Reservoir production optimization method considering complex fracture network

Abstract

The application provides a reservoir production optimization method considering a complex fracture network, which comprises the following steps: step 1, dividing the reservoir into two regions according to the development scale of fractures in the reservoir; step 2, respectively establishing a single pipe model and a double medium model for the divided regions, and solving the models by dividing grids on the pipe; step 3, performing history matching on the observed values and the simulated values by means of a set multi-smooth data assimilation algorithm to obtain an optimized model; and step 4, using the optimized model to optimize the injection and production capacities of each well with the economic net present value of the reservoir as an objective function. The reservoir production optimization method considering the complex fracture network performs well in improving the oil recovery rate and achieves remarkable optimization effect. The method can effectively deal with the complex fracture network in the reservoir, improve the production efficiency and maximize the economic benefits of the oilfield.

Description

Technical Field

[0001] This invention relates to the field of numerical simulation technology for fractured reservoirs, and in particular to a reservoir production optimization method that considers complex fracture networks. Background Technology

[0002] Most oil reservoirs contain fractures, such as those in tight limestone and mudstone. During reservoir development, hydraulic fracturing is sometimes used to artificially create fractures, such as in shale reservoirs, to provide more flow channels. The presence of fractures alters the flow characteristics of fluids; therefore, the influence of fractures must be fully considered when describing fluid flow in oil reservoirs. Furthermore, fractures exist in various forms, ranging from large-scale fractures spanning hundreds or thousands of meters to small-scale fractures with shorter extensions but wider distribution. Effectively characterizing the fracture system is the primary challenge in numerical simulation of fractured reservoirs.

[0003] Chinese patent application CN201310138269.1 discloses a numerical simulation method for large-scale fractured reservoirs, belonging to the field of oil and gas field development. The method uses large-scale fractures in the reservoir as the inner boundary for dimensionality reduction, establishing a reservoir geometric model. Then, a triangular mesh is used to geometrically partition the reservoir geometric model into discrete units, simultaneously generating reservoir geometric information description files, geometric discrete information files, and well information description files. Numerical calculation units are constructed based on these discrete units, establishing a mathematical model characterizing the fluid flow in such reservoirs. An integral method is used to establish the numerical calculation format for each numerical calculation unit, forming a large-scale system of linear algebraic equations, which are then solved to achieve numerical simulation of large-scale fractured reservoirs. This invention can accurately characterize large-scale fractured reservoirs, has clear physical meaning, and is highly adaptable to numerical simulation of large-scale fractured reservoirs.

[0004] Chinese patent application CN202010005164.9 discloses a numerical simulation method and system for tight oil reservoirs based on matrix-fracture unsteady crossflow. The model assumes that hydraulic fractures have the highest conductivity, followed by natural fractures, and then the matrix. The model uses discrete fractures to represent hydraulic fractures and a dual-medium approach to represent the matrix and natural fractures. During mesh generation, a denser mesh is used at the hydraulic fracture sites, explicitly representing the hydraulic fractures, while the natural fractures and matrix share a single mesh. After obtaining the discrete equations, a closed equation system is obtained by adding well constraints, auxiliary equations, and boundary conditions. The pressure and saturation at each time step are calculated using Newton's iteration method. This invention improves the simulation efficiency of reservoirs with large numbers of fractures; by utilizing the continuous medium assumption to handle the flow exchange between the matrix and natural fractures, the simulation program can accurately characterize the unsteady crossflow problem between the matrix and fractures.

[0005] Chinese patent application CN202210642759.4 discloses a reservoir optimization method based on an approximate gradient algorithm and an embedded discrete fracture model, comprising the following steps: S1, establishing a numerical simulation model of a fractured reservoir based on reservoir numerical simulation software and an embedded discrete fracture model; S2, setting economic and algorithm parameters and initializing the well control vector; S3, obtaining the set of disturbances for the well control vector and running the reservoir numerical simulation software to obtain the net present value corresponding to each disturbance vector; S4, calculating the approximate gradient and updating the well control vector; S5, finding the optimal well control vector using a bisection method and repeating steps S3, S4, and S5 until the iteration termination condition is met. This invention establishes a numerical simulation model of a fractured reservoir that reasonably characterizes the seepage features of fractures in the reservoir using an embedded discrete fracture modeling method, and conducts research on the optimization of reservoir injection and production parameters for a conceptual model with one injection and one production cycle and an oilfield-scale model with multiple injection and multiple production cycles.

[0006] The methods mentioned above generally suffer from high computational cost and low optimization efficiency, failing to solve the technical problem we wanted to address. Therefore, we have invented a new reservoir production optimization method that considers complex fracture networks. Summary of the Invention

[0007] The purpose of this invention is to provide a reservoir production optimization method that takes into account complex fracture networks in oil reservoirs, which can effectively address the complex fracture networks present in oil reservoirs, improve production efficiency, and maximize the economic benefits of oil fields.

[0008] The objective of this invention can be achieved through the following technical measures: a reservoir production optimization method considering complex fracture networks, which includes:

[0009] Step 1: Divide the reservoir into two regions according to the scale of fracture development in the reservoir;

[0010] Step 2: Establish single-pipe model and dual-medium model for the divided regions respectively, and solve the model by dividing the pipeline into meshes;

[0011] Step 3: Use the ensemble multi-smoothing data assimilation algorithm to perform historical fitting between the observed values ​​and the simulated values ​​to obtain the optimized model;

[0012] Step 4: Using the optimized model, with the reservoir economic net present value as the objective function, optimize the injection and production volume of each well.

[0013] The objective of this invention can also be achieved through the following technical measures:

[0014] In step 1, the reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing in the reservoir, and the reservoir is divided into two regions according to the development scale of the fractures in the reservoir.

[0015] In step 1, the reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing. The fractures are described by arranging fracture nodes at their endpoints and each inflection point. The purpose is to distinguish the fracture nodes from ordinary well points. The fracture nodes are connected according to the fracture trend, and then the fracture nodes are connected to the well points.

[0016] In step 1, the reservoir is divided into two regions according to the scale of fracture development: a region with large-scale fracture distribution and a region with small-scale fracture distribution.

[0017] In step 1, large-scale fractures refer to fractures with a length greater than one-third of the average well spacing in the reservoir, while small-scale fractures refer to fractures with a length less than one-third of the average well spacing in the reservoir. The two areas may overlap.

[0018] In step 2, a single-pipe model and a dual-medium model are constructed for the divided region, and a mesh is generated on the pipe. Based on the physical property parameters of the nodes, the parameters on the mesh are solved implicitly.

[0019] In step 2, a single-pipeline model is formed by connecting the discretized well points and fracture nodes from step 1 into a connected network composed of connected pipes. The flow equations of the matrix system are used for the pipes between wells and between wells and fracture nodes, and the flow equations of the large fracture are used for the pipes between fractures. A grid is divided on each pipe to form a single-pipeline network model.

[0020] In step 2, a dual-medium model is used, where each connected pipe is considered as two flow systems: a bedrock system and a fracture system. The bedrock system is primarily responsible for storage, while the fracture system is primarily responsible for seepage. There is fluid exchange between the two. Therefore, in the small-scale fracture region where the dual-medium model is constructed, all pipes are represented by two layers of mesh, with each mesh characterized by its own parameters.

[0021] In step 2, a grid is divided on the pipeline. The grids on the same connection unit have equal seepage cross-sectional area and permeability. When the connection unit is between well points, the permeability is the average value of the two well points. When the connection unit is between a well point and a fracture node, the permeability is the permeability within the control volume of the well point.

[0022] In step 2, the physical properties are the permeability, porosity, thickness, and area of ​​the mesh, and the mesh parameters that need to be solved are physical quantities such as pressure and saturation.

[0023] In step 3, in order to reduce the difference between the established model and the reservoir, the observed values ​​and simulated values ​​are historically fitted using an ensemble multi-smoothing data assimilation algorithm to obtain an optimized model.

[0024] In step 3, the parameters that need to be fitted include the well-to-well connectivity volume V in the single-pipe model. ij,M and the volume V between the cracks ij,F And the connected volume V of the matrix and fractures in the dual-medium model. ij,m V ij,f .

[0025] In step 4, the reservoir's economic net present value is used as a performance index to establish an objective function for reservoir injection and production optimization. Combined with relevant constraints, the reservoir's production optimization problem is transformed into a problem of solving a mathematical control model, thereby optimizing the injection and production volume of each well.

[0026] In step 4, after the established model has been historically fitted and simulated for a sufficiently long time, the updated model with historical fit is used as the forward model. The production optimization time is divided into Nc consecutive time intervals of equal length, and each time interval is called a control step.

[0027] In step 4, the control step, it is necessary to specify the operating conditions for each well. In a specific control step, each well is operated under constant operating conditions.

[0028] In step 4, the operating conditions, i.e. the control variables, are the total rate or phase rate of each well. Here, the total rate is used for production optimization.

[0029] This invention presents a reservoir production optimization method considering complex fracture networks. This method offers moderate computational complexity and high accuracy. It constructs a one-dimensional dual-medium model by discretizing the reservoir system with a one-dimensional grid between wells, enabling numerical simulation and optimization of fractured reservoirs. The proposed method demonstrates excellent performance in improving oil recovery, achieving significant optimization results. This method effectively addresses complex fracture networks in reservoirs, enhancing production efficiency and maximizing the economic benefits of oilfields. Attached Figure Description

[0030] Figure 1 This is a flowchart of a specific embodiment of the reservoir production optimization method considering complex fracture networks of the present invention;

[0031] Figure 2 This is a crack distribution diagram in a specific embodiment of the present invention;

[0032] Figure 3This is a single-pipe connection diagram of a single-pipe model in a specific embodiment of the present invention;

[0033] Figure 4 This is a diagram of a single-pipe connection using a dual-medium model in a specific embodiment of the present invention;

[0034] Figure 5 This is a diagram showing the distribution of reservoir well points and fractures in a specific embodiment of the present invention;

[0035] Figure 6 This is a pipe connection diagram of two models in a specific embodiment of the present invention;

[0036] Figure 7 This is a comparison chart of historical fitting results in a specific embodiment of the present invention;

[0037] Figure 8 This is a production optimization result diagram with an optimization time step of 180 days in a specific embodiment of the present invention;

[0038] Figure 9 This is a production optimization result diagram with an optimization time step of 90 days in a specific embodiment of the present invention;

[0039] Figure 10 This is a production optimization result diagram with an optimization time step of 60 days in a specific embodiment of the present invention. Detailed Implementation

[0040] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0041] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, and / or combinations thereof.

[0042] like Figure 1 As shown, Figure 1This is a flowchart of the reservoir production optimization method considering complex fracture networks according to the present invention. The reservoir production optimization method considering complex fracture networks includes the following steps: The reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing. Fluid flow between wells is divided into two categories: simple flow in fracture-free regions and complex flow in regions with primary fractures. A single-channel model and a dual-medium model are established respectively, and the models are solved by dividing the channels into grids. Then, a multi-smoothing data assimilation algorithm is used for history fitting, and the fitted model is used to establish production optimization with the reservoir's economic net present value as the objective.

[0043] The reservoir production optimization method of the present invention considering complex fracture networks includes the following steps:

[0044] Step 1: Discretize the reservoir into a series of interconnected fracture nodes based on the original large-scale fractures or high-conductivity fractures generated by hydraulic fracturing in the reservoir, and divide the reservoir into two regions according to the development scale of the fractures in the reservoir.

[0045] Step 2: Construct a single-pipe model and a dual-medium model for the divided region, and divide the pipeline into meshes. Based on the physical property parameters of the nodes, perform a fully implicit solution for the parameters on the mesh.

[0046] Step 3: In order to reduce the difference between the established model and the reservoir, the observed values ​​and simulated values ​​are historically fitted using an ensemble multi-smoothing data assimilation algorithm to obtain an optimized model.

[0047] Step 4: Establish the reservoir injection and production optimization objective function with the reservoir economic net present value as the performance index, and combine it with relevant constraints to transform the reservoir production optimization problem into a mathematical control model solution problem, thereby optimizing the injection and production of each well.

[0048] In step 1, the reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing. The fractures are described by arranging fracture nodes at their endpoints and each inflection point. The purpose is to distinguish the fracture nodes from ordinary well points. The fracture nodes are connected according to the fracture trend, and then the fracture nodes are connected to the well points.

[0049] In step 1, the reservoir is divided into two regions based on the scale of fracture development: regions with large-scale fractures and regions with small-scale fractures. Figure 2 As shown.

[0050] Large-scale fractures refer to fractures whose length is greater than one-third of the average well spacing in the reservoir.

[0051] The term "small-scale fracture" refers to a fracture whose length is less than one-third of the average well spacing in the reservoir.

[0052] The two regions may overlap.

[0053] The single-pipe model in step 2 connects the discretized well points and fracture nodes from step 1 into a connected network of pipes. The pipes between wells and between wells and fracture nodes use the flow equations of the matrix system, while the pipes between fractures use the flow equations of large fractures. A mesh is then created on each pipe, as shown below. Figure 3 As shown, a single-pipe network model is formed.

[0054] In step 2, the dual-medium model treats each connected pipe as two flow systems: a bedrock system and a fracture system. The bedrock system primarily functions as a storage system, while the fracture system primarily functions as a seepage system, with fluid exchange between them. Therefore, in the small-scale fracture region of the dual-medium model, all pipes are represented by two layers of mesh, each mesh characterized by its own parameters, such as... Figure 4 As shown.

[0055] In step 2, a grid is divided on the pipeline. The grids on the same connection unit have equal seepage cross-sectional area and permeability. When the connection unit is between well points, the permeability is the average value of the two well points. When the connection unit is between a well point and a fracture node, the permeability is the permeability within the control volume of the well point.

[0056] In step 2, the physical properties include the permeability, porosity, thickness, and area of ​​the mesh.

[0057] The mesh parameters that need to be solved are physical quantities such as pressure and saturation.

[0058] The parameters to be fitted in step 3 include the interconnected volume V between wells in the single-pipe model. ij,M and the volume V between the cracks ij,F And the connected volume V of the matrix and fractures in the dual-medium model. ij,m V ij,f .

[0059] In step 4, production optimization is performed after our established model has been historically fitted and simulated for a sufficiently long time. The updated model is then used as the forward model, and the production optimization time is divided into Nc consecutive time intervals of equal length. Each time interval is called a control step.

[0060] The control steps require specifying the operating conditions for each well. In a specific control step, each well operates under constant operating conditions.

[0061] The operating conditions (control variables) are the total rate or phase rate of each well. Here, we use the total rate for production optimization.

[0062] This invention aims to effectively characterize large-scale and small-scale fractures in oil reservoirs. Conventional grid-based numerical simulation methods for oil reservoirs usually require dividing a large number of grids to finely characterize and describe the fractures, which often takes a considerable amount of time. This invention only considers the flow between wells, between wells and fractures, and between fractures, simplifying the oil reservoir and significantly improving computational efficiency while retaining a certain level of computational accuracy.

[0063] In a specific embodiment 1 of the present invention, the reservoir production optimization method considering complex fracture networks includes the following steps:

[0064] First, the reservoir is divided into regions with large-scale fractures and regions with small-scale fractures according to the development scale of fractures in the reservoir. Then, a single-pipeline network model is established in the region with large-scale fractures, and a double-pipeline network model is established in the region with small-scale fractures. Then, in the region with large-scale fractures, the reservoir is discretized into a series of interconnected fracture nodes according to the structural characteristics of the fractures. The fractures are described by arranging fracture nodes at their endpoints and each inflection point. The fracture nodes are connected according to the trend of the fractures. Then, the fracture nodes are connected to the well points to form a connection network.

[0065] The steps for constructing a single-layer pipe network model in a large-scale crack distribution region are as follows:

[0066] The connections between well nodes form a connected network. The connected volume of each connecting pipe, taking node i and node j as an example, is calculated as follows:

[0067]

[0068]

[0069] Where i and j are nodes; V ij,M and V ij,F These represent the connected volumes between wells or between fractures, respectively, in meters. 3 S ij Let m be the area between two points. 2 ; The average thickness between the two points, in meters (m); and 1 represents the average porosity between wells or between fractures.

[0070] The formula for calculating the cross-sectional area of ​​seepage is as follows:

[0071]

[0072] The mass balance equations between wellpoints or between a wellpoint and a fracture node in the single-pipeline network model, as well as the mass balance equations between fracture nodes, are as follows:

[0073]

[0074]

[0075] Where β = o, w, o, w represent the oil phase and water phase, respectively; M represents the bedrock, F represents the fracture, and in areas other than those with small-scale fractures, only bedrock flow is considered between wells; ρ M,β and ρ F,β The densities of a certain phase fluid in bedrock and fractures are shown in kg / m³. 3 ;φ M and φ F Porosity of bedrock and fracture, respectively; 1; S M,β and S F,β , representing the saturation of the phase fluid in the bedrock and fractures, respectively, 1; t represents time, s; v M,β and v F,β These represent the seepage velocities of a certain phase fluid in bedrock and fractures, respectively, in meters. 3 / s;q M,β For source and sink terms (q>0 is a source term, i.e., injection is positive), m represents the increase in the amount of a certain phase fluid per unit volume per unit time. 3 / s.

[0076] The equations of motion are expressed as follows:

[0077]

[0078] Where β = o, w, o, w represent the oil phase and the water phase, respectively; v is the seepage velocity, m 3 / s; K is permeability, mD; μ is fluid viscosity, mPa·s; The pressure difference is the flow pressure, expressed in MPa.

[0079] The steps for constructing a dual-medium model in a small-scale crack distribution region are as follows:

[0080] The formula for calculating the connected volume of a connected element is as follows:

[0081]

[0082]

[0083] Where i and j are well nodes; V ij,m and V ij,f Let m represent the connected volumes of the matrix and the fracture between nodes i and j, respectively.3 S ij Let m be the partition area between node i and node j. 2 ; Let m be the average thickness between nodes i and j; and is the average porosity of the matrix and cracks between nodes i and j, 1;

[0084] The flow rate between bedrock and fractures depends primarily on the fluid viscosity, the pressure difference between the bedrock and fractures, and certain characteristic properties of the rock (length, area, volume, etc.). Therefore, the flow rate between fractures and bedrock can be expressed as:

[0085]

[0086] Where Q is the flow rate between the fracture and the bedrock, m 3 / s; α is the shape factor, 1 / m 2 ;ρ m,β The density of a certain phase fluid within bedrock, kg / m³ 3 ;K m Let mD be the permeability of a certain phase fluid within bedrock; μ be the permeability of that fluid. β Let P be the viscosity of a certain phase fluid, in mPa·s; m and P f The pressures are in MPa for the matrix and the fracture, respectively.

[0087] The mass balance equations for the matrix and cracks are as follows:

[0088]

[0089]

[0090] Where, ρ m,β and ρ f,β The densities of a certain phase fluid in bedrock and fractures are shown in kg / m³. 3 ;φ m and φ f Porosity of bedrock and fracture, respectively; 1; S m,β and S f,β , representing the saturation of the phase fluid in the bedrock and fractures, respectively, 1; t represents time, s; v m,β and v f,β Let m represent the seepage velocity of a certain phase fluid in the bedrock and fracture, respectively. 3 / s;q m and q f These are the source and sink terms for the matrix and fracture, respectively (q>0 is the source term, i.e., injection is positive, representing the increase in mass per unit volume per unit time), m 3 / s.

[0091] Then, the well-to-well connectivity volume V in the single-pipe model is calculated. ij,M and the volume V between the cracks ij,F And the connected volume V of the matrix and fractures in the dual-medium model. ij,m V ij,f Perform fitting.

[0092] The historical fitting objective function established to reduce the error between the constructed model and the actual model is as follows:

[0093]

[0094] Where m represents the feature parameters for historical fitting of each grid, O(m) is the established minimization objective function, g(m) is the production dynamic index obtained from the simulation of the pipeline model, and d obs C represents the observed production dynamics indicators. D Let be the main diagonal matrix of the covariance of observation errors.

[0095] The ensemble multi-smoothing data assimilation algorithm updates the reservoir model parameters using an update equation after each data assimilation step. However, the update equation does not guarantee that each parameter satisfies boundary constraints or linear equality constraints. Therefore, we add constraints to enforce these constraints on the parameters.

[0096] The constraints are as follows:

[0097]

[0098] Among them, V tot,m and V tot,f N represents the total volume of the matrix and the fracture, respectively. w For well number, V ij,M V ij,F and V ij,m V ij,f These are the volumes of matrix and fractures within the connected units between nodes i and j in the single-channel network model and the dual-medium model, respectively. The purpose of these constraints is to ensure that the sum of the volumes of the matrix within the reservoir equals the total volume of the matrix, and the sum of the volumes of the fractures equals the total volume of the fractures.

[0099] Finally, the objective function for reservoir injection and production optimization, established using net present value (NPV) as the indicator, is as follows:

[0100]

[0101] Where, r o The price per unit of oil, CNY / m 3 ;r w The unit price for processed water is CNY / m³. 3 ;r wi The unit price of injected water, CNY / m 3b is the annual discount rate; N t It is the total number of simulation time steps; t n It is the nth time step; Δt n It is the time step size of the nth time step; P and I correspond to the number of production wells and the number of injection wells, respectively; and Let m be the oil production and water production of the j-th production well at time step n. 3 / day; Let m be the volume of fluid injected into the j-th injection well at time step n. 3 / day;

[0102] The control variables are represented as follows:

[0103]

[0104] Where N w Indicates the number of wells, u i Let N represent the phase velocity of the i-th well, and there will be different control variables at each time step, so the dimension of the total control variables is: N u =N c ·N w , where N u N represents the dimension of the total control variables. c This represents the number of time steps.

[0105] Based on the actual conditions of the oilfield, the control variables are constrained as follows:

[0106]

[0107] in, and These represent the lower and upper boundary conditions of the well-control conditions at the j-th time step, respectively.

[0108] Conceptual Example 1:

[0109] The constructed reservoir model is as follows Figure 5 As shown, the reservoir is divided into two regions. The left region contains two large-scale fractures, while the right region contains irregular small fractures. The constructed reservoir model has a grid size of 42*42*1 and a grid size of 10m*10m*5m. There are a total of fourteen wells in the reservoir, including four water injection wells and ten production wells. ECLIPSE was used to simulate production of the reservoir block for 4920 days, and the data from the first 3120 days were used as real data to perform historical fitting on the constructed model.

[0110] First, complete the connection diagrams for the single-pipe model and the dual-medium model, such as... Figure 6As shown, circles represent well points and fracture nodes, black pipes represent large-scale fractures, white pipes represent the fracture system in the dual-medium model, and gray pipes represent the matrix system. The left area is the single-channel network model region, and the right area is the dual-medium model region. Using the method described in this paper, a single forward simulation takes 15.2 seconds, significantly less than the 53.4 seconds taken by a conventional reservoir numerical simulator. Figure 7 The image shows the block's cumulative oil production fitting results after historical fitting of single-well oil production rates over the first 3120 days. Black asterisks represent historical data, and black solid lines represent the simulated cumulative oil production results of this invention. The fitting effect is shown to be good. This embodiment uses a 180-day time step, optimizing each well for 10 time steps, totaling five years of production optimization. This significantly improves the cumulative oil production obtained by continuing production under the original production regime. The comparison results of cumulative oil production before and after optimization are shown below. Figure 8 As shown in the figure, the solid black line represents the cumulative oil production after optimization, the dashed black line represents the cumulative oil production under the previous system, and the vertical dashed line represents the dividing line between historical fitting and production optimization.

[0111] Conceptual Example 2:

[0112] This embodiment uses the same reservoir parameters and injection-production regime as Conceptual Example 1, and employs the same historical data fitting results for production optimization. The difference between this embodiment and Conceptual Example 1 is that this embodiment uses a 90-day optimization step size, optimizing each well for 20 time steps, with a total optimization period of five years. The comparison of cumulative oil production before and after optimization is as follows: Figure 9 As shown in the figure, the solid black line represents the cumulative oil production after optimization, the dashed black line represents the cumulative oil production under the previous system, and the vertical dashed line represents the dividing line between historical fitting and production optimization.

[0113] Conceptual Example 3:

[0114] This embodiment uses the same reservoir parameters and injection-production regime as Conceptual Example 1, and employs the same historical data fitting results for production optimization. The difference between this embodiment and Conceptual Example 1 is that this embodiment uses a 60-day optimization step size, optimizing each well for 30 time steps, with a total optimization period of five years. The comparison of cumulative oil production before and after optimization is as follows: Figure 10 As shown in the figure, the solid black line represents the cumulative oil production after optimization, the dashed black line represents the cumulative oil production under the previous system, and the vertical dashed line represents the dividing line between historical fitting and production optimization.

[0115] The above three conceptual examples were designed to test the impact of different optimization strategies on the optimization results. The results show that, under the same optimization cycle, optimization strategies with smaller optimization steps and higher optimization frequency will produce better optimization results.

[0116] Finally, it should be noted that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0117] Except for the technical features described in the specification, all other technologies are known to those skilled in the art.

Claims

1. A reservoir production optimization method considering complex fracture networks, characterized in that, The reservoir production optimization method that considers complex fracture networks includes: Step 1: Divide the reservoir into two regions according to the scale of fracture development in the reservoir; Step 2: Establish single-pipe model and dual-medium model for the divided regions respectively, and solve the model by dividing the pipeline into meshes; Step 3: Use the ensemble multi-smoothing data assimilation algorithm to perform historical fitting between the observed values ​​and the simulated values ​​to obtain the optimized model; Step 4: Using the optimized model, with the reservoir economic net present value as the objective function, optimize the injection and production volume of each well; In step 1, the reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing in the reservoir, and the reservoir is divided into two regions according to the development scale of the fractures in the reservoir. In step 1, the reservoir is divided into two regions according to the scale of fracture development: a region with large-scale fracture distribution and a region with small-scale fracture distribution. In step 2, a single-pipe model is formed by connecting the discretized well points and fracture nodes from step 1 into a connected network composed of connected pipes. The pipes between wells and between wells and fracture nodes adopt the flow equation of the matrix system, while the pipes between fractures adopt the flow equation of large-scale fractures. A grid is divided on each pipe to form a single-pipe model.

2. The reservoir production optimization method considering complex fracture networks according to claim 1, characterized in that, In step 1, the reservoir is discretized into a series of interconnected fracture nodes based on the large-scale primary fractures or high-conductivity fractures generated by hydraulic fracturing. The fractures are described by arranging fracture nodes at their endpoints and each inflection point. The purpose is to distinguish the fracture nodes from ordinary well points. The fracture nodes are connected according to the fracture trend, and then the fracture nodes are connected to the well points.

3. The reservoir production optimization method considering complex fracture networks according to claim 2, characterized in that, In step 1, large-scale fractures refer to fractures with a length greater than one-third of the average well spacing in the reservoir, while small-scale fractures refer to fractures with a length less than one-third of the average well spacing in the reservoir. The two areas may overlap.

4. The reservoir production optimization method considering complex fracture networks according to claim 1, characterized in that, In step 2, a single-pipe model and a dual-medium model are constructed for the divided region, and a mesh is generated on the pipe. Based on the physical property parameters of the nodes, the parameters on the mesh are solved implicitly.

5. The reservoir production optimization method considering complex fracture networks according to claim 3, characterized in that, In step 2, a dual-medium model is used, where each connected pipe is considered as two flow systems: a bedrock system and a fracture system. The bedrock system is primarily responsible for storage, while the fracture system is primarily responsible for seepage. There is fluid exchange between the two. Therefore, in the small-scale fracture region where the dual-medium model is constructed, all pipes are represented by two layers of mesh, with each mesh characterized by its own parameters.

6. The reservoir production optimization method considering complex fracture networks according to claim 3, characterized in that, In step 2, a grid is divided on the pipeline. The grids on the same connection unit have equal seepage cross-sectional area and permeability. When the connection unit is between well points, the permeability is the average value of the two well points. When the connection unit is between a well point and a fracture node, the permeability is the permeability within the control volume of the well point.

7. The reservoir production optimization method considering complex fracture networks according to claim 3, characterized in that, In step 2, the physical properties are the permeability, porosity, thickness, and area of ​​the mesh, and the mesh parameters that need to be solved are physical quantities such as pressure and saturation.

8. The reservoir production optimization method considering complex fracture networks according to claim 1, characterized in that, In step 3, in order to reduce the difference between the established model and the reservoir, the observed values ​​and simulated values ​​are historically fitted using an ensemble multi-smoothing data assimilation algorithm to obtain an optimized model.

9. The reservoir production optimization method considering complex fracture networks according to claim 8, characterized in that, In step 3, the parameters that need to be fitted include the well-to-well connectivity volume V in the single-pipe model. ij,M and the volume V between the cracks ij,F And the connected volume V of the matrix and fractures in the dual-medium model. ij,m V ij,f .

10. The reservoir production optimization method considering complex fracture networks according to claim 1, characterized in that, In step 4, the reservoir's economic net present value is used as a performance index to establish an objective function for reservoir injection and production optimization. Combined with relevant constraints, the reservoir's production optimization problem is transformed into a problem of solving a mathematical control model, thereby optimizing the injection and production volume of each well.

11. The reservoir production optimization method considering complex fracture networks according to claim 10, characterized in that, In step 4, after the established model has been historically fitted and simulated for a sufficiently long time, the updated model with historical fit is used as the forward model. The production optimization time is divided into Nc consecutive time intervals of equal length, and each time interval is called a control step.

12. The reservoir production optimization method considering complex fracture networks according to claim 11, characterized in that, In step 4, the control step, it is necessary to specify the operating conditions for each well. In a specific control step, each well is operated under constant operating conditions.

13. The reservoir production optimization method considering complex fracture networks according to claim 12, characterized in that, In step 4, the operating conditions, i.e. the control variables, are the total rate or phase rate of each well. Here, the total rate is used for production optimization.

Images