| With the increasing scarcity of conventional easily exploitable oil reservoir resources,heavy oil will be an important strategic resource in the future.Due to the high viscosity of heavy oil and its difficulty in extraction,so the current main method is to use heating and viscosity reduction recovery technology,such as steam huff and puff,steam assisted gravity drainage,etc.During the process of high temperature steam injection from the wellhead into the reservoir for heating,the tubular column and wellbore are heated and the wellbore serves in high temperature environment for a long time.In the harsh environment,the casing is deformed due to thermal expansion and contraction.Due to the binding of the well wall,the casing generates large thermal stresses under the high temperature,which causes changes of the casing material properties.In this paper,we apply the basic theory of heat transfer,combine the law of conservation of energy,the law of conservation of momentum and the law of conservation of mass of the fluid in the warp barrel,establish the physical model of casing-cement ring-stratum in a thermal recovery well under the steam haff and pull process,and investigate the effects of thermal conductivity of insulated tubing,steam injection pressure parameters and steam injection temperature parameters on the temperature and stress fields of the casing.In this paper,the total heat transfer coefficient of the wellbore was calculated by numerical iteration method,and the error of the calculated total heat transfer coefficient was 6%,and the temperature of the inner wall of the casing was obtained in the process of calculating the total heat transfer coefficient.Comparing the casing inner wall temperature obtained from the formula iteration calculation and the casing inner wall temperature obtained from the finite element simulation,the error between the two is 10.8%.Comparing the casing inner wall temperature obtained from the iterative calculation and the casing inner wall temperature obtained from the finite element simulation,the error between the two is 10.8%.As the temperature of the injected steam increases,the temperature of the inner wall of the casing increases and the equivalent effect force increases.When the steam injection temperature increased from 150 ℃ to 350 ℃,the casing equivalent force increased by about 43%.When the steam injection pressure increased,the stress value of the casing decreased,and the decrease of the casing stress value caused by the increase of the steam injection pressure from 4 MPa to 18 MPa was about 3.7%,and the change of the steam injection pressure had a small effect on the casing equivalent stress value.In this paper,the simulation data of temperature field and stress field are used as the training set of the neural network.The parameters of well depth,formation temperature,steam injection temperature and steam injection pressure are used as the input of the neural network,and the temperature and equivalent stress of the casing are used as the output of the neural network to establish the neural network model.The secondary simulation of the temperature field and stress field of the casing is combined with the optimization of genetic algorithm.The results of the neural network operation show that the predicted value of casing temperature is within 1.2%error with the casing temperature value obtained from ANSYS simulation,and the predicted value of equivalent force is within 0.3% error with the equivalent force value from ANSYS simulation.The neural network was able to accurately predict the temperature and stress fields of the thermal recovery well,and the computation time was reduced by nearly 40% compared to the finite element simulation of ANSYS. |
