Nowadays oil and gas industries are focusing in production and development where well drilling is common and often done. One among many aspects need to be considered for drilling safety is pore pressure prediction. There are so many methods used in prediction the pressure including machine learning ANN but no one has ever done this with ANFIS which is combination of ANN and FIS machine learning. This study wants to use ANFIS for making a pore pressure distribution in 2D seismic data with 70% accuracy. Both pre-stack and post-stack seismic data are used here with well and RFT measurement. The modified Eaton-Azadpour and Eaton are two methods that applied to predict pore pressure. The methods are considered to be good in prediction pore pressure as they accommodate the rocks condition during drilling. The well parameters models are then distributed with ANFIS to find its correlation with P-impedance, S-impedance and density so we may find its distribution in 2D seismic data. The result shows that pore pressures are distributed very well but still need another study to give information regarding drilling safety.
|Journal||IOP Conference Series: Materials Science and Engineering|
|Publication status||Published - 1 Jul 2020|
|Event||2nd International Conference on Science and Innovated Engineering, i-COSINE 2019 - Malacca, Malaysia|
Duration: 9 Nov 2019 → 10 Nov 2019
- Pore Pressure