Astsauri, T. Mhd. Sofyan
(2021)
Studi Sensitivitas Parameter Low Salinity Water Injection Dalam Meningkatkan Oil Recovery Pada Carbonate Reservoir Menggunakan Random Forest Algorithm.
Other thesis, Universitas Islam Riau.
Abstract
LSWI has been validated by many experiments as a promising technique to improve perolehan minyak. Regrettably, even though nearly 60% of the world's total oil reserves are accumulated in carbonate rocks, until now there is only 1 LSWI implementation report that has been carried out in the carbonate field. This study applied a Machine Learning Algorithm based on Random Forest Regression for eliminating the insignificant parameter and evaluating the correlation between each parameter and response parameter on the LSWI process. This study is initiated by building 1000 experimental designs of LSWI parameters, Reservoir & Injection Temperature, Volume Injection, Formation Water Composition (Ca2+, Mg2+, SO42-, Na+ & Cl-) dan Injection Water Composition (Ca2+, Mg2+, SO42-, Na+ & Cl-), using Design of Experiment on CMOST by Computer Modeling Group (CMG) with Recovery Factor is set up as the response parameter. Finally, the sensitivity analysis is carried out on Random Forest Regressor based on the decrease in the mean squared error (MSE). The Random Forest Algorithm methods respectively recognized the following parameters, Injection SO42- Composition, Formation Water SO42-Composition dan Volume Injection as the top 3 lists of most significant parameters. The hyper-parameters of Random Forest also optimized and the smallest MSE is attained by numbers of tree 177 (1.7213) and R2 for both training and test data respectively are 0.96 and 0.905. The information about the significant operation parameter of LSWI process that presented in this article is potential bearing the novel to the industry. The insight into those parameters is predicted to be useful to encourage the LSWI implementation on Carbonate Reservoir. Besides, the Random Forest Algorithm should be considered for use in performing features selection whether for LSWI or other future works with R2 0.906 in this study.
Actions (login required)
 |
View Item |
