The water coning phenomenon leads to decrease the wellhead pressure with moving of water into oil production zone, which is regarded as one of most serious problems during oil production. Therefore, the development of reliable models is important to predict the water coning breakthrough time, and consequently avoid the water coning phenomenon and production of water. To this end, the artificial neural network modeling strategy optimized with particle swarm optimization, least square support vector machine (LSSVM) approach coupled with the coupled simulated annealing optimization method, and finally decision tree method are implemented in current study to accurately predict the dimensionless breakthrough time of water coning. The results obtained in the present study demonstrate that the models proposed provide acceptable results in predicting the dimensionless breakthrough time of water coning. Furthermore, comparative study conducted illustrates the superiority of LSSVM methodology in terms of accuracy compared to the other methods investigated.
展开▼