Method for grading mobilization of remaining oil driven by potential energy gradient field and well fracture communication

Abstract

The application discloses a remaining oil grading production method based on potential energy gradient field and well fracture connectivity combined driving, and relates to the technical field of oil and gas field development. The application firstly establishes a reservoir numerical model in a reservoir numerical simulator, adjusts the reservoir numerical model by fitting production history data of simulation data of the reservoir numerical model, obtains key field distribution conditions of the reservoir, then quantitatively describes a crude oil migration process according to fluid properties of the reservoir, introduces driving force and retarding force, establishes a crude oil displacement potential energy model, couples the crude oil displacement potential energy model and the reservoir numerical model, obtains potential energy gradient data and well fracture connectivity coefficient data of the reservoir, obtains a plurality of reservoir characteristic vectors after preprocessing to establish a reservoir characteristic data set, performs double-parameter clustering analysis based on a K-means clustering algorithm, obtains classification results of each reservoir grid and determines a production grade of the reservoir grid, and the accurate acquisition of remaining oil distribution and scientific grading of production potential are realized.

Description

Technical Field

[0001] This invention relates to the field of oil and gas field development technology, specifically to a method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity. Background Technology

[0002] With the continued growth in demand for oil and gas resources, most oil and gas fields have gradually entered the mid-to-late stages of development, and the focus of exploitation is gradually shifting to development adjustment and potential tapping. Considering the heterogeneity of complex reservoirs and the impact of long-term water-drive, gas-drive, and fracturing development measures during the development process, a large amount of residual oil remains in the reservoirs. Moreover, as oil and gas resources are gradually exploited, the distribution of residual oil becomes increasingly dispersed, significantly increasing the difficulty of its utilization. Therefore, accurately characterizing the spatial distribution of residual oil and effectively utilizing it has become a core challenge for stabilizing oil and gas field production and improving recovery rates. The distribution of residual oil is controlled by multiple factors such as geological structure, sedimentary microfacies, and development history, exhibiting strong heterogeneity and dynamic evolution characteristics, leading to problems such as low accuracy and insufficient reliability in traditional residual oil distribution prediction methods. Furthermore, existing technologies mostly focus on single parameters such as residual oil saturation or oil abundance, making it difficult to comprehensively reflect the distribution law of residual oil under the coupling of multiple factors, which seriously restricts the pertinence of development adjustment measures and affects the overall development benefits and resource continuity capacity of oil and gas fields.

[0003] Currently, research on residual oil analysis and utilization in oilfields mainly revolves around four methods: static geological analysis, production dynamic analysis, reservoir fine description, and reservoir numerical simulation. Static geological analysis primarily relies on static geological data such as drilling cores and well logging, experimentally determining the residual oil saturation of the reservoir to characterize the local distribution of residual oil. While direct and reliable, this method is costly and yields discontinuous residual oil saturation information, making it difficult to achieve a systematic evaluation at the entire oilfield scale. Production dynamic analysis, based on oilfield production data and combined with the principles of material balance and dynamic response characteristics, infers the macroscopic distribution and development potential of residual oil, offering good real-time performance and a global perspective. However, its solutions are usually not unique and struggle to finely characterize the spatial distribution details of residual oil. Reservoir fine description integrates multi-source information from geology, geophysics, well logging, and testing, using reservoir modeling and geostatistics for semi-quantitative and quantitative description. However, its accuracy relies excessively on expert experience and subjective judgment, and its repeatability and objectivity need improvement. Numerical simulation methods for oil reservoirs integrate the advantages of the aforementioned methods. By constructing a three-dimensional geological model and performing historical data fitting, they predict the distribution of remaining oil through reservoir parameter inversion, guiding adjustments in reservoir development. However, this method requires extremely high accuracy in historical data fitting and is constrained by computational efficiency and model complexity, thus facing significant challenges in practical applications. In summary, existing methods for analyzing and utilizing remaining oil in oilfields have limitations in terms of accuracy, cost, efficiency, and applicability, hindering the efficient tapping of remaining oil potential.

[0004] Therefore, there is an urgent need to propose a method for the graded utilization of remaining oil based on the combined driving force of potential energy gradient field and wellbore connectivity. This method should integrate multidisciplinary data and be accurate, economical, and operable, so as to achieve accurate acquisition of the distribution of remaining oil and scientific classification of its utilization potential. Summary of the Invention

[0005] This invention aims to solve the above problems and proposes a method for graded utilization of remaining oil based on the combined driving force of potential energy gradient field and well-fracture connectivity. This method achieves accurate quantitative prediction of the distribution of remaining oil and scientific classification of utilization potential, effectively guiding the utilization potential of remaining oil at each stage of oilfield development and providing clear guidance for reservoir development and adjustment.

[0006] The present invention adopts the following technical solution:

[0007] The residual oil stage recovery method based on the combined drive of potential energy gradient field and wellbore connectivity includes the following steps:

[0008] Step 1: Integrate drilling, logging, and geological data of the reservoir to establish a reservoir numerical model in the reservoir numerical simulator;

[0009] Step 2: Based on the reservoir production test data and fracturing operation monitoring data, use the reservoir numerical model to perform numerical simulation to obtain reservoir production simulation data, and fit it with the actual historical production data of the reservoir. Adjust the reservoir numerical model until the consistency between the simulated reservoir production data and the actual historical production data reaches the preset requirements. Use the adjusted reservoir numerical model to simulate and obtain the distribution of key fields in the reservoir.

[0010] Step 3: Based on the fluid properties within the reservoir, quantitatively describe the crude oil migration process and introduce driving force and retention force to establish a crude oil displacement potential energy model;

[0011] Step 4: Based on the crude oil displacement potential energy model and the adjusted reservoir numerical model, obtain the potential energy distribution field of the entire reservoir area, calculate the potential energy gradient at each location of the reservoir by spatial differentiation, obtain the potential energy gradient field of the reservoir, and obtain the potential energy gradient data of the reservoir.

[0012] Step 5: Based on the streamline simulation method, the overall streamline distribution inside the reservoir is simulated using the adjusted reservoir numerical model. The connectivity coefficient of each reservoir grid in the reservoir numerical model is calculated based on the fluid flux to obtain the well-fracture connectivity coefficient data of the reservoir.

[0013] Step 6: After preprocessing the reservoir potential gradient data and well fracture connectivity coefficient data, multiple reservoir feature vectors are obtained, a reservoir feature dataset is established, and a two-parameter clustering analysis is performed based on the K-means clustering algorithm to obtain the classification results of each reservoir grid in the reservoir numerical model.

[0014] Step 7: Based on the classification results of each reservoir grid in the reservoir numerical model, obtain the remaining oil tapping potential value of each region of the reservoir, and determine the utilization level of each region of the reservoir by combining the preset remaining oil grading and utilization standards.

[0015] Preferably, in step 1, a structural framework for a three-dimensional geological model is first constructed based on the well trajectory, stratification, and fault data of the reservoir. Combining conventional logging curves, imaging logging data, and reservoir fine description results, a phase control model is set in the structural framework to obtain a three-dimensional geological model, which is used to simulate the spatial distribution of porosity, permeability, and fluid saturation within the reservoir. Then, a reservoir fluid model is set based on the fluid property parameters of the reservoir, and the reservoir fluid model is embedded into the three-dimensional geological model to establish a reservoir numerical model, which is used to simulate reservoir production test data.

[0016] Preferably, in step 2, the reservoir numerical model is adjusted according to the actual production history data of the reservoir, so that the consistency between the reservoir production simulation data obtained by the reservoir numerical model and the actual production history data is not less than 95%, and the consistency between the fracture propagation morphology, fracture network volume and post-fracturing seepage capacity of the reservoir numerical model and the reservoir fracturing construction monitoring data is not less than 95%. Thus, the adjusted reservoir numerical model is used to simulate and obtain the distribution of key fields of the reservoir, including the oil saturation field, pressure field, fluid velocity field, interfacial tension field, viscous resistance field and flow resistance field.

[0017] Preferably, the crude oil displacement potential energy model includes pressure energy, kinetic energy, interfacial energy, viscous resistance energy, and flow resistance energy, wherein the pressure energy and kinetic energy are crude oil driving force energy, and the interfacial energy, viscous force energy, and flow resistance energy are crude oil retention force energy.

[0018] The crude oil displacement potential energy model is set as follows:

[0019] ;

[0020] in,

[0021] ;

[0022] ;

[0023] ;

[0024] ;

[0025] ;

[0026] In the formula, Displacement potential energy for crude oil; It is pressure energy; The equivalent pressure of the oil; This is the initial pressure of the oil phase; The relative permeability of the oil phase; The dynamic viscosity of crude oil; This is the function for calculating the oil phase density; Formation pressure; Kinetic energy; This is the function for calculating oil phase velocity. For interfacial force energy; The interfacial tension of the oil phase; The wetting angle between the oil phase and the rock; The radius of the rock pore capillary; It is viscous energy; This represents the oil phase migration velocity gradient. This is the energy of flow resistance; This is the drag coefficient along the route. , Reynolds number for oil phase transport; The length of the crack; The width of the crack.

[0027] Preferably, in step 4, the key field parameters at each location of the reservoir obtained from the reservoir numerical model simulation are substituted into the crude oil displacement potential energy model. The potential energy at each reservoir grid in the reservoir numerical model is calculated using the crude oil displacement potential energy model. The potential energy distribution of the entire reservoir is obtained, and the potential energy gradient at each location of the reservoir numerical model is obtained by spatial differentiation. The reservoir potential energy gradient field is established, and the potential energy gradient data of the reservoir is obtained.

[0028] The reservoir potential energy gradient field is expressed as:

[0029] ;

[0030] ;

[0031] in,

[0032] ;

[0033] ;

[0034] ;

[0035] In the formula, These are the location coordinates of the reservoir grid in the reservoir numerical model. The x-coordinate of the reservoir grid is denoted as . The vertical coordinate of the reservoir grid is denoted as . The vertical coordinates of the reservoir grid; Reservoir mesh in reservoir numerical model Potential energy vector field at the location; For the del operator; Reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For modulus calculation; , , The potential energy gradients at... direction, direction, Components in direction; , , The potential energy scalar field is respectively direction, direction, Partial derivatives in the direction; , , They are respectively direction, direction, Basis vectors in the direction; , , For adjacent reservoir grids in direction, direction, Spacing in the direction; Reservoir mesh in reservoir numerical model The potential energy scalar field at that location, Reservoir mesh in reservoir numerical model Potential energy scalar field at the location; Reservoir mesh in reservoir numerical model The potential energy scalar field at that location, Reservoir mesh in reservoir numerical model Potential energy scalar field at the location; Reservoir mesh in reservoir numerical model The potential energy scalar field at that location, Reservoir mesh in reservoir numerical model Potential energy scalar field at the location; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction.

[0036] Preferably, in step 5, the formula for calculating the connectivity coefficient is:

[0037] ;

[0038] In the formula, For the first The connectivity coefficient of each reservoir grid; Streamline number; This represents the total number of streamlines. For the first The volumetric flow rate carried by each streamline; For streamline indicator functions, when the first... The streamline passes through the first When a reservoir grid is used, The value is 1, otherwise, The value is 0.

[0039] Preferably, in step 6, firstly, the potential energy gradient data and wellbore connectivity coefficient data of the reservoir are homogenized so that the values ​​of all potential energy gradient data and wellbore connectivity coefficient data are within the range of... Then, based on the potential energy gradient and well-fracture connectivity coefficient of each reservoir network in the reservoir numerical model, the reservoir feature vector of each reservoir grid is obtained. Based on the reservoir feature vectors of all reservoir grids in the reservoir numerical model as sample data points, a reservoir feature dataset is established.

[0040] The reservoir feature dataset is divided into Clustering, resulting in Given cluster centers, the objective function is constructed by minimizing the sum of squared distances between all sample data points and their respective cluster centers. ,get:

[0041] ;

[0042] In the formula, Cluster number; The total number of clusters; For the first Each reservoir feature vector ,in, For the first Potential energy gradient of each reservoir grid, For the first The connectivity coefficient of a reservoir grid. It is the transpose matrix; For the first A cluster set; For the first The coordinates of each cluster center along the potential gradient dimension; For the first The coordinates of each cluster center along the connectivity coefficient dimension;

[0043] The objective function is optimized based on the K-means clustering algorithm. In each optimization process, the cluster centers of each cluster are first determined. Then, all reservoir grids in the reservoir numerical model are traversed, and each reservoir grid is assigned to the cluster set whose cluster center has the smallest squared Euclidean distance. Each temporary cluster is used to obtain the centroid of each temporary cluster and use it as the cluster center. When the cluster center of each temporary cluster meets the preset optimization termination criteria, the optimization of the objective function is stopped, and the classification results of each reservoir grid in the reservoir numerical model are obtained.

[0044] Preferably, the optimization termination criterion is set based on two principles: local optimal solution and engineering protection. When the cluster centers of each temporary cluster no longer change or the change is less than the preset minimum tolerance threshold after two consecutive optimizations, or when the current number of optimizations reaches the preset maximum number of optimizations, the preset optimization termination criterion is met, the clustering and classification of all sample data points in the reservoir feature dataset is completed, and the optimization of the objective function is stopped.

[0045] Preferably, in step 7, the internal area of ​​the reservoir is divided into primary, secondary, tertiary, and quaternary exploitation zones based on the remaining oil exploitation potential value. Specifically, when the remaining oil exploitation potential value of the reservoir area is greater than 0.8 and does not exceed 1, the reservoir area is determined to be a primary exploitation zone; when the remaining oil exploitation potential value of the reservoir area is greater than 0.6 and does not exceed 0.8, the reservoir area is determined to be a secondary exploitation zone; when the remaining oil exploitation potential value of the reservoir area is greater than 0.4 and does not exceed 0.6, the reservoir area is determined to be a tertiary exploitation zone; and when the remaining oil exploitation potential value of the reservoir area is greater than 0 and does not exceed 0.4, the reservoir area is determined to be a quaternary exploitation zone.

[0046] The present invention has the following beneficial effects:

[0047] (1) This invention proposes a method for graded utilization of remaining oil based on the combined driving of potential energy gradient field and well-fracture connectivity. In view of the problem that the remaining oil in underground reservoirs is not well understood and the utilization rules are unclear, this invention introduces a dual analysis method of potential energy gradient and well-fracture connectivity coefficient, coupled with a two-parameter clustering method, so that the judgment results of the graded utilization of remaining oil are more accurate, and provides a basis for the evaluation of graded utilization of remaining oil in complex reservoirs.

[0048] (2) This invention proposes a method for graded utilization of residual oil based on the combined drive of potential energy gradient field and well fracture connectivity. It gets rid of the limitation of judging the potential value of residual oil based on the traditional residual oil saturation or residual oil abundance. It introduces the fluid dynamics method and realizes the method of studying residual oil by coupling displacement kinetic energy and retention kinetic energy. It solves the problem of low accuracy in traditional residual oil research and improves the analysis accuracy of residual oil potential value.

[0049] (3) This invention proposes a method for graded utilization of remaining oil based on the combined driving of potential energy gradient field and well-fracture connectivity. Based on the clear distribution of remaining oil, the well-fracture connectivity of reservoir is introduced. The remaining oil is graded and utilized by combining the grid connectivity of each reservoir and coupling a two-parameter clustering algorithm. Finally, the accurate determination of the graded utilization of remaining oil is realized. This provides a technical means to guide the accurate acquisition of the utilization level of remaining oil in the middle and late stages of complex oilfield development and has broad practical engineering application value. Attached Figure Description

[0050] Figure 1 This is a flowchart of the residual oil stratification and mobilization method based on the combined driving of potential energy gradient field and wellbore connectivity according to the present invention.

[0051] Figure 2 This is a flowchart of the two-parameter clustering analysis method based on the K-means clustering algorithm of this invention.

[0052] Figure 3 This is a porosity distribution diagram of a typical well group reservoir according to the present invention.

[0053] Figure 4 This is a permeability distribution diagram of a typical well group reservoir according to the present invention.

[0054] Figure 5 The figure shows the distribution fields of crude oil driving force energy and crude oil retention force energy in the entire area of ​​a typical well group. In the figure, (a) is the pressure energy distribution field of a typical well group, (b) is the kinetic energy distribution field of a typical well group, (c) is the interface energy distribution field of a typical well group, and (d) is the viscous force energy distribution field of a typical well group.

[0055] Figure 6 This is the crude oil displacement potential energy distribution field of a typical well group in this invention.

[0056] Figure 7 The diagram shows the results of the reservoir's residual oil stratification and mobilization analysis.

[0057] Figure 8 This is a comparison chart showing the residual oil grading and utilization effects of the method of the present invention and the traditional residual oil grading and utilization method.

[0058] In the diagram, Z1, Z2, Z3, Z4, Z5, and Z6 are all production wells, while S1 and S2 are both injection wells. Detailed Implementation

[0059] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0060] Example 1

[0061] This invention proposes a method for staged utilization of residual oil based on the combined driving force of potential energy gradient field and wellbore connectivity, such as... Figure 1 As shown, the specific steps include:

[0062] Step 1: Integrate drilling, logging, and geological data of the reservoir to establish a reservoir numerical model in the reservoir numerical simulator.

[0063] Specifically, a three-dimensional geological model framework is first constructed based on well trajectory, stratification, and fault data of the reservoir. This framework is then combined with conventional logging curves, imaging logging data, and detailed reservoir description results (e.g., sedimentary facies description and sand body geometry description) to establish a facies control model within the framework. Using sequential indicator modeling and geostatistics theory, a three-dimensional geological model is built in the commercial numerical simulation software Petrel to simulate the spatial distribution of porosity, permeability, and fluid saturation within the reservoir. Next, a reservoir fluid model is established based on the reservoir's fluid property parameters, including oil density, crude oil viscosity, crude oil saturation pressure, formation volume factor, and gas-oil ratio. Using laboratory experimental data such as constant composition expansion and differential analysis, and fitted with the PR equation of state, a reservoir fluid model that accurately describes fluid phase changes and physical properties is established for fluid analysis. Finally, the reservoir fluid model is embedded into the three-dimensional geological model to establish a reservoir numerical model for simulating reservoir production test data.

[0064] Step 2: Based on the reservoir production test data and fracturing operation monitoring data, use the reservoir numerical model to perform numerical simulation to obtain reservoir production simulation data, and fit it with the actual historical production data of the reservoir. Adjust the reservoir numerical model until the consistency between the reservoir production simulation data obtained by the reservoir numerical model and the actual historical production data reaches the preset requirements. Use the adjusted reservoir numerical model to simulate and obtain the distribution of key fields in the reservoir.

[0065] Specifically, the reservoir numerical model is adjusted based on the actual historical production data of the reservoir to ensure that the consistency between the simulated reservoir production data and the actual historical production data is no less than 95%. This means that the overall consistency between the simulated oil production, water production, bottom hole pressure, and other key dynamic indicators and the actual production data must be no less than 95%. This ensures the high reliability of the reservoir numerical model in terms of macroscopic production dynamics. If the consistency between the simulated reservoir production data and the actual historical production data does not reach this threshold, the reservoir numerical model must be calibrated according to the principles of systematic approach and geological prior constraints. First, parameters affecting the global material and energy balance are calibrated, mainly by fine-tuning the average porosity, net-to-gross ratio, or initial oil saturation to match the overall pressure trend and total production. Secondly, the parameters controlling fluid flow capacity are modified, including scaling or anisotropically adjusting the permeability field and optimizing the shape and endpoint values ​​of the relative permeability curve. This is used to match the single-well production capacity (oil production), bottom hole pressure, and water cut increase patterns, so that the reservoir numerical model can accurately reflect the overall migration and energy changes of underground fluids.

[0066] Meanwhile, the fitting rate of the fracturing pumping program was set at 0.95, meaning that the agreement between the fracture propagation morphology, fracture network volume, and post-fracturing seepage capacity of the reservoir numerical model and the reservoir fracturing construction monitoring data should not be less than 95%. If the fitting rate of the fracturing well pumping program does not reach this threshold, the reservoir numerical model needs to be adjusted and calibrated based on fracturing construction data and geological interpretation data. First, the mechanical model controlling the macroscopic morphology of fractures is calibrated, mainly adjusting the geostress field and rock mechanical properties. Second, the engineering parameters controlling the construction response and fracture network volume are calibrated, mainly adjusting the fracturing fluid performance. This achieves a reliable representation of the artificial fracture system in the reservoir numerical model, laying the foundation for accurately simulating the flow behavior in the fracturing-stimulated area.

[0067] Finally, the adjusted reservoir numerical model was used for simulation to obtain the distribution of key fields in the reservoir, including the oil saturation field, pressure field, fluid velocity field, interfacial tension field, viscous resistance field, and flow resistance field. This provides a reliable data foundation and simulation basis for subsequent construction of the potential energy gradient field and in-depth analysis of the reservoir driving mechanism and remaining oil distribution.

[0068] Step 3: Based on the fluid properties in the reservoir, fully consider the influence of formation pressure, capillary force, oil saturation, crude oil viscosity and flow resistance on the flow of crude oil in the reservoir, quantitatively describe the crude oil migration process and introduce driving force and retention force, fully consider the reservoir properties and crude oil flow characteristics, and establish a crude oil displacement potential energy model.

[0069] Specifically, the crude oil displacement potential energy model includes pressure energy, kinetic energy, interfacial energy, viscous resistance energy, and flow resistance energy, wherein the pressure energy and kinetic energy are crude oil driving force energy, and the interfacial energy, viscous force energy, and flow resistance energy are crude oil retention force energy.

[0070] The crude oil displacement potential energy model is set as follows:

[0071] ;

[0072] in,

[0073] ;

[0074] ;

[0075] ;

[0076] ;

[0077] ;

[0078] In the formula, The displacement potential energy of crude oil is expressed in units of... ; Pressure energy, unit: ; The oil equivalent pressure is expressed in units of... ; The initial pressure of the oil phase, in units of ; This refers to the relative permeability of the oil phase. The dynamic viscosity of crude oil is expressed in units of... ; This is the function for calculating the density of the oil phase, with units of... ; Formation pressure, unit: ; Kinetic energy, unit: ; This is the function for calculating oil phase velocity, in units of... ; Interfacial force energy, unit: ; The interfacial tension of the oil phase is expressed in units of... ; The wetting angle between the oil phase and the rock, in degrees; The radius of the rock pore capillary is given in units of 1. ; Viscous energy, unit: ; This represents the oil phase migration velocity gradient, in units of... ; Flow resistance energy, unit: ; This is the drag coefficient along the route. , The Reynolds number for oil phase transport; The crack length is expressed in units of 1 / 2. ; The crack width is expressed in units of 1000 mm. .

[0079] Step 4: Based on the crude oil displacement potential energy model and the adjusted reservoir numerical model, obtain the potential energy distribution field of the entire reservoir. Calculate the potential energy gradient at each location in the reservoir by spatial differentiation to obtain the potential energy gradient field of the reservoir and obtain the potential energy gradient data of the reservoir.

[0080] Furthermore, the key field parameters at various locations in the reservoir obtained from the reservoir numerical model simulation are substituted into the crude oil displacement potential energy model. The potential energy at each reservoir grid in the reservoir numerical model is calculated using the crude oil displacement potential energy model, and the potential energy distribution of the entire reservoir is obtained. This potential energy field has high potential areas and low potential areas, and there is a certain deviation in the potential energy between each reservoir grid. By spatially differentiating the potential energy distribution field of the entire reservoir, the potential energy gradient at each location in the reservoir numerical model is obtained, and the reservoir potential energy gradient field is established to obtain the potential energy gradient data of the reservoir.

[0081] Specifically, in order to obtain the specific value of the potential energy gradient field, the potential energy gradient field of the reservoir is discretized and calculated based on the finite difference method for different reservoir grids, and expressed as:

[0082] ;

[0083] ;

[0084] in,

[0085] ;

[0086] ;

[0087] ;

[0088] In the formula, These are the location coordinates of the reservoir grid in the reservoir numerical model. The x-coordinate of the reservoir grid is denoted as . The vertical coordinate of the reservoir grid is denoted as . The vertical coordinates of the reservoir grid; For reservoir mesh in reservoir numerical model Potential energy vector field at the location; For the del operator; For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For modulus calculation, it is used to represent the value of reservoir potential energy gradient; , , The potential energy gradients at... direction, direction, Components in direction; , , The potential energy scalar field is respectively direction, direction, Partial derivatives in the direction; , , They are respectively direction, direction, Basis vectors in the direction; , , For adjacent reservoir grids in direction, direction, Spacing in the direction; Reservoir mesh in reservoir numerical model The potential energy scalar field at that location, For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; Reservoir mesh in reservoir numerical model The potential energy scalar field at that location, Reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For reservoir mesh in reservoir numerical model The potential energy scalar field at that location, For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction.

[0089] Step 5: Based on the streamline simulation method, the overall streamline distribution inside the reservoir is simulated using the adjusted reservoir numerical model. The connectivity coefficient of each reservoir grid in the reservoir numerical model is calculated according to the fluid flux, and the well-fracture connectivity coefficient data of the reservoir is obtained. That is, the fluid flux flowing through the streamlines in each reservoir grid is determined, and then the connectivity between wells and fractures is evaluated.

[0090] Specifically, the formula for calculating the connectivity coefficient is as follows:

[0091] ;

[0092] In the formula, For the first The connectivity coefficient of a reservoir grid; Streamline number; This represents the total number of streamlines. For the first The volumetric flow rate carried by each streamline; For streamline indicator functions, when the first... The streamline passes through the first When a reservoir grid is used, The value is 1, otherwise, The value is 0.

[0093] Step 6: After preprocessing the reservoir potential gradient data and wellbore connectivity coefficient data, multiple reservoir feature vectors are obtained, and a reservoir feature dataset is established. A two-parameter clustering analysis is then performed on the reservoir feature dataset based on the K-means clustering algorithm, such as... Figure 2 As shown, the classification results of each reservoir grid in the reservoir numerical model are obtained.

[0094] Furthermore, firstly, to ensure the fairness and effectiveness of the cluster analysis, the potential energy gradient data and wellbore connectivity coefficient data of the reservoir were homogenized so that the values ​​of all potential energy gradient data and wellbore connectivity coefficient data were within the range of... Then, based on the potential energy gradient and well-fracture connectivity coefficient of each reservoir network in the reservoir numerical model, the reservoir feature vector of each reservoir grid is obtained. Based on the reservoir feature vectors of all reservoir grids in the reservoir numerical model as sample data points, a reservoir feature dataset is established.

[0095] Based on prior knowledge from field experience or reservoir engineering practice, the total number of clusters is set, and the reservoir feature dataset is divided into clusters. Clustering, resulting in Given cluster centers, the objective function is constructed by minimizing the sum of squared distances between all sample data points and their respective cluster centers. ,get:

[0096] ;

[0097] In the formula, Cluster number; The total number of clusters; For the first Each reservoir feature vector ,in, For the first Potential energy gradient of each reservoir grid, For the first The connectivity coefficient of a reservoir grid. It is the transpose matrix; For the first A cluster set; For the first The coordinates of each cluster center along the potential gradient dimension; For the first The coordinates of each cluster center along the connectivity coefficient dimension.

[0098] The objective function is optimized based on the K-means clustering algorithm. In each optimization process, the cluster centers of each cluster are first determined. Then, all reservoir grids in the reservoir numerical model are traversed, and each reservoir grid is assigned to the cluster set whose cluster center has the smallest squared Euclidean distance. Each temporary cluster is used to obtain the centroid of each temporary cluster and use it as the cluster center. When the cluster center of each temporary cluster meets the preset optimization termination criterion, the optimization of the objective function is stopped, and the classification results of each reservoir grid in the reservoir numerical model are obtained.

[0099] Specifically, the optimization termination criterion is set based on two principles: local optimal solution and engineering protection. When the cluster centers of each temporary cluster no longer change or the change is less than the preset minimum tolerance threshold after two consecutive optimizations, or when the current number of optimizations reaches the preset maximum number of optimizations, the preset optimization termination criterion is met, the clustering and classification of all sample data points in the reservoir feature dataset is completed, and the optimization of the objective function is stopped.

[0100] Step 7: Based on the classification results of each reservoir grid in the reservoir numerical model, obtain the remaining oil exploitation potential value of each region of the reservoir, and determine the exploitation level of each region of the reservoir by combining the preset remaining oil classification and exploitation standards.

[0101] Specifically, based on the remaining oil exploitation potential, the internal area of ​​the reservoir is divided into four levels: Level 1, Level 2, Level 3, and Level 4 exploitation zones. When the remaining oil exploitation potential of a reservoir area is greater than 0.8 but not more than 1, the reservoir area is classified as a Level 1 exploitation zone, and strategies focused on efficiency improvement and regulation are adopted, including optimizing injection and production parameters, converting production fractures to water injection fractures, and deep-seated displacement. When the remaining oil exploitation potential of a reservoir area is greater than 0.6 but not more than 0.8, the reservoir area is classified as a Level 2 exploitation zone, and strategies focused on engineering modifications and pressure enhancement are adopted. The core measures include implementing targeted repeated fracturing and gas injection miscible flooding to utilize remaining oil; when the remaining oil potential value of a reservoir area is greater than 0.4 but not more than 0.6, the reservoir area is determined to be a Level III utilization area, and measures to replenish formation capacity are implemented, with continuous gas injection and continuous water injection as the main methods for utilizing remaining oil; when the remaining oil potential value of a reservoir area is greater than 0 but not more than 0.4, the reservoir area is determined to be a Level IV utilization area, and measures to utilize remaining oil are implemented with strategic protection or economic assessment as the guiding principle, with control of inefficient operations and conversion to chemical flooding as the main methods for utilizing remaining oil.

[0102] Example 2

[0103] This embodiment takes a typical well group in a low-permeability block as an example. The typical well group includes eight production wells, namely production well Z1, production well Z2, production well Z3, production well Z4, production well Z5, production well Z6, injection well S1, and injection well S2. Among them, the production wells are vertical wells and the injection wells are fractured horizontal wells.

[0104] For this typical well group, a residual oil grading and utilization analysis was conducted. First, field data for this typical well group was collected, including well data, geological and logging data, fluid and high-pressure property data, reservoir oil-water well injection and production data, and fracturing monitoring data. The well data included well structure, precise coordinates, full well trajectory, and completion / perforation information. The geological and logging data included stratigraphic and fault data, environmentally corrected conventional and imaging logging curves, and core analysis data. The fluid and high-pressure property data included surface fluid properties, representative formation fluid PVT experimental data, and original oil-water interface and pressure / temperature. The reservoir oil-water well injection and production data included injection and production rates, bottom hole pressure monitoring, and production test data. The fracturing monitoring data included fracturing fluid data, fracturing well construction curves, microseismic monitoring, and post-fracturing response data.

[0105] Based on the collected data, the residual oil staged utilization method proposed in Example 1, which is driven by the combined potential energy gradient field and wellbore connectivity, was used to analyze the residual oil staged utilization. The specific steps include:

[0106] Step 1: Integrate drilling, logging, and geological data of the reservoir to establish a reservoir numerical model in the reservoir numerical simulator.

[0107] In this embodiment, a three-dimensional geological model framework is constructed in the commercial numerical simulation software Petrel based on well trajectory, stratification, and fault data of the reservoir. This framework is then combined with conventional logging curves, imaging logging data, and detailed reservoir description results. Furthermore, facies-controlled models within the framework are set using detailed reservoir description results (e.g., sedimentary facies description, sand body geometry description). Sequential indicator modeling methods and geostatistics theory are applied to establish a three-dimensional geological model in Petrel to simulate the spatial distribution of porosity, permeability, and fluid saturation within the reservoir. Figure 3 and Figure 4 As shown, a detailed characterization of the reservoir's structural framework, sedimentary facies zones, and spatial distribution of property parameters was achieved. Then, a reservoir fluid model was established based on the reservoir's fluid property parameters, including oil density, crude oil viscosity, crude oil saturation pressure, formation volume factor, and gas-oil ratio. Using indoor experimental data such as constant composition expansion and differential analysis, and fitted with the PR equation of state, a reservoir fluid model capable of accurately describing fluid phase changes and physical properties was established for fluid analysis. Finally, the reservoir fluid model was embedded into a three-dimensional geological model to establish a reservoir numerical model for simulating reservoir production test data, laying a reliable geological and fluid dynamic foundation for subsequent reservoir history analysis.

[0108] Step 2: In this embodiment, production test data and fracturing operation monitoring data from eight production wells are combined with reservoir numerical model for numerical simulation to obtain production simulation data from the eight production wells. This data is then fitted with the actual historical production data of the reservoir. During the historical fitting process, a graded adjustment strategy is adopted to adjust the reservoir numerical model. First, global material and energy balance parameters (such as pore volume and water energy) are corrected to match the overall production trend. Then, local flow capacity parameters (such as permeability field and relative permeability curve) are adjusted to fit the dynamics of a single well. Finally, fracture mechanical parameters and conductivity are corrected based on fracturing operation data to ensure that the fracture network morphology in the reservoir numerical model is consistent with the monitoring results. The reservoir numerical model is adjusted until the consistency between the reservoir production simulation data obtained by the reservoir numerical model and the actual historical production data reaches the preset requirements. The adjusted reservoir numerical model is used to simulate and obtain the key field distribution of the reservoir, including the oil saturation field, pressure field, fluid velocity field, interfacial tension field, viscous resistance field, and flow resistance field. This provides a reliable data foundation and simulation basis for subsequent construction of potential energy gradient field and in-depth analysis of reservoir driving mechanism and remaining oil distribution.

[0109] Step 3: Based on the fluid properties in the reservoir, fully consider the influence of formation pressure, capillary force, oil saturation, crude oil viscosity and flow resistance on the flow of crude oil in the reservoir, quantitatively describe the crude oil migration process and introduce driving force and retention force, fully consider the reservoir properties and crude oil flow characteristics, and establish a crude oil displacement potential energy model.

[0110] Step 4: Integrating the crude oil displacement potential energy model of typical well groups and the potential energy gradient field simulated by the reservoir numerical model after historical fitting, the reservoir geological attribute field obtained from the typical simulation of the reservoir numerical model after production history fitting is input into the crude oil displacement potential energy model, as shown below. Figure 5 The pressure energy distribution, kinetic energy distribution, interfacial energy distribution, and viscous energy distribution of the entire area are shown. The potential energy distribution field of a typical well group is calculated, as follows: Figure 6 As shown, the potential energy field has high potential and low potential regions, and there is a certain deviation in potential energy between each reservoir grid. By spatially differentiating the potential energy distribution field of the entire reservoir, the potential energy gradient at each location of the reservoir numerical model is obtained, the reservoir potential energy gradient field is established, and the potential energy gradient data of the reservoir is obtained. After analysis, it is found that the high potential energy gradient in this embodiment is mainly distributed around the fractured horizontal well. The remaining oil in this area is highly enriched and has a high potential for utilization.

[0111] Step 5: Based on the streamline simulation method, the overall streamline distribution inside the reservoir where the typical well group is located is simulated using the adjusted reservoir numerical model. The connectivity coefficient of each reservoir grid in the reservoir numerical model is calculated according to the fluid flux, and the well-fracture connectivity coefficient data of the reservoir is obtained. That is, the fluid flux flowing through the streamline in each reservoir grid is determined, and then the connectivity between wells and fractures is evaluated.

[0112] Step 6: After preprocessing the reservoir potential gradient data and well fracture connectivity coefficient data, multiple reservoir feature vectors are obtained, and a reservoir feature dataset is established. Based on the K-means clustering algorithm, a two-parameter clustering analysis is performed on the reservoir feature dataset to obtain the classification results of each reservoir grid in the reservoir numerical model.

[0113] In this embodiment, firstly, to ensure the fairness and effectiveness of the cluster analysis, the reservoir potential energy gradient data and wellbore connectivity coefficient data are homogenized so that the values ​​of all potential energy gradient data and wellbore connectivity coefficient data are within the range of... Then, based on the potential energy gradient and well-fracture connectivity coefficient of each reservoir network in the reservoir numerical model, the reservoir feature vector of each reservoir grid is obtained. Based on the reservoir feature vectors of all reservoir grids in the reservoir numerical model as sample data points, a reservoir feature dataset is established.

[0114] Based on prior knowledge from field experience or reservoir engineering practice, the total number of clusters is set, and the reservoir feature dataset is divided into 4 clusters, resulting in 4 cluster centers. The objective function is constructed by minimizing the sum of squared distances between all sample data points and their respective cluster centers.

[0115] The objective function is optimized based on the K-means clustering algorithm. In each optimization process, the cluster center of each cluster is determined first. All reservoir grids in the reservoir numerical model are traversed, and each reservoir grid is assigned to the cluster set to which the cluster center with the smallest squared Euclidean distance belongs, forming 4 temporary clusters. The centroid of each temporary cluster is obtained and used as the cluster center. When the optimization of the cluster centers of each temporary cluster meets the preset optimization termination criteria, the optimization of the objective function is stopped, and the classification results of each reservoir grid in the reservoir numerical model are obtained.

[0116] Specifically, the optimization termination criterion is set based on two principles: local optimal solution and engineering protection. In this embodiment, the minimum tolerance threshold is set to 0.05 and the maximum number of optimizations is set to 500. When the cluster centers of each temporary cluster no longer change or the change is less than the preset minimum tolerance threshold or the current number of optimizations reaches the preset maximum number of optimizations, the preset optimization termination criterion is met, the clustering and classification of all sample data points in the reservoir feature dataset is completed, and the optimization of the objective function is stopped.

[0117] Step 7: Based on the classification results of each reservoir grid in the reservoir numerical model, obtain the remaining oil exploitation potential value of each region of the reservoir. Combined with the preset remaining oil grading and utilization criteria, determine the utilization level of each region of the reservoir, such as... Figure 7 As shown.

[0118] In this embodiment, the reservoir's internal area is divided into four levels of exploitation zones based on the remaining oil exploitation potential value: Level 1 exploitation zone, Level 2 exploitation zone, Level 3 exploitation zone, and Level 4 exploitation zone. These correspond to high potential energy-high connectivity zone, high potential energy-low connectivity potential enrichment zone, low potential energy-high connectivity energy deficiency zone, and low potential energy-low connectivity inefficient and difficult-to-exploit zone, respectively. Specifically, when the remaining oil exploitation potential value of the reservoir area is greater than 0.8 but not exceeding 1, the reservoir area is determined to be a Level 1 exploitation zone. A strategy focused on efficiency improvement and regulation is adopted, implementing remaining oil exploitation measures primarily involving optimizing injection and production parameters, converting production fractures to water injection fractures, and deep-seated displacement. When the remaining oil exploitation potential value of the reservoir area is greater than 0.6 but not exceeding 1, the area is classified as a Level 1 exploitation zone. When the residual oil potential is 0.8, the reservoir area is classified as a secondary development zone, and measures centered on engineering modification and pressure enhancement are implemented, along with targeted repeated fracturing and gas injection miscible flooding as the main methods for utilizing the remaining oil. When the residual oil potential of the reservoir area is greater than 0.4 but not exceeding 0.6, the reservoir area is classified as a tertiary development zone, and measures focused on supplementing formation capacity are implemented, along with continuous gas injection and continuous water injection as the main methods for utilizing the remaining oil. When the residual oil potential of the reservoir area is greater than 0 but not exceeding 0.4, the reservoir area is classified as a quaternary development zone, and measures guided by strategic protection or economic assessment are implemented, along with controlling inefficient operations and converting to chemical flooding as the main methods for utilizing the remaining oil. Simultaneously, the traditional residual oil tiered development method and the residual oil tiered development method based on the combined driving of potential energy gradient field and wellbore connectivity of this invention were used to develop and adjust the reservoir of this typical well group. The comprehensive oil production capacity of the reservoir after using the two methods was compared. Figure 8 As shown, it was found that the residual oil stratification and recovery method based on the combined driving of potential energy gradient field and wellbore connectivity of the present invention can improve the average oil production rate by about 10.5% and the final recovery rate by about 8% compared with the traditional residual oil stratification and recovery method, thus verifying the effectiveness of the residual oil stratification and recovery method based on the combined driving of potential energy gradient field and wellbore connectivity proposed in this invention.

[0119] In summary, the method of this invention can effectively solve the problem of the difficulty in utilizing remaining oil in the later stages of complex oilfield development, improve the accuracy of remaining oil analysis, and provide a basis for accurately determining the utilization potential and utilization level of remaining oil in oilfields. It has broad practical engineering application value.

[0120] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for staged utilization of residual oil based on the combined driving force of potential energy gradient field and wellbore connectivity, characterized in that, Includes the following steps: Step 1: Integrate drilling, logging, and geological data of the reservoir to establish a reservoir numerical model in the reservoir numerical simulator; Step 2: Based on the reservoir production test data and fracturing operation monitoring data, use the reservoir numerical model to perform numerical simulation to obtain reservoir production simulation data, and fit it with the actual historical production data of the reservoir. Adjust the reservoir numerical model until the consistency between the simulated reservoir production data and the actual historical production data reaches the preset requirements. Use the adjusted reservoir numerical model to simulate and obtain the distribution of key fields in the reservoir. Step 3: Based on the fluid properties within the reservoir, quantitatively describe the crude oil migration process and introduce driving force and retention force to establish a crude oil displacement potential energy model; Step 4: Based on the crude oil displacement potential energy model and the adjusted reservoir numerical model, obtain the potential energy distribution field of the entire reservoir area, calculate the potential energy gradient at each location of the reservoir by spatial differentiation, obtain the potential energy gradient field of the reservoir, and obtain the potential energy gradient data of the reservoir. Step 5: Based on the streamline simulation method, the overall streamline distribution inside the reservoir is simulated using the adjusted reservoir numerical model. The connectivity coefficient of each reservoir grid in the reservoir numerical model is calculated based on the fluid flux to obtain the well-fracture connectivity coefficient data of the reservoir. Step 6: After preprocessing the reservoir potential gradient data and well fracture connectivity coefficient data, multiple reservoir feature vectors are obtained, a reservoir feature dataset is established, and a two-parameter clustering analysis is performed based on the K-means clustering algorithm to obtain the classification results of each reservoir grid in the reservoir numerical model. Step 7: Based on the classification results of each reservoir grid in the reservoir numerical model, obtain the remaining oil tapping potential value in each region of the reservoir, and determine the utilization level in each region of the reservoir by combining the preset remaining oil grading and utilization standards. The crude oil displacement potential energy model includes pressure energy, kinetic energy, interfacial energy, viscous resistance energy, and flow resistance energy, wherein the pressure energy and kinetic energy are crude oil driving force energy, and the interfacial energy, viscous force energy, and flow resistance energy are crude oil retention force energy. The crude oil displacement potential energy model is set as follows: ； in, ； ； ； ； ； In the formula, Displacement potential energy for crude oil; It is pressure energy; The equivalent pressure of the oil; This is the initial pressure of the oil phase; This refers to the relative permeability of the oil phase. The dynamic viscosity of crude oil; This is the function for calculating the oil phase density; Formation pressure; Kinetic energy; This is the function for calculating oil phase velocity. For interfacial force energy; The interfacial tension of the oil phase; The wetting angle between the oil phase and the rock; The radius of the rock pore capillary; It is viscous energy; This represents the oil phase migration velocity gradient. This is the energy of flow resistance; This is the drag coefficient along the route. , The Reynolds number for oil phase transport; The length of the crack; The width of the crack.

2. The method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 1, a structural framework for a three-dimensional geological model is first constructed based on the well trajectory, stratification, and fault data of the reservoir. Combining conventional logging curves, imaging logging data, and reservoir fine description results, a phase control model is set in the structural framework to obtain a three-dimensional geological model, which is used to simulate the spatial distribution of porosity, permeability, and fluid saturation within the reservoir. Then, a reservoir fluid model is set based on the fluid property parameters of the reservoir. The reservoir fluid model is embedded into the three-dimensional geological model to establish a reservoir numerical model, which is used to simulate reservoir production test data.

3. The method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 2, the reservoir numerical model is adjusted based on the actual production history data of the reservoir, so that the consistency between the reservoir production simulation data obtained by the reservoir numerical model and the actual production history data is not less than 95%, and the consistency between the fracture propagation morphology, fracture network volume and post-fracturing seepage capacity of the reservoir numerical model and the reservoir fracturing construction monitoring data is not less than 95%. Thus, the adjusted reservoir numerical model is used to simulate and obtain the distribution of key fields of the reservoir, including the oil saturation field, pressure field, fluid velocity field, interfacial tension field, viscous resistance field and flow resistance field.

4. The method for staged utilization of residual oil based on the combined driving force of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 4, the key field parameters at each location of the reservoir obtained from the reservoir numerical model simulation are substituted into the crude oil displacement potential energy model. The potential energy at each reservoir grid in the reservoir numerical model is calculated using the crude oil displacement potential energy model. The potential energy distribution of the entire reservoir is obtained, and the potential energy gradient at each location of the reservoir numerical model is obtained by spatial differentiation. The reservoir potential energy gradient field is established, and the potential energy gradient data of the reservoir is obtained. The reservoir potential energy gradient field is expressed as: ； ； in, ； ； ； In the formula, These are the location coordinates of the reservoir grid in the reservoir numerical model. The x-coordinate of the reservoir grid is denoted as . The vertical coordinate of the reservoir grid is denoted as . The vertical coordinates of the reservoir grid; For reservoir mesh in reservoir numerical model Potential energy vector field at the location; For the del operator; For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For modulus calculation; , , The potential energy gradients at... direction, direction, Components in direction; , , The potential energy scalar field is respectively direction, direction, Partial derivatives in the direction; , , They are respectively direction, direction, Basis vectors in the direction; , , For adjacent reservoir grids in direction, direction, Spacing in the direction; For reservoir mesh in reservoir numerical model The potential energy scalar field at that location, For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For reservoir mesh in reservoir numerical model The potential energy scalar field at that location, For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; For reservoir mesh in reservoir numerical model The potential energy scalar field at that location, For reservoir mesh in reservoir numerical model Potential energy scalar field at the location; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction; reservoir grid and reservoir grid With reservoir grid exist Adjacent in direction.

5. The method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 5, the formula for calculating the connectivity coefficient is: ； In the formula, For the first The connectivity coefficient of a reservoir grid; Streamline number; This represents the total number of streamlines. For the first The volumetric flow rate carried by each streamline; For streamline indicator functions, when the first... The streamline passes through the first When a reservoir grid is used, The value is 1, otherwise, The value is 0.

6. The method for staged utilization of residual oil based on the combined driving force of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 6, firstly, the potential energy gradient data and wellbore connectivity coefficient data of the reservoir are homogenized so that the values ​​of all potential energy gradient data and wellbore connectivity coefficient data are within the range of... Then, based on the potential energy gradient and well-fracture connectivity coefficient of each reservoir network in the reservoir numerical model, the reservoir feature vector of each reservoir grid is obtained. Based on the reservoir feature vectors of all reservoir grids in the reservoir numerical model as sample data points, a reservoir feature dataset is established. The reservoir feature dataset is divided into Clustering, resulting in Given cluster centers, the objective function is constructed by minimizing the sum of squared distances between all sample data points and their respective cluster centers. ,get: ； In the formula, Cluster number; The total number of clusters; For the first Each reservoir feature vector ,in, For the first Potential energy gradient of each reservoir grid, For the first The connectivity coefficient of a reservoir grid. It is the transpose matrix; For the first A cluster set; For the first The coordinates of each cluster center along the potential gradient dimension; For the first The coordinates of each cluster center along the connectivity coefficient dimension; The objective function is optimized based on the K-means clustering algorithm. In each optimization process, the cluster centers of each cluster are first determined. Then, all reservoir grids in the reservoir numerical model are traversed, and each reservoir grid is assigned to the cluster set whose cluster center has the smallest squared Euclidean distance. Each temporary cluster is used to obtain the centroid of each temporary cluster and use it as the cluster center. When the cluster center of each temporary cluster meets the preset optimization termination criterion, the optimization of the objective function is stopped, and the classification results of each reservoir grid in the reservoir numerical model are obtained.

7. The method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity as described in claim 6, characterized in that, The optimization termination criterion is set based on two principles: local optimal solution and engineering protection. When the cluster centers of each temporary cluster no longer change or the change is less than the preset minimum tolerance threshold after two consecutive optimizations, or when the current number of optimizations reaches the preset maximum number of optimizations, the preset optimization termination criterion is met, the clustering and classification of all sample data points in the reservoir feature dataset is completed, and the optimization of the objective function is stopped.

8. The method for staged utilization of residual oil based on the combined drive of potential energy gradient field and wellbore connectivity as described in claim 1, characterized in that, In step 7, the internal area of ​​the reservoir is divided into primary, secondary, tertiary, and quaternary exploitation zones based on the remaining oil exploitation potential value. Specifically, when the remaining oil exploitation potential value of the reservoir area is greater than 0.8 but not more than 1, the reservoir area is determined to be a primary exploitation zone; when the remaining oil exploitation potential value of the reservoir area is greater than 0.6 but not more than 0.8, the reservoir area is determined to be a secondary exploitation zone; when the remaining oil exploitation potential value of the reservoir area is greater than 0.4 but not more than 0.6, the reservoir area is determined to be a tertiary exploitation zone; and when the remaining oil exploitation potential value of the reservoir area is greater than 0 but not more than 0.4, the reservoir area is determined to be a quaternary exploitation zone.

Images