Article Details


Oleh   Fahmi Satrio Pinandito [12216011]
Kontributor / Dosen Pembimbing : Dr. Dedy Irawan, S.T., M.T.;Muhammad Arif Naufaliansyah, B.Eng., M.Eng.Sc.;
Jenis Koleksi : S1-Tugas Akhir
Penerbit : FTTM - Teknik Perminyakan
Fakultas : Fakultas Teknik Pertambangan dan Perminyakan
Subjek : Mining & related operations
Kata Kunci : Petrophysics, porosity prediction, machine learning, logging
Sumber :
Staf Input/Edit : Suharsiyah   Ena Sukmana
File : 1 file
Tanggal Input : 28 Jun 2022

Through years, all stages of the petroleum industry depend on the retrieval and analysis of cores. One of the ways to acquire complete, vertically continuous samples that enable the visual inspection of the reservoir characteristics are through cores. Understanding the characteristics of the pore system in the reservoir requires the use of cores. It could give the information we can learn from core samples to forecast reservoir performance and strategies to maximize the hydrocarbon recovery. However, as the core sampling are considered expensive, the sample only recovered in the specific range and interval of the well, which makes some area unsampled. In this study, a practical methodology is proposed for predicting the rock parameters, in this case is porosity, in the uncored section of the wells by comparing the correlation of the core data with log data by using machine learning algorithm as a tool to become estimators of the given datasets. Several machine learning algorithms is used and to be compared and which gives the least errors are used to develop the model for the porosity prediction of the uncored well.