Proportioning optimization method for efficiently balancing chemical oil displacement system

Abstract

The invention discloses a proportioning optimization method for efficiently balancing a chemical oil displacement system. The proportioning optimization method comprises the following steps: S1, collecting and sorting experiment and field data of the existing chemical oil displacement system; the oil displacement effect can be effectively improved by continuously optimizing the ratio of the chemical agents through experiments and numerical simulation, so that the recovery efficiency of an oil field is improved, the oil-water interface characteristics can be improved, the fluid viscosity can be reduced, the oil fluidity can be improved through the reasonable ratio of the chemical agents, and by analyzing the cost benefits of different ratios, the oil displacement effect can be effectively improved. The chemical agent combination with higher economical efficiency can be selected, the expenditure in the chemical oil displacement process is reduced, the oil displacement effect and cost are comprehensively considered, the resource utilization efficiency of the chemical oil displacement system is improved by optimizing the proportion, and it can be ensured that the selected proportion is not only theoretically effective but also theoretically effective through laboratory verification and on-site small-scale tests. And the method also has practical significance in practical application.

Description

Technical Field

[0001] The present invention relates to the technical field of oilfield development technology, and more particularly to a method for optimizing the proportions of a highly efficient and balanced chemical flooding system. Background Art

[0002] As oil fields gradually enter the high water content and ultra-high water content period, the difficulty of stable production gradually increases, development contradictions become increasingly prominent, the difficulty of exploring new reserves increases, and the cost increases. Therefore, further improving the recovery rate of proven and developed reserves has become a very urgent task. Chemical flooding system refers to a technical means of increasing crude oil recovery rate by injecting specific chemical substances in oil field development. Chemical flooding is usually used as a method to enhance oil and gas recovery rate. It is mainly suitable for difficult-to-recover oil reservoirs such as low permeability, heavy oil, and complex formations. It has far-reaching strategic significance for the sustained and stable production of oil fields.

[0003] However, there are some problems with the existing chemical flooding system's proportioning method. Unreasonable proportioning will lead to the inability of the flooding agent to fully play its role to a certain extent, resulting in a decrease in recovery rate and an inability to effectively extract more crude oil from the reservoir. Unreasonable proportioning may lead to weak interaction between liquid phases and incomplete oil-water separation, thereby affecting subsequent processing efficiency. Inappropriate chemical flooding agent ratios may lead to reduced viscosity and fluidity of the injected fluid, increased injection costs and reduced driving efficiency. Mismatches between chemical agents may lead to reduced reaction efficiency and inability to maintain long-term effectiveness, affecting the sustainability of the overall flooding operation. Moreover, unbalanced proportions may also lead to the use of some unnecessary chemicals, increase the amount and difficulty of waste treatment, and may cause environmental pollution risks. Therefore, a proportioning optimization method for an efficient and balanced chemical flooding system is proposed to solve the above problems. Summary of the Invention

[0004] The present invention overcomes the deficiencies in the prior art and provides a method for optimizing the ratio of a highly efficient and balanced chemical flooding system.

[0005] The purpose of the present invention is achieved through the following technical solutions.

[0006] A method for optimizing the ratio of a highly efficient and balanced chemical flooding system, comprising the following steps:

[0007] S1. Collect and organize experimental and field data of chemical flooding system;

[0008] S2. Based on the data collected in S1, multiple chemical flooding ratio data are proposed, and the flooding effects of the chemical flooding ratio data are evaluated respectively. When the evaluation results of the chemical flooding ratio numbers meet the set standards, the process proceeds to S3 for simulation testing. When the evaluation results of the chemical flooding ratio numbers do not meet the set standards, problems existing in the current chemical flooding ratio are determined, the chemical flooding ratio data plan is modified, and the modified chemical flooding ratio data plan is re-evaluated.

[0009] S3. Use numerical simulation tools to establish the relationship between oil wells, fluids, and chemicals to test the chemical sweep efficiency of the chemical flooding ratio data obtained in S2, and determine whether the tested ratio is qualified based on the chemical sweep efficiency;

[0010] S4. Conduct multi-factor experimental design on the chemical flooding ratio data that have passed the test in S3;

[0011] S5. Conduct small-scale experiments in the laboratory, gradually test chemical flooding ratios under different configurations based on the model and experimental design established in S4, and select the optimal chemical flooding ratio;

[0012] S6. Using an optimization algorithm to refine and solve the optimal chemical flooding ratio data obtained in S5 to maximize the flooding effect and obtain refined chemical flooding ratio data;

[0013] S7, conducting a small-scale field test on the chemical flooding ratio data obtained in S6, and adjusting and optimizing the chemical flooding ratio data based on the field experimental data;

[0014] S8. Apply the chemical flooding ratio data obtained in S7 to the on-site oil production process, and adjust and optimize the chemical flooding ratio data according to the monitoring data.

[0015] The specific steps for organizing the experimental and field data of the chemical flooding system in S1 include:

[0016] S11. Clean all collected data.

[0017] Based on the mean interpolation method to fill the missing data values, the interpolation formula is expressed as,

[0018]

[0019] Among them, N is the number of samples in a data column, x i is the nth sample in the data column, and n≤N;

[0020] S12, detect outlier data based on 3σ method,

[0021] Get the mean μ of the data, expressed as

[0022] Calculate the data standard deviation σ, expressed as

[0023] based on The original data is transformed, where Z represents the standardized data after transformation and obeys the standard normal distribution;

[0024] Construct a normal distribution graph and set the intervals (μ-σ, μ+σ), (μ-2σ, μ+2σ) and (μ-3σ, μ+3σ). Values ​​outside the interval (μ-3σ, μ+3σ) are regarded as outlier data and are eliminated.

[0025] The criteria for evaluating the oil displacement effect under the chemical flooding ratio data in the collected data in S2 include:

[0026] Quantitative analysis of single indicators to detect whether the interfacial tension (IFT) has dropped to an ultra-low value <1×10 -2 mN / m, to judge whether the performance of the surfactant meets the standards;

[0027] Analyze the solution viscosity and check whether the mobility ratio of the polymer solution viscosity to the crude oil viscosity is close to 1. If the mobility ratio is less than 1, it meets the standard. If the mobility ratio is greater than 1, change the polymer concentration and re-test the mobility ratio.

[0028] Calculate the recovery rate increase value, calculate the recovery rate increase under the existing ratio through core test data, simulate the oil displacement process according to the existing ratio parameters through electronic plug-ins, compare the error between the simulated recovery rate and the actual data, and select data with data error <5% as reliable data.

[0029] Problems identified in S2 with the current chemical flooding mix include:

[0030] Calculate the amount of chemical adsorption through core test data and evaluate the effect of chemical adsorption on oil displacement efficiency;

[0031] Based on the emulsion stability data, the bottle test method is used to ensure that the stratification time is less than 24 hours. If it is greater than 24 hours, it is judged that the emulsification risk may cause wellbore blockage or difficulty in producing fluid treatment;

[0032] Analyze long-term stability issues by analyzing chemical degradation rates.

[0033] The specific steps in S3 to establish the relationship between oil wells, fluids, and chemicals based on numerical simulation tools include:

[0034] S31, use the STARS module in CMG software to simulate chemical flooding and build a reservoir geological model.

[0035] Establish a grid system to display homogeneous or layered reservoirs through Cartesian structured grids and complex fault or fracture reservoirs through corner point unstructured grids;

[0036] Set the grid scale, design the horizontal grid size to be 10-50m, and the vertical layering to be 0.5-2m;

[0037] Assign values ​​to the attribute field, generate heterogeneous distribution through geostatistics to assign values ​​to the permeability field, and assign values ​​to porosity and saturation based on interpolation of well logging and core data;

[0038] S32. Define fluid and chemical properties.

[0039] Analyze crude oil properties using pseudo-component models and viscosity-temperature curves, and construct a fluid model based on the effects of formation water salinity and ion composition on chemical agents.

[0040] The concentration-IFT relationship is input through the interfacial tension curve (IFT), and the surfactant properties are analyzed based on the Langmuir model parameters. The polymer viscosity-concentration relationship is analyzed using the power law model, and the polymer shear thinning effect is analyzed using the Carreau model parameters. The viscosity attenuation coefficient is adjusted according to the salinity. A chemical agent model is constructed based on the concentration-IFT relationship, surfactant properties, polymer viscosity-concentration relationship, polymer shear thinning effect, and viscosity attenuation coefficient.

[0041] Configure chemical flooding parameters and design an injection plan based on slug design, including pre-slug, main slug, protection slug, and injection rate. Simulate the effect of oil-water emulsion viscosity changes on seepage by increasing the water-phase viscosity of the polymer, reducing the mobility ratio, and reducing the IFT of the surfactant to increase the capillary number. Define a rock surface adsorption model, test the polymer hydrolysis rate at high temperatures, analyze the dynamic interaction of the chemical agents, and select the polymer type, molecular weight, and concentration based on the data on adsorption and high-temperature effects during the dynamic interaction of the chemical agents.

[0042] S33, initializing the model based on the reservoir pressure and saturation field, using a fixed liquid production rate or a fixed bottomhole pressure in the production well, and a fixed injection rate or a fixed wellhead pressure in the injection well as boundary conditions;

[0043] S34, perform numerical simulation,

[0044] Use historical production data to calibrate the model for historical matching, with a control error of <10%;

[0045] Run the injection scheme designed in S32, including the pre-slug, main slug, protection slug, and injection rate, based on the chemical flooding ratio data, and perform predictive simulation based on the dynamic response output by the injection scheme;

[0046] The recovery rate improvement, chemical migration front, and water cut change in the production well are set as key output parameters; the chemical sweep efficiency is obtained through streamline visualization analysis.

[0047] The specific steps of multi-factor experimental design in S4 are as follows:

[0048] S41. Collect geological characteristics of the oil field under study and determine the fluid properties in the oil field;

[0049] S42. Select experimental equipment. According to the scale and conditions of the actual oil field, select the experimental scale to simulate the fluid flow in the actual oil field and obtain the actual oil displacement effect.

[0050] S43. Use computational fluid dynamics models to simulate the flow of fluids in oil fields and obtain data on the interaction between fluids and porous media;

[0051] S44. Use response surface methodology to systematically evaluate the impact of different factors. Design a control group and an experimental group based on actual oilfield conditions, and compare the effects of the control group and the experimental group under different conditions.

[0052] S45. Collect data during the experiment, compare the collected data with actual oil field data, evaluate the accuracy of the experimental conditions, and dynamically adjust the experimental conditions based on the actual oil field data.

[0053] The specific steps of S5 include:

[0054] S51, setting a concentration gradient, setting parameters for surfactant, polymer, and base concentrations;

[0055] Surfactants, based on the critical micelle concentration (CMC) setting range, with a step size of 0.2%;

[0056] For polymer, the target viscosity was matched to the crude oil viscosity, and the concentration range was determined to be 500–2000 ppm with a step size of 500 ppm.

[0057] Alkali concentration, set based on pH requirements and precipitation risk;

[0058] S52, set interfacial tension IFT < 10 -2 mN / m, the solution viscosity matches the crude oil viscosity and the mobility ratio ≈ 1, the core recovery factor is improved by ≥15% during the chemical flooding stage, and the adsorption loss ensures that the polymer adsorption amount is ≤1.5mg / g as performance indicators for data collection;

[0059] S53. Calculate cost parameters based on the chemical cost per ton of oil, and set the cost parameters as economic indicators.

[0060] Cost = (surfactant dosage × unit price + polymer dosage × unit price) / oil addition amount (tons);

[0061] S54. Construct a multiple regression model. The multiple regression model is expressed as:

[0062] Improved recovery factor = a·C s +b·C p +c·C a +d·C s ·C p +∈,

[0063] Among them C s 、C p 、C a are surfactant, polymer, and alkali concentrations, respectively, and a, b, c, and d are regression coefficients;

[0064] S54. Set the optimal chemical flooding ratio as the optimization target, use chemical concentration, temperature, salinity and core permeability as input features, and output recovery enhancement, viscosity and IFT based on the neural network as output targets to obtain the optimal flooding ratio.

[0065] The optimization algorithm in S6 is a genetic algorithm. The specific steps include:

[0066] Determine the oil recovery effect indicators that need to be maximized,

[0067] And mathematically transform it into the objective function,

[0068] Determine the concentration range of the ingredients that need to be optimized,

[0069] Each possible ratio scheme is encoded as a chromosome, an initial population is randomly generated and the fitness of each individual is calculated. Individuals are selected based on the fitness, some individuals are selected for crossover operation to generate new individuals, certain values ​​in the individual genes are randomly changed to ensure the diversity of the population and the search range, the newly generated individuals are merged with the original individuals, and a new population is selected using fitness evaluation. After reaching the preset number of generations, the fitness value no longer increases significantly, or other termination conditions are met, the iteration ends, and the individual with the highest fitness is selected as the final optimized ratio scheme. The oil recovery effect value corresponding to the ratio is output, and the optimal ratio obtained is verified under laboratory or field conditions.

[0070] The optimization algorithm in S6 is a method that further refines the optimal ratio by combining particle swarm optimization. The specific steps include:

[0071] Determine the oil displacement effect index that needs to be maximized and mathematically transform it into an objective function to determine the concentration range of the components that need to be optimized, including surfactant, polymer and salt concentration data;

[0072] Randomly generate an initial particle swarm, each particle represents a possible ratio scheme, determine the size of the particle swarm and the parameters of the PSO, calculate the fitness value of each particle, and update it according to the particle's speed and current fitness;

[0073] If the fitness of the current particle is better than its historical optimal fitness, its individual optimal position is updated. If the fitness of the current particle is better than the overall optimal fitness in the entire group, the global optimal position is updated. The particle position and speed are updated, the fitness is evaluated, and the individual and global optimal solutions are updated iteratively until the termination condition is met.

[0074] The ratio corresponding to the particles with the highest fitness is selected as the optimal chemical flooding ratio, and the corresponding flooding effect is recorded. The obtained optimal ratio is verified under laboratory or field conditions.

[0075] The specific steps for adjusting and optimizing the chemical flooding ratio in S7 and S8 include:

[0076] A1. Injection-end data monitoring: Using online sensors to detect surfactant and polymer concentrations in the injection fluid, the error range is determined to be (-5%, +5%). Pressure gauges and flow meters are used to monitor injection pressure and flow changes at the wellhead in real time.

[0077] Combined with chromatographic analysis to detect chemical breakthrough time, analyze the produced liquid composition, record the water content and oil production every 8 hours to analyze the displacement dynamics, and track the scope of chemical impact through tracer monitoring;

[0078] A2. Remove outliers from the collected data and unify the timestamps, align injection and production data, adjust the adsorption coefficient and permeability field parameters based on real-time data, calibrate the numerical model, update the training model online, and predict the future dynamics of the monitoring data in real time;

[0079] A3. Make dynamic ratio adjustments.

[0080] When the surfactant concentration in the produced fluid is greater than 0.1%, reduce the surfactant injection concentration by 10%;

[0081] When the wellhead pressure continues to rise by more than 10%, a 0.05PV gel slug is injected as a profile control agent to dilute the chemical concentration;

[0082] Increase polymer concentration or adjust slug size when water cut decreases and stagnates for 7 consecutive days;

[0083] Replace low-cost chemicals when economic costs exceed budget, i.e., when the cost per ton of oil exceeds the threshold;

[0084] A4. Algorithm-driven optimization: Based on model predictive control (MPC), we predict oil production for the next 30 days and optimize chemical injection concentration for the next seven days. We define a reward function based on reinforcement learning (RL), dynamically adjust the ratio through DQN algorithm training, and achieve iterative optimization of the oil displacement ratio.

[0085] The beneficial effects of the present invention are:

[0086] This proposal provides a method for optimizing the chemical ratio of an efficient and balanced chemical flooding system. Through experiments and numerical simulations, the chemical ratio is continuously optimized, effectively improving flooding efficiency and thus increasing oilfield recovery. A reasonable chemical ratio can improve oil-water interface properties, reduce fluid viscosity, and enhance oil mobility. By analyzing the cost-effectiveness of different ratios, more economical chemical combinations can be selected, reducing costs during the chemical flooding process. By comprehensively considering flooding efficiency and costs, the resource utilization efficiency of the chemical flooding system can be improved by optimizing the ratio. Furthermore, through laboratory validation and small-scale field trials, the selected ratio is ensured to be not only theoretically effective but also practically meaningful. Based on feedback from field monitoring, dynamic adjustments can be made in real time, enabling the ratio to respond promptly to changing field conditions and enhance effectiveness. Numerical simulation tools and experimental design provide theoretical support for the optimization process, helping to predict and verify performance under different ratios and improving decision-making efficiency and accuracy. Furthermore, optimizing the ratio can reduce potential negative environmental impacts during chemical flooding, and selecting environmentally friendly chemicals can help mitigate ecological risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 It is a step diagram of the present invention. DETAILED DESCRIPTION

[0088] Example

[0089] like Figure 1 As shown, a method for optimizing the ratio of a highly efficient and balanced chemical flooding system comprises the following specific steps:

[0090] S1. Collect and organize experimental and field data of chemical flooding system;

[0091] S2. Based on the data collected in S1, multiple chemical flooding ratio data are proposed, and the flooding effects of the chemical flooding ratio data are evaluated respectively. When the evaluation results of the chemical flooding ratio numbers meet the set standards, the process proceeds to S3 for simulation testing. When the evaluation results of the chemical flooding ratio numbers do not meet the set standards, problems existing in the current chemical flooding ratio are determined, the chemical flooding ratio data plan is modified, and the modified chemical flooding ratio data plan is re-evaluated.

[0092] S3. Use numerical simulation tools to establish the relationship between oil wells, fluids, and chemicals to test the chemical sweep efficiency of the chemical flooding ratio data obtained in S2, and determine whether the tested ratio is qualified based on the chemical sweep efficiency;

[0093] S4. Conduct multi-factor experimental design on the chemical flooding ratio data that have passed the test in S3;

[0094] S5. Conduct small-scale experiments in the laboratory, gradually test chemical flooding ratios under different configurations based on the model and experimental design established in S4, and select the optimal chemical flooding ratio;

[0095] S6. Using an optimization algorithm to refine and solve the optimal chemical flooding ratio data obtained in S5 to maximize the flooding effect and obtain refined chemical flooding ratio data;

[0096] S7, conducting a small-scale field test on the chemical flooding ratio data obtained in S6, and adjusting and optimizing the chemical flooding ratio data based on the field experimental data;

[0097] S8. Apply the chemical flooding ratio data obtained in S7 to the on-site oil production process, and adjust and optimize the chemical flooding ratio data according to the monitoring data.

[0098] Furthermore, the experimental data of the chemical flooding system in S1 include:

[0099] formulation parameters, surfactant type, polymer molecular weight, and base concentration;

[0100] Physical and chemical performance data, including interfacial tension (IFT) measured by a spinning drop interfacial tensiometer, viscosity at different shear rates measured by a rheometer, and emulsion stability data measured by observing the emulsion separation time using the bottle test method;

[0101] Record core permeability, displacement pressure and recovery enhancement data;

[0102] Organize the historical data of formula screening and adsorption loss in the collected data;

[0103] The field data of the chemical flooding system in S1 include:

[0104] Injection parameters, chemical concentration, injection rate and slug design data;

[0105] Dynamic response data, including oil production changes, water cut decline, and pressure response;

[0106] Reservoir characteristics, temperature, salinity and crude oil viscosity.

[0107] The specific steps for organizing the experimental and field data of the chemical flooding system in S1 include:

[0108] S11. Clean all collected data.

[0109] Based on the mean interpolation method to fill the missing data values, the interpolation formula is expressed as,

[0110]

[0111] Among them, N is the number of samples in a data column, x i is the nth sample in the data column, and n≤N;

[0112] S12, detect outlier data based on 3σ method,

[0113] Get the mean μ of the data, expressed as

[0114] Calculate the data standard deviation σ, expressed as

[0115] based on The original data is transformed, where Z represents the standardized data after transformation and obeys the standard normal distribution;

[0116] Construct a normal distribution graph and set the intervals (μ-σ, μ+σ), (μ-2σ, μ+2σ) and (μ-3σ, μ+3σ). Values ​​outside the interval (μ-3σ, μ+3σ) are regarded as outlier data and are eliminated.

[0117] The purpose of S1 is to collect and clean the collected data, eliminate outliers, and facilitate subsequent modeling and analysis based on the data.

[0118] S2 evaluates the data collected in S1.

[0119] The criteria for evaluating the oil displacement effect under the ratio in the collected data in S2 include:

[0120] Quantitative analysis of single indicators to detect whether the interfacial tension (IFT) has dropped to an ultra-low value <1×10 -2 mN / m, to judge whether the performance of the surfactant meets the standards;

[0121] Analyze the solution viscosity and check whether the mobility ratio of the polymer solution viscosity to the crude oil viscosity is close to 1. If the mobility ratio is less than 1, it meets the standard. If the mobility ratio is greater than 1, change the polymer concentration and re-test the mobility ratio.

[0122] Calculate the recovery increase using core test data to calculate the recovery increase under the existing mix ratio. Use an electronic plug-in to simulate the flooding process based on the existing mix parameters. Compare the simulated recovery error with the actual data, and select data with an error of <5% as reliable data. The electronic plug-in uses Eclipse software.

[0123] This step evaluates each mix ratio scheme and selects the ones that meet the requirements. The mix ratio schemes that do not meet the standards are analyzed to determine the problems with the current mix ratio.

[0124] The specific steps in S2 to determine the problems with the current ratio include:

[0125] The amount of chemical adsorption was calculated based on core experimental data to evaluate its impact on oil recovery efficiency.

[0126] Among them, chemicals with large adsorption capacity have poor oil recovery efficiency and are likely to have negative impacts on reservoirs and produced fluids. Therefore, chemicals with large adsorption capacity are generally not selected.

[0127] Based on the emulsion stability data, the bottle test method is used to make the stratification time less than 24 hours. If it is greater than 24 hours, it is judged that the wellbore is blocked or the produced fluid treatment is difficult due to the emulsification risk.

[0128] Analyze long-term stability issues by analyzing chemical degradation rates.

[0129] When the problems are identified, such as high adsorption, severe emulsification, and poor long-term stability, the chemical flooding ratio modification plan will select a ratio with relatively low adsorption, less severe emulsification, and good long-term stability. This stage is to adjust the ratio that does not meet the standards by combining static experiments, such as adsorption experiments, emulsification experiments, and long-term stability experiments.

[0130] In S3, numerical simulation is used to test the ratio scheme selected in S2, and the appropriate ratio scheme is selected through the chemical agent sweep efficiency.

[0131] The specific steps in S3 to establish the relationship between oil wells, fluids, and chemicals based on numerical simulation tools include:

[0132] S31, use the STARS module in CMG software to simulate chemical flooding and build a reservoir geological model.

[0133] Establish a grid system to display homogeneous or layered reservoirs through Cartesian structured grids and complex fault or fracture reservoirs through corner point unstructured grids;

[0134] Set the grid scale, design the horizontal grid size to be 10-50m, and the vertical layering to be 0.5-2m;

[0135] Assigning values ​​to the attribute field, generating heterogeneous distribution through geostatistics to assign values ​​to the permeability field, and combining well logging and core data interpolation to assign values ​​to porosity and saturation;

[0136] S32. Define fluid and chemical properties.

[0137] Analyze crude oil properties by combining pseudo-component models with viscosity-temperature curves, and construct a fluid model by considering the effects of formation water salinity and ion composition on chemical agents;

[0138] Combine the interfacial tension curve (IFT) with the input concentration-IFT relationship and analyze the surfactant properties based on the Langmuir model parameters. Use the power law model to analyze the polymer viscosity-concentration relationship and the Carreau model parameters to analyze the polymer shear thinning effect. Adjust the viscosity attenuation coefficient according to the salinity to construct a chemical agent model.

[0139] Configure chemical flooding parameters and design an injection plan based on the slug design, including the pre-slug, main slug, and protection slug, as well as the injection rate. Simulate the effect of oil-water emulsion viscosity changes on seepage by increasing the water-phase viscosity of the polymer, reducing the mobility ratio, and reducing the IFT of the surfactant to increase the capillary number. Define a rock surface adsorption model, test the hydrolysis rate of the polymer at high temperatures, and analyze the dynamic interactions of the chemical agents.

[0140] S33, initializing the model based on the reservoir pressure and saturation field, using a fixed liquid production rate or a fixed bottomhole pressure in the production well, and a fixed injection rate or a fixed wellhead pressure in the injection well as boundary conditions;

[0141] S34, perform numerical simulation,

[0142] Use historical production data to calibrate the model for historical matching, with a control error of <10%;

[0143] Run the injection scheme designed in S32, including the pre-slug, main slug, protection slug, and injection rate, based on the chemical flooding ratio data, and perform predictive simulation based on the dynamic response output by the injection scheme;

[0144] The recovery rate improvement, chemical migration front, and water cut change in the production well are set as key output parameters; the chemical sweep efficiency is obtained through streamline visualization analysis.

[0145] Among them, S34 uses numerical simulation to output numerical values ​​and graphs. The scope of the impact can be intuitively seen through the streamline diagram, and the specific impact efficiency can be quantified through the numerical values. The larger the impact range, the greater the impact efficiency and the better the oil displacement effect.

[0146] The specific steps of multi-factor experimental design in S4 are as follows:

[0147] S41. Collect geological characteristics of the oil field under study and determine the fluid properties in the oil field;

[0148] S42. Select experimental equipment. According to the scale and conditions of the actual oil field, select an appropriate experimental scale to simulate the fluid flow in the actual oil field and obtain the actual oil displacement effect.

[0149] S43. Use computational fluid dynamics models to simulate the flow of fluids in oil fields and study the interaction between fluids and porous media;

[0150] S44. Use response surface methodology to systematically evaluate the impact of different factors. Design a control group and an experimental group based on actual oilfield conditions, and compare the effects of the control group and the experimental group under different conditions.

[0151] S45. Collect data during the experiment, compare the collected data with actual oil field data, evaluate the accuracy of the experimental conditions, and dynamically adjust the experimental conditions based on the actual oil field data.

[0152] Furthermore, the geological characteristics in S41 include rock type, porosity, permeability and physical and chemical properties of rocks, and the physical and chemical properties of rocks include mineral composition, saturated water and oil-gas ratio;

[0153] Fluid properties include the viscosity, density, surface tension, and relative wettability of the crude oil and the composition and pH of the well water;

[0154] Experimental equipment includes high-pressure and high-temperature vessels, micro-chromatographic equipment, and dynamic models;

[0155] The data collected in S45 include oil-water separation efficiency, flow rate and recovery factor.

[0156] Preferably, in this embodiment, S41 and S42 select a suitable experimental scale according to the scale and conditions of the actual oil field to simulate the fluid flow in the actual oil field. The specific steps include:

[0157] Combined with the scale of the oil field, core experiments and numerical simulations are carried out based on small oil fields, physical simulations and numerical simulations are carried out based on large oil fields, core experiments are mainly used for homogeneous reservoirs, and physical simulations are carried out based on fractured heterogeneous reservoirs.

[0158] The core size is scaled with the reservoir thickness, the actual pore network is replicated through CT scanning of the core, the flow rate is adjusted proportionally to maintain the flow state consistent, the injection pressure or interfacial tension is adjusted to make the ratio of displacement force to capillary force consistent, and the injected pore volume multiple PV number is matched with the oil field development time.

[0159] The core flooding experiment steps are as follows:

[0160] a. Clean and dry the core, measure porosity and permeability, inject degassed crude oil to irreducible water saturation, record water flooding recovery, inject chemicals to monitor pressure and chemical flooding recovery, calculate chemical flooding oil increment, and analyze adsorption loss.

[0161] b. Construct a three-dimensional sandfill model to set up the injection and production well pattern. Simulate heterogeneity based on the alternating distribution of high-permeability strips and low-permeability zones. Use tracers to detect the volume swept by the chemical agent, and use a pressure sensor array to monitor the dominant direction of the flow channel.

[0162] The temperature and pressure in the experiment were set to match the actual oilfield environment, ensuring that the concentration range of surfactants and polymers was close to the concentrations used in actual oilfield applications. The steps to obtain the actual oil displacement effect included:

[0163] 1) Collect temperature and pressure data:

[0164] a. Obtain reservoir temperature through well logging data or reservoir engineering reports:

[0165] b. Determine formation pressure based on bottom hole pressure test or reservoir simulation results;

[0166] c. Analyze fluid properties based on crude oil viscosity and formation water salinity data;

[0167] d. Analyze the typical concentration ranges of surfactants and polymers in similar oil fields based on historical chemical data.

[0168] 2) Load the core into a high-temperature, high-pressure core displacement holder, apply confining pressure to the reservoir pressure based on pressure sensor monitoring, heat to the target temperature with a precision temperature control system, and allow the core to stabilize in thermal equilibrium for 2 hours or more before injecting preheated formation water to simulate the initial conditions of the reservoir;

[0169] 3) Surfactants and polymers were formulated according to the actual formulation of the oil field. Brine matching the salinity of the formation water was used. Natural cores with a permeability (200 mD) and porosity (20%) were selected to approximate the target reservoir. Reservoirs with fractures were subjected to heterogeneous treatment. The cores were washed with toluene and methanol to remove residual oil. After drying, the cores were vacuum-saturated with formation water, the pore volume was measured, and the degassed crude oil was saturated to the irreducible water saturation.

[0170] 4) Inject formation water to a water cut of 95% during the water flooding phase and record the base recovery rate. In the chemical flooding phase, inject a chemical slug, simulating the field injection rate using the flow rate. Continue to inject formation water during the subsequent water flooding phase and monitor the recovery rate until it stabilizes.

[0171] 5) Monitor and collect pressure gradient parameter data to evaluate the effect of chemicals on flow resistance. Calculate the incremental oil production from chemical flooding using recovery enhancement parameter data. Determine chemical adsorption parameter data using material balance method or chromatographic analysis. Verify performance indicators, including interfacial tension (IFT) and mobility control capability, by measuring and calculating the mobility ratio before and after flooding using a spinning drop tensiometer at high temperature.

[0172] 6) Set up multiple concentration gradient experiments to form different cross-combinations of surfactants and polymers, analyze the responses of recovery factor, IFT, and viscosity, and determine the optimal concentration range.

[0173] S43 uses computational fluid dynamics models to simulate the flow of fluids in oil fields and study the interaction between fluids and porous media. The specific steps include:

[0174] 1) Simulate the flow characteristics of oil, water, and gas in porous media, analyze the oil displacement efficiency or chemical migration effect as the research goal, obtain core CT scan data, use Avizo or Simpleware to perform 3D reconstruction, then export it to STL format and import it into CFD software to generate geometry and reconstruct the actual pore structure;

[0175] 2) Set up the physical model

[0176] a. Porous media properties: Calculate permeability using isotropic or anisotropic tensors based on Darcy's law;

[0177] b. Porosity: Calculate porosity data using CT data;

[0178] c. Fluid properties: The oil phase has a density of 800 kg / m 3 , viscosity 5mPa·s; water phase property is density 1000kg / m 3 , viscosity 1mPa·s;

[0179] d. Construct a multiphase flow model: Combine the VOF method to track the phase interface and use the Eulerian-Eulerian model to process the dispersed phase.

[0180] Set the boundary conditions as follows: set constant pressure or flow rate data at the inlet, set static pressure conditions at the outlet, and set a no-slip boundary on the wall;

[0181] 3) Set up a pressure-based solver to handle pressure-velocity coupling using the SIMPLE algorithm. Discretize the momentum equations in a second-order upwind format, setting the convergence criteria to residuals < 1e-6 for the continuity equation and < 1e-8 for the energy equation.

[0182] 4) Modeling the fluid-pore interaction, adding a surfactant adsorption source term to realize the adsorption effect, expressed as S ads =k ads C (1-θ)-k des ·θ

[0183] Where C is the solution concentration and θ is the surface coverage;

[0184] Describe the polymer / surfactant concentration distribution based on coupled mass transfer equations based on chemical flooding simulation analysis;

[0185] 5) Identify hyperpermeability channels and dead-end pores through velocity field distribution, assess flow resistance based on pressure gradient, analyze displacement efficiency using phase saturation distribution, and compare recovery factors with core flooding experiments, controlling the error to <5%. Observe the flow within the pores using microfluidics experiments for experimental verification.

[0186] S44 uses the response surface methodology to systematically evaluate the impact of different factors. Experimental groups, including control and experimental groups, are designed based on actual oilfield conditions. The impacts are compared under different conditions. Specifically:

[0187] 1) Screening of factors that significantly affect oil displacement, including surfactant concentration C s , polymer concentration C p , alkali concentration C a and temperature T, and selected oil displacement efficiency ΔR, interfacial tension IFT, and solution viscosity μ as technical indicators;

[0188] 2) Construct a second-order polynomial model, expressed as

[0189]

[0190] The coefficients were obtained by fitting the experimental data using the least squares method, and the insignificant terms were eliminated through stepwise regression to ensure p>0.05 to simplify the model.

[0191] 3) Set the maximization of oil recovery efficiency, minimization of IFT, and minimization of cost per ton of oil as optimization objectives, set weights through the expectation function method, calculate the comprehensive score, and use the Pareto front to screen the non-inferior solution set.

[0192] The experiment was repeated three times for the optimal solution to confirm that the fluctuation of oil displacement efficiency ΔR was less than 5%. After the parameters were input into Eclipse based on numerical simulation verification, the error of the predicted well group effect was controlled to be less than 10%.

[0193] During the small-scale experiment of S5, the parameters monitored in real time included flow characteristics, production and oil-water separation. Statistical tools were used to analyze the experimental data to determine the optimal ratio. The statistical tool selected was variance analysis.

[0194] The specific steps of S5 include:

[0195] S51, setting a concentration gradient, setting parameters for surfactant, polymer, and base concentrations;

[0196] Surfactants, based on the critical micelle concentration (CMC) setting range, with a step size of 0.2%;

[0197] For polymer, the target viscosity was matched to the crude oil viscosity, and the concentration range was determined to be 500–2000 ppm with a step size of 500 ppm.

[0198] Alkali concentration, set based on pH requirements and precipitation risk;

[0199] S52, set interfacial tension IFT < 10 -2 mN / m, the solution viscosity matches the crude oil viscosity and the mobility ratio ≈ 1, the core recovery factor is improved by ≥15% during the chemical flooding stage, and the adsorption loss ensures that the polymer adsorption amount is ≤1.5mg / g as performance indicators for data collection;

[0200] S53. Calculate cost parameters based on the chemical cost per ton of oil, and set the cost parameters as economic indicators.

[0201] Cost = (surfactant dosage × unit price + polymer dosage × unit price) / oil addition amount (tons);

[0202] S54. Construct a multiple regression model. The multiple regression model is expressed as:

[0203] Improved recovery factor = a·C s +b·C p +c·C a +d·C s ·C p +∈,

[0204] Among them C s 、C p 、C a are surfactant, polymer, and alkali concentrations, respectively, and a, b, c, and d are regression coefficients;

[0205] S54. Set the optimal chemical flooding ratio as the optimization target, use chemical concentration, temperature, salinity and core permeability as input features, and output recovery enhancement, viscosity and IFT based on the neural network as output targets to obtain the optimal flooding ratio.

[0206] The optimization algorithm in S6 is a genetic algorithm. The specific steps include:

[0207] Determine the oil recovery effect indicators that need to be maximized (for example, recovery rate, oil-water separation efficiency, etc.),

[0208] And mathematically transform it into the objective function,

[0209] Determine the concentration range of the ingredients that need to be optimized (such as surfactants, polymers, salts, etc.),

[0210] Each possible ratio scheme is encoded as a chromosome, an initial population is randomly generated and the fitness of each individual is calculated. Individuals are selected based on the fitness, some individuals are selected for crossover operation to generate new individuals, certain values ​​in the individual genes are randomly changed to ensure the diversity of the population and the search range, the newly generated individuals are merged with the original individuals, and a new population is selected using fitness evaluation. After reaching the preset number of generations, the fitness value no longer increases significantly, or other termination conditions are met, the iteration ends, and the individual with the highest fitness is selected as the final optimized ratio scheme. The oil recovery effect value corresponding to the ratio is output, and the optimal ratio obtained is verified under laboratory or field conditions.

[0211] Preferably, in this embodiment, the displacement ratio is optimized based on a genetic algorithm, and the objective function is constructed as follows:

[0212]

[0213] Where ΔR is the recovery factor improvement value, which is used to directly reflect the oil displacement efficiency. IFT is the interfacial tension mN / m, which needs to be minimized. C ton is the cost per ton of oil, which needs to be minimized, ω1, ω2, ω3 are weight coefficients, IFT base is the base interfacial tension, C threshold is the economic threshold;

[0214] The concentration range of the ingredients that need to be optimized specifically includes:

[0215] 1) Determine the surfactant concentration range based on the critical micelle concentration or preliminary experimental results;

[0216] 2) Determine the polymer concentration based on solution viscosity requirements and economic costs;

[0217] 3) Determine salt concentration by matching formation water salinity;

[0218] Use fitness evaluation to select a new population. The specific steps are:

[0219] 1) Based on the roulette wheel selection method, the selection probability is allocated according to the fitness ratio, and the fitness ratio of each individual is calculated as The cumulative probability distribution is expressed as Select individuals by generating random numbers;

[0220] 2) Generate offspring through single-point or uniform crossover, and update the population by slightly perturbing the gene values ​​according to probability;

[0221] The fitness acquisition steps include:

[0222] 1) Measure the recovery enhancement and interfacial tension in core flooding experiments, and calculate the cost per ton of oil based on the market price of chemicals for cost accounting;

[0223] 2) Use Eclipse to build a reservoir model, input ratio parameters to simulate recovery and pressure response, calculate the time of a single simulation, train the neural network (ANN) based on historical data, and predict the fitness of the new ratio;

[0224] 3) Efficiency optimization: Run multiple experiments or simulation tasks simultaneously for parallel computing, and reduce the sampling density in areas where fitness changes slowly to achieve adaptive sampling.

[0225] The optimization algorithm in S6 is a method that further refines the optimal ratio by combining particle swarm optimization. The specific steps include:

[0226] Identify the oil recovery performance indicators that need to be maximized (e.g., recovery factor, oil-water separation efficiency, etc.), mathematically transform them into objective functions, and determine the concentration range of the components that need to be optimized, including surfactant, polymer, and salt concentration data;

[0227] Randomly generate an initial particle swarm, each particle represents a possible ratio scheme, determine the size of the particle swarm and the parameters of the PSO, calculate the fitness value of each particle, and update it according to the particle's speed and current fitness;

[0228] If the fitness of the current particle is better than its historical optimal fitness, its individual optimal position is updated. If the fitness of the current particle is better than the overall optimal fitness in the entire group, the global optimal position is updated. The particle position and speed are updated, the fitness is evaluated, and the individual and global optimal solutions are updated iteratively until the termination condition is met.

[0229] The ratio corresponding to the particles with the highest fitness is selected as the optimal chemical flooding ratio, and the corresponding flooding effect is recorded. The obtained optimal ratio is verified under laboratory or field conditions.

[0230] Furthermore, this solution provides two solutions: genetic algorithm and particle swarm algorithm, which can be used selectively according to actual conditions, or they can be used in combination by optimizing the ratio with genetic algorithm and then optimizing with particle swarm algorithm.

[0231] The specific steps for adjusting and optimizing the chemical flooding ratio in S5 and S6 include:

[0232] A1. Injection-end data monitoring: Using online sensors to detect surfactant and polymer concentrations in the injection fluid, the error range is determined to be (-5%, +5%). Pressure gauges and flow meters are used to monitor injection pressure and flow changes at the wellhead in real time.

[0233] Combined with chromatographic analysis to detect chemical breakthrough time, analyze the produced liquid composition, record the water content and oil production every 8 hours to analyze the displacement dynamics, and track the scope of chemical impact through tracer monitoring;

[0234] A2. Remove outliers from the collected data and unify the timestamps, align injection and production data, adjust the adsorption coefficient and permeability field parameters based on real-time data, calibrate the numerical model, update the training model online, and predict the future dynamics of the monitoring data in real time;

[0235] A3. Make dynamic ratio adjustments.

[0236] When the surfactant concentration in the produced fluid is greater than 0.1%, reduce the surfactant injection concentration by 10%;

[0237] When the wellhead pressure continues to rise by more than 10%, a 0.05PV gel slug is injected as a profile control agent to dilute the chemical concentration;

[0238] Increase polymer concentration or adjust slug size when water cut decreases and stagnates for 7 consecutive days;

[0239] d. Replace low-cost chemicals when the economic cost exceeds the budget, i.e., the cost per ton of oil exceeds the threshold;

[0240] A4. Algorithm-driven optimization: Based on model predictive control (MPC), we predict oil production for the next 30 days and optimize chemical injection concentration for the next seven days. We define a reward function based on reinforcement learning (RL), dynamically adjust the ratio through DQN algorithm training, and achieve iterative optimization of the oil displacement ratio.

[0241] The embodiments of the present invention are described in detail above, but the contents described are only preferred embodiments of the present invention and should not be considered to limit the scope of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for optimizing the ratio of an efficient and balanced chemical flooding system, characterized in that: The specific steps include: S1. Collect and organize experimental and field data of chemical flooding system; S2. Based on the data collected in S1, multiple chemical flooding ratio data are proposed, and the flooding effects of the chemical flooding ratio data are evaluated respectively. When the evaluation results of the chemical flooding ratio numbers meet the set standards, the process proceeds to S3 for simulation testing. When the evaluation results of the chemical flooding ratio numbers do not meet the set standards, problems existing in the current chemical flooding ratio are determined, the chemical flooding ratio data plan is modified, and the modified chemical flooding ratio data plan is re-evaluated. S3. Use numerical simulation tools to establish the relationship between oil wells, fluids, and chemicals to test the chemical sweep efficiency of the chemical flooding ratio data obtained in S2, and determine whether the tested ratio is qualified based on the chemical sweep efficiency; S4. Conduct multi-factor experimental design on the chemical flooding ratio data that have passed the test in S3; S5. Conduct small-scale experiments in the laboratory, gradually test chemical flooding ratios under different configurations based on the model and experimental design established in S4, and select the optimal chemical flooding ratio; S6. Using an optimization algorithm to refine and solve the optimal chemical flooding ratio data obtained in S5 to maximize the flooding effect and obtain refined chemical flooding ratio data; S7, conducting a small-scale field test on the chemical flooding ratio data obtained in S6, and adjusting and optimizing the chemical flooding ratio data based on the field experimental data; S8. Apply the chemical flooding ratio data obtained in S7 to the on-site oil production process, and adjust and optimize the chemical flooding ratio data according to the monitoring data.

2. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The specific steps for organizing the experimental and field data of the chemical flooding system in S1 include: S11. Clean all collected data. Based on the mean interpolation method to fill the missing data values, the interpolation formula is expressed as, Among them, N is the number of samples in a data column, x i is the nth sample in the data column, and n≤N; S12, detect outlier data based on 3σ method, Get the mean μ of the data, expressed as Calculate the data standard deviation σ, expressed as based on The original data is transformed, where Z represents the standardized data after transformation and obeys the standard normal distribution; Construct a normal distribution graph and set the intervals (μ-σ, μ+σ), (μ-2σ, μ+2σ) and (μ-3σ, μ+3σ). Values ​​outside the interval (μ-3σ, μ+3σ) are regarded as outlier data and are eliminated.

3. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The criteria for evaluating the oil displacement effect under the chemical flooding ratio data in the collected data in S2 include: Quantitative analysis of single indicators to detect whether the interfacial tension (IFT) has dropped to an ultra-low value <1×10 -2 mN / m, to judge whether the performance of the surfactant meets the standards; Analyze the solution viscosity and check whether the mobility ratio of the polymer solution viscosity to the crude oil viscosity is close to 1. If the mobility ratio is less than 1, it meets the standard. If the mobility ratio is greater than 1, change the polymer concentration and re-test the mobility ratio. Calculate the recovery rate increase value, calculate the recovery rate increase under the existing ratio through core test data, simulate the oil displacement process according to the existing ratio parameters through electronic plug-ins, compare the error between the simulated recovery rate and the actual data, and select data with data error <5% as reliable data.

4. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: Problems identified in S2 with the current chemical flooding mix include: Calculate the amount of chemical adsorption through core test data and evaluate the effect of chemical adsorption on oil displacement efficiency; Based on the emulsion stability data, the bottle test method is used to ensure that the stratification time is less than 24 hours. If it is greater than 24 hours, it is judged that the emulsification risk may cause wellbore blockage or difficulty in producing fluid treatment; Analyze long-term stability issues by analyzing chemical degradation rates.

5. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The specific steps in S3 to establish the relationship between oil wells, fluids, and chemicals based on numerical simulation tools include: S31, use the STARS module in CMG software to simulate chemical flooding and build a reservoir geological model. Establish a grid system to display homogeneous or layered reservoirs through Cartesian structured grids and complex fault or fracture reservoirs through corner point unstructured grids; Set the grid scale, design the horizontal grid size to be 10-50m, and the vertical layering to be 0.5-2m; Assign values ​​to the attribute field, generate heterogeneous distribution through geostatistics to assign values ​​to the permeability field, and assign values ​​to porosity and saturation based on interpolation of well logging and core data; S32. Define fluid and chemical properties. Analyze crude oil properties using pseudo-component models and viscosity-temperature curves, and construct a fluid model based on the effects of formation water salinity and ion composition on chemical agents. The concentration-IFT relationship is input through the interfacial tension curve (IFT), and the surfactant properties are analyzed based on the Langmuir model parameters. The polymer viscosity-concentration relationship is analyzed using the power law model, and the polymer shear thinning effect is analyzed using the Carreau model parameters. The viscosity attenuation coefficient is adjusted according to the salinity. A chemical agent model is constructed based on the concentration-IFT relationship, surfactant properties, polymer viscosity-concentration relationship, polymer shear thinning effect, and viscosity attenuation coefficient. Configure chemical flooding parameters and design an injection plan based on slug design, including pre-slug, main slug, protection slug, and injection rate. Simulate the effect of oil-water emulsion viscosity changes on seepage by increasing the water-phase viscosity of the polymer, reducing the mobility ratio, and reducing the IFT of the surfactant to increase the capillary number. Define a rock surface adsorption model, test the polymer hydrolysis rate at high temperatures, analyze the dynamic interaction of the chemical agents, and select the polymer type, molecular weight, and concentration based on the data on adsorption and high-temperature effects during the dynamic interaction of the chemical agents. S33, initializing the model based on the reservoir pressure and saturation field, using a fixed liquid production rate or a fixed bottomhole pressure in the production well, and a fixed injection rate or a fixed wellhead pressure in the injection well as boundary conditions; S34, perform numerical simulation, Use historical production data to calibrate the model for historical matching, with a control error of <10%; Run the injection scheme designed in S32, including the pre-slug, main slug, protection slug, and injection rate, based on the chemical flooding ratio data, and perform predictive simulation based on the dynamic response output by the injection scheme; The recovery rate improvement, chemical migration front, and water cut change in the production well are set as key output parameters; the chemical sweep efficiency is obtained through streamline visualization analysis.

6. According to the method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, the specific steps of the multi-factor experimental design in S4 are as follows: S41. Collect geological characteristics of the oil field under study and determine the fluid properties in the oil field; S42. Select experimental equipment. According to the scale and conditions of the actual oil field, select the experimental scale to simulate the fluid flow in the actual oil field and obtain the actual oil displacement effect. S43. Use computational fluid dynamics models to simulate the flow of fluids in oil fields and obtain data on the interaction between fluids and porous media; S44. Use response surface methodology to systematically evaluate the impact of different factors. Design a control group and an experimental group based on actual oilfield conditions, and compare the effects of the control group and the experimental group under different conditions. S45. Collect data during the experiment, compare the collected data with actual oil field data, evaluate the accuracy of the experimental conditions, and dynamically adjust the experimental conditions based on the actual oil field data.

7. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The specific steps of S5 include: S51, setting a concentration gradient, setting parameters for surfactant, polymer, and base concentrations; Surfactants, based on the critical micelle concentration (CMC) setting range, with a step size of 0.2%; For polymer, the target viscosity was matched to the crude oil viscosity, and the concentration range was determined to be 500–2000 ppm with a step size of 500 ppm. Alkali concentration, set based on pH requirements and precipitation risk; S52, set interfacial tension IFT < 10 -2 mN / m, the solution viscosity matches the crude oil viscosity and the mobility ratio ≈ 1, the core recovery factor is improved by ≥15% during the chemical flooding stage, and the adsorption loss ensures that the polymer adsorption amount is ≤1.5mg / g as performance indicators for data collection; S53. Calculate cost parameters based on the chemical cost per ton of oil, and set the cost parameters as economic indicators. Cost = (surfactant dosage × unit price + polymer dosage × unit price) / oil addition amount (tons); S54. Construct a multiple regression model. The multiple regression model is expressed as: Improved recovery factor = a·C s +b·C p +c·C a +d·C s ·C p +∈, Among them C s 、C p 、C a are surfactant, polymer, and alkali concentrations, respectively, and a, b, c, and d are regression coefficients; S54. Set the optimal chemical flooding ratio as the optimization target, use chemical concentration, temperature, salinity and core permeability as input features, and output recovery enhancement, viscosity and IFT based on the neural network as output targets to obtain the optimal flooding ratio.

8. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The optimization algorithm in S6 is a genetic algorithm. The specific steps include: Determine the oil recovery effect indicators that need to be maximized, And mathematically transform it into the objective function, Determine the concentration range of the ingredients that need to be optimized, Each possible ratio scheme is encoded as a chromosome, an initial population is randomly generated and the fitness of each individual is calculated. Individuals are selected based on the fitness, some individuals are selected for crossover operation to generate new individuals, certain values ​​in the individual genes are randomly changed to ensure the diversity of the population and the search range, the newly generated individuals are merged with the original individuals, and a new population is selected using fitness evaluation. After reaching the preset number of generations, the fitness value no longer increases significantly, or other termination conditions are met, the iteration ends, and the individual with the highest fitness is selected as the final optimized ratio scheme. The oil recovery effect value corresponding to the ratio is output, and the optimal ratio obtained is verified under laboratory or field conditions.

9. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 1, characterized in that: The optimization algorithm in S6 is a method that further refines the optimal ratio by combining particle swarm optimization. The specific steps include: Determine the oil displacement effect index that needs to be maximized and mathematically transform it into an objective function to determine the concentration range of the components that need to be optimized, including surfactant, polymer and salt concentration data; Randomly generate an initial particle swarm, each particle represents a possible ratio scheme, determine the size of the particle swarm and the parameters of the PSO, calculate the fitness value of each particle, and update it according to the particle's speed and current fitness; If the fitness of the current particle is better than its historical optimal fitness, its individual optimal position is updated. If the fitness of the current particle is better than the overall optimal fitness in the entire group, the global optimal position is updated. The particle position and speed are updated, the fitness is evaluated, and the individual and global optimal solutions are updated iteratively until the termination condition is met. The ratio corresponding to the particles with the highest fitness is selected as the optimal chemical flooding ratio, and the corresponding flooding effect is recorded. The obtained optimal ratio is verified under laboratory or field conditions.

10. The method for optimizing the ratio of a high-efficiency balanced chemical flooding system according to claim 8, characterized in that: The specific steps for adjusting and optimizing the chemical flooding ratio in S7 and S8 include: A1. Injection-end data monitoring: Using online sensors to detect surfactant and polymer concentrations in the injection fluid, the error range is determined to be (-5%, +5%). Pressure gauges and flow meters are used to monitor injection pressure and flow changes at the wellhead in real time. Combined with chromatographic analysis to detect chemical breakthrough time, analyze the produced liquid composition, record the water content and oil production every 8 hours to analyze the displacement dynamics, and track the scope of chemical impact through tracer monitoring; A2. Remove outliers from the collected data and unify the timestamps, align injection and production data, adjust the adsorption coefficient and permeability field parameters based on real-time data, calibrate the numerical model, update the training model online, and predict the future dynamics of the monitoring data in real time; A3. Make dynamic ratio adjustments. When the surfactant concentration in the produced fluid is greater than 0.1%, reduce the surfactant injection concentration by 10%; When the wellhead pressure continues to rise by more than 10%, a 0.05PV gel slug is injected as a profile control agent to dilute the chemical concentration; Increase polymer concentration or adjust slug size when water cut decreases and stagnates for 7 consecutive days; Replace low-cost chemicals when economic costs exceed budget, i.e., when the cost per ton of oil exceeds the threshold; A4. Algorithm-driven optimization: Based on model predictive control (MPC), we predict oil production for the next 30 days and optimize chemical injection concentration for the next seven days. We define a reward function based on reinforcement learning (RL), dynamically adjust the ratio through DQN algorithm training, and achieve iterative optimization of the oil displacement ratio.

Images