while meeting the accuracy requirements.

[0036] Step 3-2: During the simulated displacement process, design a laminar flow model. Whether it is multiphase flow or unidirectional flow, the flow equations of the fluid property model should follow the following equations:

[0037]

[0038]

[0039] The equation for the boundary condition wall is: u = 0;

[0040] Entry equation: u = -U0n;

[0041] Exit equation:

[0042] Symmetric equation: u·n=0;

[0043] K n -(K n ·n)n=0,K n =Kn;

[0044] Step 3-3: In the simulated displacement process, design a rare substance transport model and the transport property equation:

[0045]

[0046] Equation for zero flux: -n·J i =0;

[0047] Inflow equation: c i =c 0,i ;

[0048] Outflow equation:

[0049] Symmetric equation: -n·J i =0;

[0050] In the above formula, t is time, ρ is fluid density, u is the fluid velocity vector, p is pressure, I is the unit tensor, K is the stress tensor, F is the body force, Uon is the given velocity at the inlet, Kn is the stress tensor, c is the concentration of the dilute substance, and c i It is the concentration of the i-th dilute substance, c oi It refers to the concentration of dilute substances at the inflow boundary. Let R denote the gradient operator, D denote the dilute material diffusion coefficient, n be the unit normal vector at the outlet boundary, and R be the gradient operator. i Rate of change of concentration of dilute substance, J i This represents the flux of rare substances.

[0051] Steps 3-4: Simulation includes two discretized physical fields: laminar flow and rare material transport. Transient equations are established using linear Lagrangian element geometric functions to calculate the changes in concentration, pressure, and velocity fields. Finally, a three-dimensional model is used to dynamically display the characteristics of crude oil flow in the core, the distribution characteristics of the remaining oil enrichment zone, and calculate the oil displacement efficiency.

[0052] Specific Implementation Method Two: Combination Figures 2 to 8 This embodiment describes a simulation using the carbon dioxide flooding simulation method based on a digital core model as described in Specific Embodiment 1: an on-site implementation was carried out using Oilfield A as an example.

[0053] (1) Establish a digital core model;

[0054] like Figure 2 and Figure 3 As shown, a cuboid model with a size of 2*2*6cm was designed based on the pore structure of the actual core of the oilfield, namely: a three-dimensional digital core model. The model is set with 55759 grid vertices, 111650 triangles, 5638 edge elements, and 549 vertex elements.

[0055] (2) Simulation of miscible and immiscible displacement;

[0056] like Figures 4 to 6 As shown, where, Figure 4 This is a graph showing the concentration distribution at time T = 2 seconds. Figure 5 The figures show the velocity distribution at time T = 2 s and the pressure distribution at time T = 0.69 s. Under both miscible and immiscible conditions, and with different single variables such as crude oil density, dynamic viscosity, and CO2 injection rate, the dynamic and intuitive characteristics of crude oil distribution, pressure distribution, and other parameters within the pore structure space are dynamically and intuitively displayed.

[0057] (3) Evaluation of results;

[0058] like Figure 7 and Figure 8 As shown, Figure 7 This is a schematic diagram illustrating the change in displacement rate over time. Figure 8 This is a diagram showing the volume fraction distribution of the fluid at time T = 0.5 s. By utilizing the volume fraction distribution during the displacement process, the displacement rate under different displacement conditions can be dynamically evaluated, and parameters can be dynamically adjusted to achieve an intuitive and rapid evaluation of the displacement effect and select the optimal solution.

[0059] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0060] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A simulation method for carbon dioxide flooding based on a digital core model, characterized by: This method is implemented by the following steps: Step 1: Construct a simulated digital core model based on the pore structure characteristics of the actual core. Step 2: Using the digital core model described in Step 1, combined with geomechanics, single-well fracture analysis and attribute model, output pore structure parameters such as porosity, permeability, tortuosity and specific surface area through machine deep learning and artificial intelligence algorithms; Step 3: Use multi-parameter single-variable simulation to simulate the CO2 displacement process in underground oil reservoirs and obtain the oil displacement efficiency, flow characteristics and residual oil distribution characteristics in the pore structure space.

2. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: The specific process of step one is as follows: Step 1-1: Perform high-resolution CT digital imaging on small rock fragments to obtain the three-dimensional spatial structure of the opaque porous medium and information on fluid flow and transport within it; Steps 1-2: Process the original image using noise reduction, smoothing, and segmentation methods to separate the pores from the rock mineral matrix components and obtain a segmented image; Steps 1-3: Obtain porous medium pore structure information based on the segmented images described in Steps 1-2.

3. The carbon dioxide flooding simulation method based on a digital core model according to claim 2, characterized in that: The pore structure information of the porous medium in steps 1-3 includes the particle size distribution, pore radius distribution, and microcrack distribution of the porous medium.

4. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: In step two, the attribute model includes rock type, rock physical properties, and sandstone-mudstone superposition relationship.

5. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: In step two, pore structure parameters such as porosity, tortuosity, and specific surface area are obtained by calculating and processing digital core samples using an indirect method. The Kozeny-Carman formula or pore network model is then used to calculate the seepage properties of the porous medium.

6. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: In step two, a direct method is used to extract the pore structure that characterizes the pore shape and wall wettability of the porous medium, and MATLAB software is used to indirectly calculate the pore structure parameters such as porosity, permeability, tortuosity, and specific surface area from the CT image data.

7. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: In step three, the multiple parameters and single variables include crude oil density, dynamic viscosity, CO2 injection rate, and diffusion coefficient.

8. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: The displacement process in step three is as follows: Step 3:

1. Set the parameters of the digital core model, including the number of vertices, triangles, edge elements, and vertex elements; Step 32: Design a laminar flow model based on the parameters set in Step 31; Step 3: Using the laminar flow model from Step 2, simulate the two discretized physical fields, including laminar flow and rarefied mass transfer. Establish transient equations using linear Lagrangian element geometric functions to obtain the changes in concentration, pressure, and velocity fields.

9. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: In step 3, based on the changes in concentration, pressure, and velocity fields, the characteristics of crude oil flow and the distribution of remaining oil-rich areas in the core are displayed in real time using a three-dimensional model, and the oil displacement efficiency is calculated.

10. The carbon dioxide flooding simulation method based on a digital core model according to claim 1, characterized in that: By utilizing the volume fraction distribution during the displacement process, the displacement rate under different displacement conditions is dynamically evaluated, and the parameter variables are dynamically adjusted.

Images