| Predicting the pore pressure is the premise of achieving excellent, efficient and safe drilling operation and it is also the key part of protecting oil&gas layer. Marine carbonate basin in the middle-west of our country usually has very difficult geological structure and multivariate underground pressure system, so it is very significant to efficiently predict its pore pressure in the process of exploration and development of the basin.This thesis analyzes the mechanism of abnormal pore pressure through learning related theory knowledge, research condition and development tendency and concludes the main influencing factors of causing abnormal pore pressure in the carbonate reservoir. The thesis also analyzes the principle and defect of current prediction methods and finds the proper carbonate pore pressure prediction method by combining the forming mechanism of abnormal pressure in the carbonate.In this thesis, an experiencing model of pressure prediction is established based on the relationship of acoustic velocity and effective stress and porosity which is summarized in the experiments of the acoustic properties of carbonate rock samples of carbonate rock strata laws. Analyzing the formation factors of clay content and gas saturation which can influence the formation acoustic velocity, this thesis introduces clay content and gas saturation into the pore pressure prediction model. The pressure prediction software is designed using MATLAB, and the software can predict pressure in the carbonate formation using modified prediction model.At last because of the factors that has complicated relationship can be trained and self studied in the artificial neural net, we use the BP model in the MATLAB artificial neural net toolbox to predict pore pressure. This produces a new thought and method for pressure prediction of carbonate formation, this method is impacted by the number of study samples, it should be used in where the measured formation pressure data is rich.
|
PDF Full Text Request