![]() | Only 14 pages are availabe for public view |
Abstract The purpose of this thesis is to deliver a new model to predict the Fold-Of-Increase in Productivity index (i.e., well productivity) for a multilayer reservoir stimulated by limited entry fracturing. The reliability prediction of well productivity after limited entry fracturing treatment will facilitate realistic decision-making of the fracturing treatment. For this purpose, a new perforations friction model was architected. The classical Unified Fracture Design model was re-casted to include the further simulated perforations friction that could successfully divert the fracturing fluid to the end of the treatment. The solution methodology involves the application of artificial neural network (ANN). Actual historical field data of limited entry fracturing treatments have been used to train, validate and test the new ANN perforations friction model and the re-casted classical UFD. A statistical analysis comparison between the newly proposed ANN models and the previous models demonstrates that the results from ANN models are the most reliable estimation of the actual historical data. |