Optimization Of Water-alternating-gas Scheme In CO₂ Flooding And Storage Integration Project Based On Intelligent Algorithm

The CO₂ flooding is in order to improve the oilfield production rate in the CO₂injection technology,is widely used in tertiary oil recovery and the development of low permeability reservoirs,etc.In the context of striving to achieve"carbon neutrality",more and more attention has been paid to the side of carbon dioxide flooding that can achieve carbon storage while driving oil.This requires us to evaluate the effects of carbon sequestration and sometimes economic benefits as well as the effects of CO₂flooding.Water-alternating-gas is the most important development mode of carbon dioxide flooding,and its injection and production scheme has the characteristics of many parameters to be optimized and complex laws of displacement and storage,so it is difficult to achieve accurate optimization of multi-parameters and multi-objectives of water-alternating-gas scheme by using traditional optimization methods.In the comprehensive utilization of the numerical simulation and optimization algorithm,artificial neural network technology,build a set of complete water alternating flooding scheme optimization method,and with big emotion Wells in Jilin oilfield H59blocks of low permeability reservoir as the research object to practice proposed optimization method.Firstly,the numerical simulation model of the target reservoir is established on the basis of consulting the reservoir data of the study area.The concrete injection and production scheme of water-gas alternations was proposed,and six parameters to be optimized were selected,including the size of carbon dioxide slug injected in each cycle,water-gas ratio,gas injection rate,water injection rate,production well flow pressure and total carbon dioxide injection.Three objective functions were constructed to evaluate the displacement,storage and economic benefits of the water-gas alternate scheme in the study area.Then,1800 sets of data of water-air alternating random scheme are simulated by numerical simulation method.Three neural networks which can fit the objective function are obtained by using the data training.Finally,the single objective,double objective and triple objective optimization of the water-air injection scheme in the study area is realized by combining the neural network with various optimization algorithms.Optimization results show that less carbon dioxide slug,higher gas ratio,gas injection at low speed,high speed injection,low flowing pressure and production as much as possible help to maximize the total co2 injection to improve oil recovery;A slightly larger CO₂ slug,a low water-gas ratio,high-speed water injection,high-pressure extraction,and the maximum amount of CO₂ injected can help maximize CO₂storage;High water-gas ratio,low-speed gas injection,low-pressure production and appropriate total carbon dioxide injection are helpful to improve the economic benefits of the project.Pareto optimization model is more suitable for multi-objective optimization than linear weighting method.It can provide a Pareto optimal solution set for decision-makers’reference.There are some contradictions among the three objective functions,which cannot reach the optimal value at the same time.The economic net present value is mainly controlled by the amount of crude oil produced and the amount of carbon dioxide sources.Sometimes,it is necessary to increase the income by reducing the carbon dioxide storage,which conflicts with the goal of carbon storage.Therefore,technological innovation should be adopted to reduce the cost of gas sources and strengthen the economic subsidies for carbon storage.Research results show that the optimization method is used to effectively optimize the water alternating scheme,have important reference value for engineering practice.