Biological source composite system plug removal and oil displacement integrated numerical simulation method

Abstract

The invention provides a plug removal and oil displacement integrated numerical simulation method for a biological source compound system, and belongs to the technical field of biological source compound systems. A plug removal kinetic model and an oil displacement synergistic model are established, and a plug removal and oil displacement coupling numerical model is constructed; the method comprises the following steps: establishing a reservoir permeability dynamic evolution prediction model, introducing a blockage dissolution fraction as an intermediate variable, solving a multi-physics field coupling equation set by adopting an operator splitting technology and a self-adaptive time step control strategy, establishing a geological conceptual model to simulate the whole injection process of a blocking remover and an oil-displacing agent, establishing a ratio optimization double-layer game model for collaborative optimization, and calculating the reservoir permeability dynamic evolution prediction model. The oil increasing quantity index is output, and the proportion of the blocking remover and the surface active oil-displacing agent in the biological source composite system is optimized, so that the technical problem that the oil increasing effect is poor due to the fact that the proportion of the blocking remover and the oil-displacing agent is difficult to optimize accurately in the stratum blocking removal and oil displacement process of the biological source composite system is solved.

Description

Technical Field

[0001] This invention belongs to the technical field of bio-based compound systems, and more specifically, it relates to a numerical simulation method for integrated deblocking and oil displacement of bio-based compound systems. Background Technology

[0002] In oilfield development, formation blockage is a key factor restricting production improvement. Traditional unblocking technologies mainly employ chemical solvents or bio-based composite systems to remove formation blockages. Bio-based composite systems dissolve blockages with unblocking agents and combine them with surfactants to improve oil recovery, offering advantages such as environmental friendliness and high efficiency, and have been applied in multiple oilfields. However, in existing technologies, the ratio of unblocking agents to surfactants in bio-based composite systems lacks quantitative optimization methods. Field applications mainly rely on experience to determine the ratio parameters, making it difficult to accurately describe the synergistic mechanism between unblocking and oil displacement processes. This leads to situations where the unblocking effect and oil displacement efficiency cannot be simultaneously optimized when the ratio is unreasonable. In other words, existing technologies suffer from the technical problem of inaccurately optimizing the ratio of unblocking agents to surfactants in the formation unblocking and oil displacement process, resulting in poor oil recovery. Summary of the Invention

[0003] In view of this, the present invention provides a numerical simulation method for the integrated unblocking and oil displacement of a bio-based composite system, which can solve the technical problem in the prior art where the ratio of unblocking agent to oil displacement agent is difficult to optimize accurately during the formation unblocking and oil displacement process of a bio-based composite system, resulting in poor oil enhancement effect.

[0004] This invention is implemented as follows: It provides a numerical simulation method for integrated unblocking and oil displacement in a bio-based composite system. This method constructs a database of the physicochemical properties of the bio-based composite system and a random field model of the spatial distribution of blockages. It establishes a unblocking kinetic model and an oil displacement efficiency enhancement model, constructs a coupled numerical model for unblocking and oil displacement, and establishes a dynamic evolution prediction model for reservoir permeability. It introduces the dissolution fraction of blockages as an intermediate variable, employs operator splitting technology and an adaptive time step control strategy to solve the multi-physics coupled equations, establishes a geological conceptual model to simulate the entire process of injecting unblocking agents and oil displacement agents, and establishes a two-layer game model for ratio optimization. The upper-layer model, with unblocking effect as the objective, and the lower-layer model, with oil displacement efficiency as the objective, are used for collaborative optimization. The method outputs an oil increase index and optimizes the ratio of unblocking agents and surfactants in the bio-based composite system.

[0005] Specifically, the step of constructing the physicochemical property database of the bio-derived composite system involves setting the molecular weight of each component based on indoor experiments, obtaining data on the change of oil-water interfacial tension with the concentration of surfactant oil displacement agent based on indoor experiments, and fitting the data on the change of oil-water interfacial tension with the concentration of surfactant oil displacement agent into a power-law decay model mathematical function.

[0006] Specifically, the step of establishing a random field model for the spatial distribution of blockages involves using random field theory combined with geostatistical methods, and generating multiple equally probable heterogeneous distributions of blockages using Monte Carlo simulation technology. The uncertainty of the spatial distribution of blockages is described by a probability density function, and the spatial correlation of blockages is described by a variogram function.

[0007] The probability density function adopts a log-normal distribution, and the variogram adopts a spherical model.

[0008] Specifically, the step of establishing the deblocking and oil displacement coupled numerical model involves establishing a deblocking kinetic model, calculating the reaction kinetic equation using data processing software based on data obtained from indoor experiments, selecting a concentration-based power-law model to describe the reaction rate, and establishing a deblocking agent transport equation to describe the convection diffusion and chemical reaction process of the deblocking agent in the porous medium.

[0009] Specifically, the establishment of the oil displacement enhancement model involves determining the relative permeability curves under the action of the surfactant oil displacement agent based on indoor experiments. The relative permeability curves include the relative permeability curves of the aqueous phase and the relative permeability curves of the oil phase.

[0010] Specifically, the step of establishing a dynamic evolution prediction model for reservoir permeability involves establishing the relationship between porosity and permeability based on the Carman-Kozeny equation, and introducing the dissolution fraction of plugging material as an intermediate variable, whereby the dissolution fraction of plugging material is the ratio of the mass of dissolved plugging material to the total mass of the initial plugging material.

[0011] The reservoir permeability dynamic evolution prediction model is trained using a neural network algorithm. The neural network algorithm includes two hidden layers in the input layer and an output layer. The input parameters include the initial permeability and porosity of the blockage dissolution fraction, and the output parameter is the current permeability. The neural network algorithm is trained using the backpropagation algorithm.

[0012] Specifically, the step of using operator splitting technology and adaptive time step control strategy to solve the multiphysics coupling equations involves using operator splitting technology to decompose the multiphysics coupling problem into a fluid flow subproblem, a chemical reaction subproblem, and a concentration diffusion subproblem, and solving each subproblem sequentially in each time step.

[0013] Among them, an implicit iterative scheme is introduced to improve the stability of time integration. The implicit iterative scheme adopts the Crank-Nicolson scheme and sets an adaptive time step control strategy. The relative error between the solutions of two adjacent time steps is calculated. When the relative error is greater than 0.01, the time step is reduced to 0.5 times the original, and when the relative error is less than 0.001, the time step is increased to 1.2 times the original.

[0014] Among them, the Newton-Raphson iterative method is used to accelerate the solution of the nonlinear equation system, and the iterative convergence criterion is set as the residual being less than 1 / 3. The maximum number of iterations is 50. When the iteration fails to converge, the time step is automatically reduced and the calculation is restarted.

[0015] Specifically, the step of establishing a geological concept model involves establishing a geological concept model based on actual formation parameters, importing the molecular weight of each component and the oil-water interfacial tension reduction curve, introducing reaction kinetic equations and relative permeability curves, introducing a reservoir permeability dynamic evolution prediction model, and setting injection and production parameters.

[0016] The geological concept model adopts a one-injection-one-production mode. In the blockage model, the blockage material is injected first to simulate formation blockage, then water is injected, the production well and the injection well are shut down, and finally water is continuously injected. The cumulative oil production is recorded as the oil production in the blockage state. In the unblocking model, the blockage material is injected first to simulate formation blockage, then a bio-based composite system is injected, the production well and the injection well are shut down, and finally water is continuously injected. The cumulative oil production is recorded as the oil production in the unblocking state.

[0017] Specifically, the step of establishing a two-layer game model for ratio optimization involves establishing an upper-layer model with the goal of unblocking effect and a lower-layer model with the goal of oil displacement efficiency. The objective function of the upper-layer model is used to maximize the dissolution rate of the blockage, and the objective function of the lower-layer model is used to maximize the oil displacement efficiency. The two objective functions are correlated with each other through the concentration ratio of the unblocking agent and the surfactant oil displacement agent as a coupling term.

[0018] Among them, a sequential solution strategy is adopted to solve the two-layer game model of ratio optimization. First, the parameters of the lower-layer model are fixed to solve the upper-layer model to obtain the optimal unblocking agent configuration. Then, the parameters of the upper-layer model are fixed to solve the lower-layer model to obtain the optimal surfactant displacement agent configuration. The solution is iterated until both layers of the model reach Nash equilibrium.

[0019] Specifically, the steps of outputting the oil increase index and optimizing the ratio of unblocking agent and surfactant in the bio-based composite system involve determining the initial optimized ratio range based on the equilibrium solution of the bi-layer game model for ratio optimization, designing multiple schemes for the ratio of unblocking agent to surfactant, running a geological concept model to calculate the oil increase for each scheme, and comparing the oil increase results of multiple schemes to output the optimal ratio scheme.

[0020] The beneficial effects of this invention are as follows: This invention establishes a kinetic model for unblocking and an oil displacement efficiency enhancement model, constructs a coupled numerical model for unblocking and oil displacement, and achieves accurate simulation of formation property changes during the unblocking process based on a dynamic evolution prediction model of reservoir permeability. It employs a two-layer game model for ratio optimization, with the upper-layer model targeting unblocking effect and the lower-layer model targeting oil displacement efficiency working together for optimization. The optimal ratio scheme is obtained through iterative calculation using a sequential solution strategy until a Nash equilibrium is reached, thus solving the problem of the lack of quantitative optimization methods for the ratio of bio-derived composite systems.

[0021] This invention accurately describes the chemical reaction process between the unblocking agent and the blockage material, as well as the influence of the surfactant on the relative permeability, through a coupled numerical model. By using a two-layer game model, the two objective functions of unblocking and oil displacement are correlated through a concentration ratio coupling term, which realizes the quantitative characterization of the synergistic mechanism of the two processes and avoids the problem of unreasonable ratios caused by traditional empirical methods.

[0022] In summary, this invention solves the technical problem mentioned in the background art of poor oil enhancement effect caused by the difficulty in accurately optimizing the ratio of unblocking agent and oil displacement agent in the formation unblocking and oil displacement process of bio-derived composite systems. Attached Figure Description

[0023] Figure 1 This is a flowchart of the method of the present invention.

[0024] Figure 2 This is a graph showing the relative permeability under the action of the surface-active oil displacement agent in the examples.

[0025] Figure 3 This is a geological concept diagram of one injection and one extraction process in the embodiment.

[0026] Figure 4 The graph shows the relationship between the amount of oil added and the ratio of the unblocking agent and the surfactant in the examples. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0028] like Figure 1 The diagram shown is a flowchart of a numerical simulation method for integrated unblocking and oil displacement in a bio-based composite system provided by this invention. This method includes the following steps: S1. Construct a database of physicochemical properties of bio-derived composite systems, including the molecular weight of each component and the reduction curve of oil-water interfacial tension; S2. Establish a random field model for the spatial distribution of blockages, and use Monte Carlo simulation to generate multiple heterogeneous distributions of blockages; S3. Establish a de-blocking dynamics model and an oil displacement efficiency enhancement model, and construct a de-blocking-oil displacement coupled numerical model; S4. Establish a dynamic evolution prediction model for reservoir permeability, and introduce the dissolution fraction of plugging material as an intermediate variable; S5. Using operator splitting technology and adaptive time step control strategy, solve the multiphysics coupling equations. S6. Establish a geological concept model based on the unblocking-oil displacement coupled numerical model, set formation parameters and simulate the entire process of unblocking agent and oil displacement agent injection; S7. Establish a two-layer game model for ratio optimization, with the upper-layer model aimed at unblocking effect and the lower-layer model aimed at oil displacement efficiency being optimized collaboratively. S8 outputs oil increase index and optimizes the ratio of unblocking agent and surfactant oil displacement agent in the bio-derived composite system.

[0029] The step of constructing the physicochemical property database of the bio-derived composite system in step S1 is as follows: Based on indoor experiments, the molecular weight of the unblocking agent is set to 60. The molecular weight of the surface-active oil displacement agent was set to 500. The molecular weight of the blockage is set to 2000. The molecular weight of the residue after the blockage reaction is set to 245. The molecular weight of water is set to 18. The data on the change of oil-water interfacial tension with the concentration of surfactant were obtained from indoor experiments. Specifically, the oil-water interfacial tension was 25.0 wt% when the surfactant concentration was 0.0 wt%. When the concentration of the surfactant-induced oil displacement agent is 0.1 wt%, the oil-water interfacial tension is 1.2. When the concentration of the surfactant-driven oil displacement agent is 0.2 wt%, the oil-water interfacial tension is 0.086. When the concentration of the surfactant-driven oil displacement agent is 0.3 wt%, the oil-water interfacial tension is 0.079. When the concentration of the surfactant-driven oil displacement agent is 0.4 wt%, the oil-water interfacial tension is 0.073. When the concentration of the surfactant-driven oil displacement agent is 0.5 wt%, the oil-water interfacial tension is 0.068. The data on the variation of oil-water interfacial tension with the concentration of surfactant were fitted to a power-law decay model mathematical function, where the original oil-water interfacial tension without surfactant was 25.0. The lowest oil-water interfacial tension achievable by surfactant-driven oil displacement agents is 0.06. The optimal concentration required to achieve the lowest oil-water interfacial tension is half of 0.08 wt%, with an exponential coefficient of 3.15. The molecular weights of each component are used to establish the unblocking kinetic model in step S3, and the oil-water interfacial tension reduction curve is used to establish the oil displacement efficiency enhancement model in step S3.

[0030] The step S2, establishing a random field model for the spatial distribution of blockages, involves: using random field theory combined with geostatistical methods, and employing Monte Carlo simulation to generate 50 equally probable heterogeneous distributions of blockages; describing the uncertainty of the spatial distribution of blockages using a probability density function, which adopts a log-normal distribution, with a log-mean of 5.2 and a log-standard deviation of 0.8 for the blockage concentration; and describing the spatial correlation of blockages using a variogram, which adopts a spherical model with a principal direction range of 80°. The secondary direction stroke is 40. Vertical range is 5 Each heterogeneous distribution of blockage material is marked and a database of heterogeneous blockage material distributions is established. In the numerical simulation of step S6, a heterogeneous distribution of blockage material distributions is randomly selected as the initial blockage condition to establish a three-dimensional numerical model that reflects the true blockage characteristics.

[0031] The random field theory described here is used to describe the random spatial distribution characteristics of blockage concentration. The inputs include statistical parameters of blockage concentration and spatial correlation parameters, and the output is a spatial distribution field of blockage concentration. The statistical parameters of blockage concentration are derived from the logarithmic mean and logarithmic standard deviation set in step S2, and the spatial correlation parameters are derived from the principal direction range, secondary direction range, and vertical range set in step S2. The spatial distribution field of blockage concentration is used to establish a database of heterogeneous blockage distribution in step S2. The probability density function describes the probability of blockage concentration occurring near a certain value. The inputs include observed blockage concentration values ​​and distribution parameters, and the output is the probability density of the concentration distribution. The observed blockage concentration values ​​are derived from indoor experiments, and the distribution parameters are derived from the logarithmic mean and logarithmic standard deviation set in step S2. The probability density of the concentration distribution is used to generate a heterogeneous blockage distribution in step S2. The variogram is used to describe the correlation of blockage concentration at different spatial locations. The input includes spatial distance and direction angle, and the output is the semivariogram value. The spatial distance is derived from the grid step size set in step S6, and the direction angle is derived from the principal and secondary directions set in step S2. The semivariogram value is used in step S2 to determine the spatial correlation of blockage. The spherical model is used to fit the relationship between the variogram and spatial distance. The input includes spatial distance and range parameters, and the output is the theoretical semivariogram value. The spatial distance is derived from the grid step size set in step S6, and the range parameters are derived from the principal direction range, secondary direction range, and vertical range set in step S2. The theoretical semivariogram value is used in step S2 to verify the fitting accuracy of the variogram.

[0032] The step S3, establishing the unblocking-oil displacement coupled numerical model, involves: establishing an unblocking kinetic model; calculating the reaction kinetics using data processing software based on indoor experimental data; obtaining the reaction kinetic equation as follows: 3.33 molar amounts of unblocking agent react with 111.11 molar amounts of blockage material to produce 13.61 molar amounts of blockage material residue and 1 molar amount of water; based on experimental observations showing a linear relationship between the reaction rate and the unblocking agent concentration, the reaction order is determined to be 1; the pre-exponential factor for the reaction between the unblocking agent and the blockage material is 5; the above reaction kinetic equation uses a concentration-based power-law model to describe the reaction rate. The reaction rate is the product of the reaction rate constant and the concentrations of the unblocking agent and the blockage material. An unblocking agent transport equation is established to describe the convection, diffusion, and chemical reaction processes of the unblocking agent in the porous medium. An oil displacement enhancement model is established, and the relative permeability curves under the action of the surfactant are determined based on indoor experiments. These relative permeability curves include the relative permeability curves of the aqueous phase and the oil phase. The reaction kinetic equation is used to solve the multiphysics coupling equations in step S5, and the relative permeability curves are used to simulate the oil displacement process in step S6.

[0033] The reaction kinetic equation describes the chemical reaction process between the unblocking agent and the blockage. The inputs include the concentrations of the unblocking agent and the blockage, and the output is the concentration changes of each component after the reaction. The unblocking agent concentration is derived from the injection parameters set in step S6, and the blockage concentration is derived from the heterogeneous distribution database of the blockage established in step S2. The concentration changes of each component after the reaction are used to calculate the blockage dissolution fraction in step S4. The reaction order describes the relationship between the reaction rate and the reactant concentration; a reaction order of 1 indicates a linear relationship between the reaction rate and the reactant concentration. The pre-exponential factor describes the inherent rate of the chemical reaction; a larger pre-exponential factor value indicates a faster reaction. The power-law model quantitatively describes the chemical reaction rate. The inputs include the reaction rate constant and the reactant concentration, and the output is the reaction rate. The reaction rate constant is derived from the pre-exponential factor set in step S3, and the reactant concentration is derived from the unblocking agent concentration and the blockage concentration in step S3. The reaction rate is used to establish the unblocking agent transport equation in step S3. The unblocking agent transport equation describes the transport process of the unblocking agent in the formation. The inputs include porosity, Darcy velocity, diffusion coefficient, and chemical reaction term. The output is the distribution of the unblocking agent concentration over time and space. The porosity is derived from the matrix porosity set in step S6. The Darcy velocity is derived from the fluid flow subproblem solved in step S5. The diffusion coefficient is derived from indoor experimental measurements. The chemical reaction term is derived from the power-law model established in step S3. The distribution of the unblocking agent concentration over time and space is used to calculate the dissolution fraction of the blockage in step S4.

[0034] The step S4, establishing a dynamic evolution prediction model for reservoir permeability, involves: establishing the relationship between porosity and permeability based on the Carman-Kozeny equation, whereby permeability is directly proportional to the cube of porosity and inversely proportional to 1 minus the square of porosity; introducing the dissolution fraction of plugging material as an intermediate variable, where the dissolution fraction is the ratio of the mass of dissolved plugging material to the initial total mass of plugging material; measuring permeability changes at different dissolution stages through indoor core dissolution experiments, where the dissolution stages include permeability values ​​at plugging material dissolution fractions of 0, 0.2, 0.4, 0.6, 0.8, and 1.0; and training the permeability prediction model using a neural network algorithm. The permeability evolution prediction model comprises a neural network algorithm with an input layer, two hidden layers, and an output layer. The input layer has 3 neurons, the first hidden layer has 10 neurons, the second hidden layer has 6 neurons, and the output layer has 1 neuron. The input parameters include the dissolution fraction of plugging material, the initial permeability, and the porosity. The output parameter is the current permeability. The neural network algorithm is trained using a backpropagation algorithm with 120 training samples, 30 validation samples, and 30 test samples. After training, the neural network algorithm is embedded into numerical simulation software to accurately simulate the dynamic changes in reservoir properties. The current permeability is used to update the permeability parameters of the fluid flow subproblem in step S5.

[0035] The Carman-Kozeny equation describes the quantitative relationship between pore structure and permeability in porous media. The inputs include porosity and pore characteristic parameters, and the output is the permeability value. The porosity is derived from the matrix porosity set in step S6, and the pore characteristic parameters are derived from indoor experimental measurements. The permeability value is used to train the neural network algorithm in step S4. The dissolution fraction of the blockage material quantifies the degree of blockage material dissolution. The inputs include the mass of dissolved blockage material and the initial total mass of the blockage material, and the output is the dissolution fraction. The mass of dissolved blockage material is derived from the concentration changes of each component after the reaction calculated by the reaction kinetic equation in step S3, and the initial total mass of the blockage material is derived from the heterogeneous distribution database of the blockage material established in step S2. The dissolution fraction is used as an input parameter for the neural network algorithm in step S4. The neural network algorithm is used to establish a nonlinear mapping relationship. The inputs include the dissolution fraction of the blockage material, the initial permeability, and the porosity. The output is the predicted current permeability. The dissolution fraction of the blockage material is derived from the value calculated in step S4. The initial permeability is derived from the matrix permeability set in step S6. The porosity is also derived from the matrix porosity set in step S6. The predicted current permeability is used to update the permeability parameters of the fluid flow subproblem in step S5. The backpropagation algorithm is used to adjust the weight parameters of the neural network algorithm. The inputs include training samples and the target output. The output is the optimized network weights. The training samples are derived from the permeability change data measured in the indoor core dissolution experiment in step S4. The target output is also derived from the permeability change data measured in the indoor core dissolution experiment in step S4. The optimized network weights are used for setting the parameters of the neural network algorithm in step S4.

[0036] The steps in step S5, which employ operator splitting technology and adaptive time step control strategy to solve the multiphysics coupling equations, are as follows: The multiphysics coupling problem is decomposed into a fluid flow subproblem, a chemical reaction subproblem, and a concentration diffusion subproblem using operator splitting technology; within each time step, the fluid flow subproblem is solved sequentially to obtain the pressure field and velocity field; the chemical reaction subproblem is solved to obtain the concentration changes of each component; and the concentration diffusion subproblem is solved to obtain the final concentration distribution; an implicit iterative scheme is introduced to improve the stability of time integration, wherein the implicit iterative scheme adopts the Crank-Nicolson scheme; an adaptive time step control strategy is set, calculating the relative error between the solutions of two adjacent time steps; when the relative error is greater than 0.01, the time step is reduced to 0.5 times the original value; when the relative error is less than 0.001, the time step is increased to 1.2 times the original value; the time step range is [0.01, 2.0]. The Newton-Raphson iterative method is used to accelerate the solution of the nonlinear equation system, and the iterative convergence criterion is set as the residual being less than 1 / 3. The maximum number of iterations is 50; when the iteration does not converge, the time step is automatically reduced and the calculation is repeated; the pressure field and velocity field are used to calculate the Darcy velocity of the unblocking agent transport equation in step S3, the concentration changes of each component are used to calculate the dissolution fraction of the blockage in step S4, and the final concentration distribution is used to output the concentration field distribution in step S6.

[0037] The operator splitting technique is used to decouple the coupled problem into multiple independent subproblems. The input includes a set of coupled equations and a splitting operator, and the output is the solution to each subproblem. The coupled equations originate from the unblocking agent transport equation and reaction kinetic equation established in step S3. The splitting operator is derived from the mathematical definition of the operator splitting technique. The solutions to each subproblem are used in step S5 to solve the multiphysics coupling problem step-by-step. The implicit iterative scheme is used to improve the stability of numerical computation. The input includes the state variables at the current and next time steps, and the output is a stable time integral result. The state variable at the current time step originates from the calculation result of the previous time step in step S5, and the state variable at the next time step originates from the calculation result of the current time step in step S5. The stable time integral result is used to advance the time step in step S5. The adaptive time step control strategy dynamically adjusts computational accuracy and efficiency. Inputs include relative error and the current time step, and output is the adjusted time step. The relative error originates from the difference between the solutions of two adjacent time steps calculated in step S5. The current time step originates from the time step used in the previous time step in step S5. The adjusted time step is used for the calculation of the next time step in step S5. The Newton-Raphson iteration method solves a system of nonlinear equations. Inputs include the system of equations and initial guesses. Output is the numerical solution of the system of equations. The system of equations originates from the chemical reaction subproblem in step S5. The initial guesses originate from the calculation result of the previous time step in step S5. The numerical solution of the system of equations is used to obtain the concentration changes of each component in step S5. The Crank-Nicolson scheme is used for time discretization, employing the average value of the current and next time steps for calculation.

[0038] Specifically, step S6, which involves establishing a geological conceptual model based on the unblocking-oil displacement coupled numerical model, comprises: establishing a geological conceptual model based on actual stratigraphic parameters, wherein the geological conceptual model has a size of 200. ×200 ×30 The grid size is 40×40×5, and the grid step size is 5. ×5 ×6 The initial formation pressure was 10. The initial formation temperature was 50℃, the initial oil saturation was 0.68, and the reservoir depth was 1000 m. The reservoir thickness is 30. The water saturation is 0.32, the matrix porosity is 0.25, and the matrix permeability is 200. Import the molecular weight of each component and the oil-water interfacial tension reduction curve determined in step S1; introduce the reaction kinetic equation of the reaction between the unblocking agent and the plugging material and the relative permeability curve under the action of the surfactant oil displacement agent determined in step S3; introduce the reservoir permeability dynamic evolution prediction model determined in step S4; set the injection and production parameters, and the geological concept model adopts a one-injection-one-production mode; in the plugging model, first inject the plugging material 36 Simulate formation blockage, then inject water 12 Shut down production well and injection well 4 Finally, water is continuously injected, and the cumulative oil production at this point is recorded as the oil production in the blocked state; in the unblocking model, 36 units of blockage material are injected first. Simulated formation blockage, re-injection of a bio-based composite system 12 Shut down production well and injection well 4 Finally, water is continuously injected, and the cumulative oil production at this point is recorded as the oil production in the unblocked state. The oil increase is calculated by subtracting the oil production in the blocked state from the oil production in the unblocked state. The oil production in the blocked state is used to calculate the oil increase in step S6, and the oil increase is used to evaluate the effect of the bio-based composite system in step S8.

[0039] The step S7, establishing the two-layer game model for ratio optimization, involves: establishing an upper-layer model with the goal of unblocking effect and a lower-layer model with the goal of oil displacement efficiency; the objective function of the upper-layer model is to maximize the dissolution rate of the blockage, with inputs including the concentration of the unblocking agent and the injection time of the unblocking agent, and the output being the dissolution rate of the blockage; the objective function of the lower-layer model is to maximize the oil displacement efficiency, with inputs including the concentration of the surfactant and the injection time of the surfactant, and the output being the oil displacement efficiency; the two objective functions are correlated through the ratio of the concentrations of the unblocking agent and the surfactant as a coupling term; the constraints of the upper-layer model include the unblocking agent concentration range being ∈ [0.1, 5.0] wt% and the unblocking agent injection time range being ∈ [6, 24]. The lower-level model constraints include a surfactant concentration range of [0.05, 2.0] wt% and a surfactant injection time range of [6, 24]. A sequential solution strategy is adopted to solve the bi-layer game model for ratio optimization. First, the parameters of the lower-layer model are fixed to solve the upper-layer model to obtain the optimal unblocking agent configuration. Then, the parameters of the upper-layer model are fixed to solve the lower-layer model to obtain the optimal surfactant displacement agent configuration. The solution is iterated until both layers of the model reach Nash equilibrium. The criterion for determining the Nash equilibrium is that the change in the concentration ratio of unblocking agent to surfactant displacement agent is less than 1% in two consecutive iterations. The optimal unblocking agent configuration is used to determine the initial optimized ratio range in step S8, and the optimal surfactant displacement agent configuration is used to determine the initial optimized ratio range in step S8.

[0040] The objective function of the upper-level model is expressed as follows: the dissolution rate of the blockage per unit time is proportional to the product of the natural logarithm of the unblocking agent concentration and the square root of the unblocking agent injection time. The objective function is dimensionless, with each term divided by its standard value. The unblocking agent concentration is derived from the upper-level model constraints set in step S7, and the unblocking agent injection time is derived from the upper-level model constraints set in step S7. The dissolution rate of the blockage per unit time is used to evaluate the unblocking effect in step S7. The objective function of the lower-level model is expressed as follows: the oil displacement efficiency is proportional to the product of the concentration of the surfactant oil displacement agent to the power of 0.6 and the injection time of the surfactant oil displacement agent to the power of 0.8, divided by 1, and then the square of the ratio of the concentration of the surfactant oil displacement agent to the critical micelle concentration is added. The objective function is dimensionless, and each term is divided by its standard value. The concentration of the surfactant oil displacement agent comes from the lower-level model constraint conditions set in step S7. The injection time of the surfactant oil displacement agent comes from the lower-level model constraint conditions set in step S7. The critical micelle concentration comes from the indoor experimental measurement value. The oil displacement efficiency is used to evaluate the oil displacement effect in step S7.

[0041] The coupling term is used to associate the decision variables of the upper and lower level models. The inputs include the concentrations of the unblocking agent and the surfactant displacement agent, and the output is a concentration ratio constraint. The unblocking agent concentration originates from the input parameters of the upper level model in step S7, and the surfactant displacement agent concentration originates from the input parameters of the lower level model in step S7. The concentration ratio constraint is used in step S7 to limit the feasible solution space of the upper and lower level models. The sequential solution strategy is used to alternately optimize the two level models. The inputs include the objective function and constraints of the upper level model and the lower level model. The output is the optimal ratio to reach Nash equilibrium. The objective function and constraints of the upper level model originate from the upper level model established in step S7, and the objective function and constraints of the lower level model originate from the lower level model established in step S7. The optimal ratio to reach Nash equilibrium is used in step S8 to determine the initial optimization ratio range. The Nash equilibrium state indicates that both layers of the model have reached the optimal solution of their respective objectives and no longer change their strategies.

[0042] The step S8, which outputs the oil increase index and optimizes the ratio of unblocking agent and surfactant in the bio-based composite system, involves the following steps: Based on the equilibrium solution of the bi-layer game model for ratio optimization obtained in step S7, the initial optimized ratio range is determined; five schemes are designed with unblocking agent to surfactant ratios of 1:2, 1:1, 2:1, 3:1, and 4:1; the geological concept model established in step S6 is run for each scheme to calculate the oil increase; comparing the oil increase results of the five schemes, the oil increase in scheme 1, with an unblocking agent to surfactant ratio of 1:2, is 4883. When the ratio of the unblocking agent to the surfactant-based oil displacement agent in Scheme 2 is 1:1, the oil increase is 4932. When the ratio of the unblocking agent to the surfactant-driven oil displacement agent in Scheme 3 is 2:1, the oil increase is 5017. When the ratio of the unblocking agent to the surfactant-driven oil displacement agent in Scheme 4 is 3:1, the oil increase is 5071. When the ratio of the unblocking agent to the surfactant-driven oil displacement agent in Scheme 5 is 4:1, the oil increase is 5061. The results above show that Scheme 4 is the most effective, i.e., when the ratio of unblocking agent to surfactant-driven oil displacement agent is 3:1, the oil increase is 5071. It maximizes oil production; outputs the optimal blending scheme and the corresponding oil production, pressure field distribution, saturation field distribution, blockage dissolution distribution, and permeability evolution curves; and guides on-site scheme design based on the optimal blending scheme.

[0043] Optionally, the present invention also provides a method for implementing a numerical simulation system based on a bio-based composite system for unblocking and oil displacement to remove formation blockages, wherein the computer is equipped with a readable storage medium that stores program instructions, and the program instructions execute the above-described method when the computer is run.

[0044] It should be noted that the key technical ideas of this invention include the following three aspects. The first key technical idea is to construct a probabilistic model of the spatial distribution of plugging material using random field theory combined with Monte Carlo simulation technology. This method describes the statistical characteristics and spatial correlation of plugging material concentration through a log-normal distribution and a spherical variogram, generating multiple equally probable heterogeneous distributions. Compared with traditional deterministic models, this method can fully consider the uncertainty and randomness of formation plugging, avoiding prediction biases that may be caused by a single distribution pattern, and improving the reliability and applicability of simulation results. The second key technical idea is to establish a reservoir permeability dynamic evolution prediction model based on a neural network algorithm. This model introduces the dissolution fraction of plugging material as an intermediate variable, and establishes a nonlinear mapping relationship between the dissolution fraction, initial permeability, porosity, and current permeability through a multi-layer neural network. Compared with traditional empirical formulas or simplified models, this method can accurately capture the complex nonlinear laws of permeability evolution and achieve accurate prediction of dynamic changes in reservoir properties. The third key technical approach is to establish a two-layer game theory model for ratio optimization. This model treats unblocking effectiveness and oil displacement efficiency as two independent optimization objectives, and uses Nash equilibrium theory to find the optimal balance point. Compared with traditional single-objective optimization methods, this method can achieve synergistic optimization between the two processes of unblocking and oil displacement, avoiding the problem of neglecting one aspect for the other. The synergistic effect of these three key technical approaches lies in the fact that the random field model provides initial conditions reflecting the actual characteristics of blockage, the neural network model realizes the dynamic updating of physical property parameters, and the two-layer game theory model finds the optimal ratio scheme based on these conditions. The three form a complete technical chain from uncertainty quantification to dynamic evolution prediction to ratio optimization. Compared with traditional methods that only consider a single factor or use static models, this synergistic technical system can comprehensively and accurately simulate the entire process of unblocking and oil displacement, significantly improving the scientific nature of ratio optimization and its guiding value for field applications.

[0045] It should be noted that this invention also solves the following technical problem: In numerical simulation of formation unblocking, the blockage material in the reservoir exhibits a highly heterogeneous distribution. Traditional simulation methods use the assumption of uniform distribution or simple partitioning to describe the spatial distribution of the blockage material, which cannot accurately reflect the randomness and spatial correlation of the blockage material, resulting in a large deviation between the simulation results and the actual formation conditions. This invention establishes a random field model of the spatial distribution of the blockage material, uses random field theory combined with geostatistical methods, and utilizes Monte Carlo simulation technology to generate multiple equally probable heterogeneous distributions of the blockage material. The uncertainty of the spatial distribution of the blockage material is described by the probability density function, and the spatial correlation of the blockage material is described by the variogram function. A three-dimensional numerical model reflecting the real blockage characteristics is established, which accurately captures the influence of the heterogeneous distribution of the blockage material on the unblocking effect and improves the simulation accuracy.

[0046] Furthermore, during the unblocking process, reservoir permeability dynamically changes as the plugging material dissolves. Traditional methods use static permeability parameters or simple linear relationships to describe permeability evolution, which cannot accurately characterize the complex nonlinear relationship between permeability and the degree of plugging material dissolution, leading to inaccurate predictions of formation property recovery processes. This invention establishes a dynamic evolution prediction model for reservoir permeability, introduces the dissolution fraction of plugging material as an intermediate variable to connect chemical reaction processes and property changes, and uses a neural network algorithm to train the permeability evolution prediction model to establish a nonlinear mapping relationship between the dissolution fraction of plugging material, initial permeability, porosity, and current permeability. The trained neural network is then embedded into numerical simulation software to achieve real-time updates of permeability parameters, accurately simulating the dynamic changes in reservoir properties and providing a reliable basis for evaluating the unblocking effect.

[0047] Specifically, the principle of this invention is as follows: The reason this invention can solve the problem of ratio optimization lies in establishing a complete multi-physics coupled simulation framework and a two-layer optimization system. It quantitatively describes the reaction process between the unblocking agent and the blockage material through the reaction kinetic equation in the unblocking kinetic model. Combined with the reservoir permeability dynamic evolution prediction model, it introduces the dissolution fraction of the blockage material as an intermediate variable to establish the evolution relationship of permeability with the unblocking process, achieving accurate quantitative evaluation of the unblocking effect. Through the relative permeability curve and the oil-water interfacial tension reduction curve in the oil displacement enhancement model, it describes the influence mechanism of the surfactant on the oil displacement process, achieving accurate evaluation of oil displacement efficiency. The ratio optimization two-layer game model maximizes the dissolution rate of the blockage material in the upper model and maximizes the oil displacement efficiency in the lower model. It links the two objective functions through a concentration ratio coupling term and uses a sequential solution strategy to alternately optimize until a Nash equilibrium is reached, ensuring that both the unblocking and oil displacement processes simultaneously achieve their respective optimal objectives. This achieves accurate optimization of the ratio of the bio-based composite system, conforming to the basic logic of multi-objective collaborative optimization.

[0048] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0049] The specific implementation of step S1 involves determining the physicochemical properties of each component in the bio-based composite system through indoor experiments, fitting the collected data using the nonlinear least squares method to obtain the characteristic parameters of the power-law decay model. The mathematical function of the power-law decay model is expressed as follows: ; In the formula, The oil-water interfacial tension is expressed in units of... ; The original oil-water interfacial tension, in units of The value is 25.0; The minimum oil-water interfacial tension, in units of The value is 0.06; The concentration of the surfactant-induced oil displacement agent is expressed in wt%; The concentration required to achieve the lowest oil-water interfacial tension is half of the optimal concentration, expressed in wt%, and the value is 0.08. The coefficient is an exponential coefficient, dimensionless, with a value of 3.15. This model can accurately describe the quantitative relationship between surfactant concentration and interfacial tension, providing reliable input data for oil displacement enhancement models.

[0050] The specific implementation of step S2 is based on constructing a probabilistic model of the spatial distribution of blockages using random field theory and geostatistical methods. First, it is defined that the concentration of blockages follows a log-normal distribution, and the probability density function is expressed as follows: ; In the formula, The probability density of blockage concentration, in units of ; The concentration of the blockage substance is expressed in units of [unit missing]. ; The value is the logarithmic mean of the blockage concentration, dimensionless, and 5.2. The value is the logarithmic standard deviation of the blockage concentration, dimensionless, with a value of 0.8; Pi is a dimensionless mathematical constant with a value of 3.14159. It is the natural logarithm function; It is an exponential function; The square root function is used. Next, a spherical variogram model is employed to describe the spatial correlation of blockages. The variogram is expressed as follows: when hour, ; when hour, ; In the formula, This is the semivariance value, in units of ; Spatial distance, unit: ; The variance of the gold nugget is expressed in units of 1000g. A value of 0.1 indicates random variation smaller than the sampling scale; The arch height is expressed in units of 1 / 2000. A value of 0.9 indicates spatial structural variation; For variable range, the unit is The values ​​are 80° in the primary direction, 40° in the secondary direction, and 5° in the vertical direction, representing the effective distance of spatial correlation. Then, Monte Carlo simulation is used to generate 50 equally probable heterogeneous distributions of blockage material based on the above statistical parameters. Finally, all realizations are numbered and a database is established. In subsequent numerical simulations, realizations are randomly selected as initial blockage conditions. This method can fully consider the uncertainty of formation blockage and improve the reliability of simulation results.

[0051] The specific implementation of step S3 involves establishing a coupled mathematical model describing the unblocking and oil displacement processes. First, the stoichiometric relationship of the reaction is determined based on indoor batch reaction experimental data. The reaction kinetic equation is expressed as follows: ; In the formula, This refers to the molar concentration of the unblocking agent, in units of... ; Molar concentration of the blockage material, in units of ; The molar concentration of the residue after the blockage reaction is given in units of . ; Molar concentration of water, in units of The reaction order was determined to be 1 by analyzing the concentration versus time curve, the pre-exponential factor was determined to be 5, and the reaction rate was described using a power-law model, as follows: ; In the formula, The reaction rate is expressed in units of 1000 mcg. ; The reaction rate constant is expressed in units of 1000 m / s. The value is 5. The transport equation for the unblocking agent in the porous medium is established as follows: ; In the formula, Porosity is a dimensionless quantity. For time, the unit is ; Darcy velocity vector, unit: ; This is the diffusion coefficient, in units of... Through indoor experiments, the empirical value is... ; For gradient operators; It is a divergence operator; The sign is for the partial derivative. Next, an oil displacement enhancement model is established. The relative permeability curves of the aqueous and oil phases under the action of surfactants are determined through core displacement experiments. Finally, the unblocking kinetics model is coupled with the oil displacement enhancement model to form a complete mathematical framework that can simultaneously describe the dissolution of blockages and the displacement of crude oil.

[0052] The specific implementation of step S4 involves establishing a predictive model for the dynamic changes of reservoir physical parameters during the unblocking process. First, based on the Carman-Kozeny equation, a theoretical relationship between porosity and permeability is established, as follows: ; In the formula, Current penetration rate, in units of ; Initial penetration rate, in units of ; The current porosity is dimensionless. The initial porosity is dimensionless. Next, the dissolution fraction of the blockage material is defined as follows: ; In the formula, The fraction of blockage dissolved is dimensionless. The mass of dissolved blockage material is expressed in units of [missing information]. It is obtained by time integration of the reaction kinetic equation; The initial total mass of the blockage, in units of The permeability is obtained by integrating the blockage concentration field over the entire computational domain in step S2. Then, permeability data at different dissolution stages are measured using indoor core dissolution experiments to obtain permeability values ​​at dissolution fractions of 0, 0.2, 0.4, 0.6, 0.8, and 1.0. A neural network model is established, with input parameters including blockage dissolution fraction, initial permeability, and porosity, and output parameter being the current permeability. The network is trained using the backpropagation algorithm, employing 120 training samples, 30 validation samples, and 30 test samples. After training, the neural network is embedded in numerical simulation software to achieve real-time prediction of the dynamic evolution of permeability. This method can accurately capture the nonlinear evolution law.

[0053] The specific implementation of step S5 involves using advanced numerical computation techniques to solve the multiphysics coupling equations. First, the complex coupling problem is decomposed into three independent sub-problems—fluid flow, chemical reaction, and concentration diffusion—using operator splitting techniques. Each sub-problem is solved sequentially within each time step. Second, the Crank-Nicolson implicit iterative scheme is used for time discretization, as described below: ; In the formula, For the first The concentration of the unblocking agent at each time step, in units of ; For the first The concentration of the unblocking agent at each time step, in units of ; For time steps, the unit is ; It is a spatial differential operator, which includes convection, diffusion and reaction terms; Let be the time step number, which is dimensionless. Then, an adaptive time step control strategy is designed, and the relative error is calculated as follows: ; In the formula, This is a relative error, dimensionless. for Norm operator, defined as ,in The total number of grid cells is dimensionless. For the first Concentration values ​​for each grid cell, in units of , is the grid number, dimensionless. When At that time, the time step was adjusted to ;when At that time, the time step was adjusted to The step size is limited to 0.01 to 2.0. Between, among The adjusted time step, in units of , The time step before adjustment, in units of The Newton-Raphson iterative method is used to solve the nonlinear equation system, and the residual convergence criterion is set to be less than... The maximum number of iterations is 50. When the iteration fails to converge, the time step is automatically reduced and the calculation is recalculated. This solution strategy can significantly improve computational efficiency while ensuring computational accuracy.

[0054] The specific implementation method of step S6 is the same as described above, and will not be repeated in detail here.

[0055] The specific implementation of step S7 involves establishing a two-layer game model for ratio optimization and solving for the Nash equilibrium. First, an upper-layer model is constructed with the objective of maximizing the dissolution rate of the blockage. The objective function is expressed as follows: ; In the formula, The objective function value of the upper-level model is dimensionless. The concentration of the unblocking agent is expressed in wt%; This is a reference concentration for the unblocking agent, in wt%, with a value of 1.0. This represents the maximum concentration of the unblocking agent, expressed in wt%, with a value of 5.0. The time for injecting the unblocking agent is expressed in units of time. ; For reference time, the unit is... The value is 12; Maximum injection time, in units of The value is 24. The constraints are set as follows: unblocking agent concentration range 0.1 to 5.0 wt%, injection time range 6 to 24 seconds. Next, a lower-level model is constructed with the goal of maximizing oil displacement efficiency. The objective function is expressed as follows: ; In the formula, This represents the objective function value of the lower-level model, which is dimensionless. The concentration of the surfactant-induced oil displacement agent is expressed in wt%; The value is a reference concentration for surfactant-induced oil displacement agents, expressed in wt%, with a value of 0.5. The injection time of the surfactant-induced oil displacement agent is expressed in units of... ; The critical micelle concentration, expressed in wt%, was determined through indoor experiments and empirically valued at 0.12. The constraints were set as follows: surfactant concentration range of 0.05 to 2.0 wt%, and injection time range of 6 to 24 hours. By using the concentration ratio of the unblocking agent and the surfactant oil displacement agent as a coupling term to associate two-layer models, and then employing a sequential solution strategy to alternately optimize the two-layer models, iterating repeatedly until the concentration ratio change between two consecutive iterations is less than 1%, reaching a Nash equilibrium state, this model can find the optimal balance between unblocking effect and oil displacement efficiency.

[0056] The specific implementation method of step S8 is the same as described above, and will not be repeated in detail here.

[0057] It should be explained that the power-law decay model describes the nonlinear reduction effect of surfactant concentration on interfacial tension by introducing half-saturation concentration and exponential coefficient. This model is based on Langmuir adsorption theory. When the surfactant concentration is low, the interfacial tension decreases rapidly. As the concentration increases, the rate of decrease gradually slows down until it approaches the minimum value. The model parameters are obtained by fitting experimental data and can accurately predict the changes in interfacial tension at different concentrations, providing a quantitative basis for optimizing the dosage of oil displacement agents.

[0058] The spherical variogram model uses a piecewise function to describe the decay of the spatial correlation of blockage with distance. When the spatial distance is less than the range, the correlation decays according to a cubic polynomial. After the range is exceeded, the correlation tends to zero. This model can reflect the anisotropic characteristics of formation blockage. By setting range parameters in different directions, the differences between the main direction and the secondary direction are characterized. Combined with Monte Carlo simulation to generate multiple equally probable realizations, the uncertainty of blockage distribution is fully considered, thereby improving the reliability and representativeness of the simulation results.

[0059] It needs to be explained that the unblocking agent transport equation comprehensively considers three physicochemical processes: convective transport, molecular diffusion, and chemical reaction. Among them, the convection term... Describes the macroscopic transport and diffusion terms of the unblocking agent as it flows through the fluid. Describes concentration gradient-driven microscopic transport, reaction term This equation describes the chemical consumption of unblocking agents and blockages. It can comprehensively characterize the spatiotemporal evolution of unblocking agents in porous media, providing a scientific basis for optimizing injection schemes.

[0060] It should be explained that the Carman-Kozeny equation establishes the theoretical relationship between porosity and permeability. In this equation, permeability is proportional to the cube of porosity and is related to the tortuosity of the pore structure. By introducing initial porosity and initial permeability for normalization, it is possible to predict the dynamic evolution of permeability caused by changes in porosity, thus providing a basis for evaluating the physical property parameters for unblocking effectiveness.

[0061] It should be explained that the dissolution fraction of plugging material is defined as the ratio of the mass of dissolved plugging material to the total mass of the initial plugging material. This parameter can intuitively quantify the plugging process. As an input parameter of the neural network model, it connects the chemical reaction process with the physical property evolution process, establishes a bridge between reaction kinetics and reservoir properties, and realizes multi-physics field coupled simulation.

[0062] It should be explained that the Crank-Nicolson scheme uses the average of the current time and the next time step for time discretization. This scheme has second-order accuracy and unconditional stability, which can maintain computational accuracy over a large time step and significantly improve computational efficiency, making it suitable for long-scale reservoir numerical simulation.

[0063] It should be explained that the adaptive time step control strategy dynamically adjusts the step size by monitoring the relative error of solutions in adjacent time steps. When the error is large, the step size is reduced to ensure accuracy, and when the error is small, the step size is increased to improve efficiency. This strategy can achieve an automatic balance between computational accuracy and computational efficiency, avoiding the problems of insufficient accuracy or computational waste that may be caused by manually setting a fixed step size.

[0064] It should be explained that the objective function of the two-layer game model is dimensionless to eliminate the dimensional differences between different physical quantities. The objective function of the upper-layer model describes the nonlinear effects of unblocking agent concentration and injection time on the unblocking effect through logarithmic and square root functions. The objective function of the lower-layer model describes the synergistic effect of surfactant concentration and injection time on oil displacement efficiency and the inhibitory effect after exceeding the critical micelle concentration through power and fractional functions. The two layers of models are interconnected through concentration ratio coupling terms, and the sequential solution strategy seeks the Nash equilibrium solution through alternating optimization. This model can find the optimal balance between the two objectives of unblocking effect and oil displacement efficiency, providing theoretical guidance for the optimization of bio-based composite systems, achieving synergistic effects of unblocking and oil displacement, and significantly improving recovery rate and oil production.

[0065] To better understand and implement this invention, a specific application scenario of the invention is provided below as Example 2: To verify the effectiveness of the invention, technicians built a numerical simulation analysis environment in the laboratory. By establishing a three-dimensional geological model based on indoor core experimental data, the entire process of unblocking and oil displacement in the bio-derived composite system was simulated, and the ratio of unblocking agent to surfactant-based oil displacement agent was optimized. The laboratory previously conducted numerous core displacement experiments and chemical reaction kinetic experiments, obtaining complete basic data, which provided reliable parameter support for the numerical simulation.

[0066] First, the technicians constructed a database of the physicochemical properties of the bio-based composite system. Based on the molecular weights of the unblocking agent, surfactant-based oil-displacing agent, blockage material, post-reaction residue of the blockage material, and water determined in indoor experiments, the molecular weight of the unblocking agent was specifically set to 60. The molecular weight of the surfactant-driven oil displacement agent is 500. The molecular weight of the blockage is 2000. The molecular weight of the residue after the blockage reaction is 245. The molecular weight of water is 18. The interfacial tension data of oil and water at different concentrations of surfactants were measured under laboratory conditions using an interfacial tension meter, as shown in Table 1.

[0067] Table 1. Data on the variation of oil-water interfacial tension with the concentration of surfactant-based oil displacement agent.

[0068] By fitting the data in Table 1 using a power-law decay model, the initial oil-water interfacial tension without surface-active displacement agents was found to be 25.0. The lowest oil-water interfacial tension achievable by surfactant-driven oil displacement agents is 0.06. The optimal concentration required to achieve the lowest oil-water interfacial tension is half of 0.08 wt%, with an exponential coefficient of 3.15. These parameters provide an accurate physical property basis for subsequent numerical simulations.

[0069] Subsequently, technicians established a random field model for the spatial distribution of plugging material. Fifty equally probable heterogeneous distributions of plugging material were generated using Monte Carlo simulation techniques. The plugging material concentration followed a log-normal distribution with a log-mean of 5.2 and a log-standard deviation of 0.8. Spatial correlation was described using a spherical variogram model, with a primary directional range of 80 m, a secondary directional range of 40 m, and a vertical range of 5 m. This random field model can realistically reflect the heterogeneous characteristics of actual reservoir plugging, providing more realistic initial conditions for numerical simulation.

[0070] Regarding the establishment of the unblocking kinetic model, based on laboratory chemical reaction kinetic experimental data, data processing software was used to calculate and determine the reaction kinetic equation as follows: 3.33 molar amounts of the unblocking agent react with 111.11 molar amounts of the blockage material to produce 13.61 molar amounts of the blockage material residue and 1 molar amount of water. By observing the linear relationship between the laboratory reaction rate and the concentration of the unblocking agent, the reaction order was determined to be 1. The pre-exponential factor of the reaction between the unblocking agent and the blockage material was set to 5, and a power-law model was used to describe the reaction rate. An unblocking agent transport equation was established to comprehensively describe the convection, diffusion, and chemical reaction processes of the unblocking agent in porous media. Regarding the oil displacement enhancement model, based on laboratory core displacement experiments, the relative permeability curves of the aqueous and oil phases under the action of the surfactant were determined, such as... Figure 2 As shown, the curve reflects the ability of surface-active oil displacement agents to significantly improve the flow characteristics of oil and water two phases.

[0071] Technicians further established a dynamic evolution prediction model for reservoir permeability. Based on the Carman-Kozeny equation, a quantitative relationship between porosity and permeability was established, introducing the dissolution fraction of plugging material as an intermediate variable to connect chemical reactions and changes in physical properties. Permeability changes were measured using laboratory core dissolution experiments when the dissolution fraction of plugging material was 0, 0.2, 0.4, 0.6, 0.8, and 1.0. A neural network algorithm was used to train the permeability evolution prediction model. This neural network includes 3 neurons in the input layer, 10 neurons in the first hidden layer, 6 neurons in the second hidden layer, and 1 neuron in the output layer. The input parameters are the dissolution fraction of plugging material, initial permeability, and porosity; the output parameter is the current permeability. The backpropagation algorithm was used to train the neural network, with 120 training samples, 30 validation samples, and 30 test samples. After training, the neural network was embedded into numerical simulation software, achieving accurate prediction of dynamic changes in reservoir physical properties.

[0072] In terms of numerical solution, an operator splitting technique is employed to decompose the multiphysics coupling problem into three sub-problems: fluid flow, chemical reaction, and concentration diffusion. Each sub-problem is solved sequentially within each time step, and a Crank-Nicolson implicit iterative scheme is introduced to improve time integration stability. An adaptive time step control strategy is implemented: when the relative error between solutions from two adjacent time steps is greater than 0.01, the time step is reduced to 0.5 times its original value; when the relative error is less than 0.001, the time step is increased to 1.2 times its original value, with a time step range of 0.01 to 2.0 hours. The Newton-Raphson iterative method is used to accelerate the solution of the nonlinear equations, and the iterative convergence criterion is set as a residual less than... The maximum number of iterations is 50.

[0073] A geological conceptual model was established based on stratigraphic parameter data obtained in the laboratory. The model dimensions were 200m×200m×30m, with a grid size of 40×40×5 and a grid step size of 5m×5m×6m. The stratigraphic parameter settings are shown in Table 2.

[0074] Table 2 Parameter Table of Reservoir Numerical Simulation Model

[0075] Import the constructed molecular weight of each component and the oil-water interfacial tension reduction curves, introduce the unblocking kinetic equation and relative permeability curve, and introduce the reservoir permeability dynamic evolution prediction model. For example... Figure 3As shown, the model uses a one-injection-one-production mode for numerical simulation. In the blockage model, the blockage material is injected for 36 hours to simulate the formation blockage process, followed by water injection for 12 hours, then the production well and injection well are shut down for 4 hours to allow for a well-suspension reaction. Finally, water injection continues until the simulation ends, and the cumulative oil production at this point is recorded as the oil production in the blockage state. In the unblocking model, the blockage material is injected for 36 hours to simulate the formation blockage process, followed by the injection of a bio-based composite system for 12 hours, then the production well and injection well are shut down for 4 hours to allow for a well-suspension reaction. Finally, water injection continues until the simulation ends, and the cumulative oil production at this point is recorded as the oil production in the unblocking state. The increase in oil production is calculated by subtracting the blockage state oil production from the unblocking state oil production.

[0076] Technicians established a two-layer game model for ratio optimization. The upper-layer model aims to maximize the dissolution rate of the blockage, with inputs being the concentration of the unblocking agent and the injection time. Constraints include an unblocking agent concentration range of 0.1 to 5.0 wt% and an injection time range of 6 to 24 hours. The lower-layer model aims to maximize the oil displacement efficiency, with inputs being the concentration of a surfactant-based oil displacement agent and the injection time. Constraints include a surfactant-based oil displacement agent concentration range of 0.05 to 2.0 wt% and an injection time range of 6 to 24 hours. The two objective functions are interrelated through the ratio of unblocking agent to surfactant-based oil displacement agent concentrations as a coupling term. A sequential solution strategy is used to solve the two-layer game model for ratio optimization. First, the parameters of the lower-layer model are fixed to solve the upper-layer model to obtain the optimal unblocking agent configuration. Then, the parameters of the upper-layer model are fixed to solve the lower-layer model to obtain the optimal surfactant-based oil displacement agent configuration. This iterative solution continues until the change in the ratio of unblocking agent to surfactant-based oil displacement agent concentrations between two consecutive iterations is less than 1%, reaching a Nash equilibrium.

[0077] Based on the equilibrium solution of the two-layer game model, the initial optimal ratio range was determined, and five schemes were designed with the ratio of plugging agent to surfactant oil displacement agent being 1:2, 1:1, 2:1, 3:1, and 4:1, respectively. A geological conceptual model was run for each scheme to calculate the oil increase. The oil increase results for different ratio schemes are shown in Table 3.

[0078] Table 3. Oil Increment by Different Proportions of Unblocking Agent and Surfactant Oil Displacement Agent

[0079] Table 3 shows that Scheme 4 is the most effective, that is, when the ratio of unblocking agent to surfactant oil displacement agent is 3:1, the oil increase reaches 5071. The optimal ratio is the one that yields the highest increase in oil production among the five options. The optimal ratio and its corresponding increases in oil production, pressure field distribution, saturation field distribution, blockage dissolution distribution, and permeability evolution curves are output, providing a scientific basis for on-site design.

[0080] This invention represents a significant technological advancement over traditional methods. Traditional unblocking and oil displacement schemes rely primarily on empirical judgment and single physical experiments, failing to accurately describe the chemical reaction kinetics between the unblocking agent and the plugging material, and struggling to predict the dynamic evolution of reservoir permeability. This invention establishes an unblocking kinetic model and an oil displacement efficiency enhancement model, constructing a coupled numerical model of unblocking-oil displacement, achieving fully coupled simulation of chemical reactions and multiphase flow. By introducing the dissolution fraction of the plugging material as an intermediate variable, a dynamic evolution prediction model of reservoir permeability is established using a neural network algorithm, solving the problem of traditional methods' inability to dynamically track changes in reservoir properties. The use of operator splitting technology and an adaptive time step control strategy significantly improves solution efficiency while maintaining computational accuracy. The established two-layer game model for ratio optimization achieves systematic optimization of the ratio of unblocking agent to surfactant-driven oil displacement agent through the synergistic effect of the upper-layer model optimizing the unblocking effect and the lower-layer model optimizing the oil displacement efficiency, overcoming the inefficiencies and unreliable results of traditional trial-and-error methods. The established stochastic field model of the spatial distribution of blockage material was realized by generating multiple equally probable heterogeneous distributions of blockage material using Monte Carlo simulation technology. This model realistically reflects the heterogeneous characteristics of formation blockage and avoids the limitation of deterministic models in being unable to characterize the complexity of actual formations.

[0081] It should be noted that the variables involved in this invention are explained in detail in Table 4.

[0082] Table 4. Variable Explanation Table

[0083] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A numerical simulation method for integrated unblocking and oil displacement in a bio-based composite system, characterized in that, By constructing a database of the physicochemical properties of a bio-based composite system and a random field model of the spatial distribution of plugging agents, a plugging kinetic model and an oil displacement efficiency enhancement model are established. A coupled numerical model for plugging and oil displacement is constructed, and a dynamic evolution prediction model for reservoir permeability is established. The dissolution fraction of plugging agents is introduced as an intermediate variable. Operator splitting technology and adaptive time step control strategy are used to solve the multi-physics field coupled equations. A geological conceptual model is established to simulate the entire process of injecting plugging agents and oil displacement agents. A two-layer game model for ratio optimization is established, with the upper-layer model aiming at the plugging effect and the lower-layer model aiming at the oil displacement efficiency working together to optimize the ratio of plugging agents and surfactants in the bio-based composite system.

2. The method according to claim 1, characterized in that, The steps for constructing the physicochemical property database of the bio-derived composite system are as follows: the molecular weight of each component is set according to the indoor experiment; the data on the change of oil-water interfacial tension with the concentration of surfactant oil displacement agent is obtained from the indoor experiment; and the data on the change of oil-water interfacial tension with the concentration of surfactant oil displacement agent is fitted into a power-law decay model mathematical function.

3. The method according to claim 2, characterized in that, The steps for establishing a random field model of the spatial distribution of blockages specifically involve using random field theory combined with geostatistical methods, generating multiple equally probable heterogeneous distributions of blockages using Monte Carlo simulation technology, describing the uncertainty of the spatial distribution of blockages through a probability density function, and describing the spatial correlation of blockages through a variogram function.

4. The method according to claim 3, characterized in that, The probability density function adopts a log-normal distribution, and the variogram adopts a spherical model.

5. The method according to claim 4, characterized in that, The steps for establishing the coupled numerical model for unblocking and oil displacement specifically involve establishing an unblocking kinetic model, obtaining reaction kinetic equations based on data from indoor experiments and using data processing software, selecting a concentration-based power-law model to describe the reaction rate, and establishing an unblocking agent transport equation to describe the convection, diffusion, and chemical reaction processes of the unblocking agent in porous media.

6. The method according to claim 5, characterized in that, The establishment of the oil displacement enhancement model specifically involves determining the relative permeability curves under the action of surfactant oil displacement agents based on indoor experiments. The relative permeability curves include the relative permeability curves of the aqueous phase and the relative permeability curves of the oil phase.

7. The method according to claim 6, characterized in that, The steps for establishing a dynamic evolution prediction model for reservoir permeability specifically involve establishing the relationship between porosity and permeability based on the Carman-Kozeny equation, and introducing the dissolution fraction of plugging material as an intermediate variable, whereby the dissolution fraction of plugging material is the ratio of the mass of dissolved plugging material to the total mass of the initial plugging material.

8. The method according to claim 7, characterized in that, The reservoir permeability dynamic evolution prediction model is trained using a neural network algorithm. The neural network algorithm includes an input layer, two hidden layers, and an output layer. The input parameters include the dissolution fraction of plugging material, the initial permeability, and the porosity. The output parameter is the current permeability. The neural network algorithm is trained using a backpropagation algorithm.

9. The method according to claim 8, characterized in that, The steps of solving the multiphysics coupling equations using operator splitting technology and adaptive time step control strategy are as follows: the multiphysics coupling problem is decomposed into a fluid flow subproblem, a chemical reaction subproblem, and a concentration diffusion subproblem using operator splitting technology, and each subproblem is solved sequentially in each time step.

10. The method according to claim 9, characterized in that, An implicit iterative scheme is introduced to improve the stability of time integration. The implicit iterative scheme adopts the Crank-Nicolson scheme and sets an adaptive time step control strategy. The relative error between the solutions of two adjacent time steps is calculated. When the relative error is greater than 0.01, the time step is reduced to 0.5 times the original size. When the relative error is less than 0.001, the time step is increased to 1.2 times the original size.

Images