Analisis Faktor Signifikansi Adsorpsi Surfaktan Pada Injeksi Eor Dengan Pendekatan Artificial Neural Network

Tri Oktavia, Desy (2022) Analisis Faktor Signifikansi Adsorpsi Surfaktan Pada Injeksi Eor Dengan Pendekatan Artificial Neural Network. Other thesis, Universitas Islam Riau.

Text 173210086.pdf - Submitted Version Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Surfactants are widely used especially in the petroleum industry for the Enhanced Oil Recovery (EOR) process because of their ability to affect rock surface properties and the oil-water interface. However, in the process, the main obstacle encountered is the adsorption of surfactants on the surface. The adsorption can be caused by several factors, namely surfactant concentration, reservoir temperature, surfactant density, Ca 2 + concentration, molecular weight and salt concentration. This factor will be the parameter in this research. With the aim of knowing the adsorption factors that most influence the adsorption of the surfactant. The research uses an Artificial neural network (ANN)-backpropagation approach as one of the input data so that it can produce output data in the form of ranking for factors that affect surfactant adsorption. Then used Random Forest as a comparison of the accuracy of the data obtained in the ANN approach used. Reservoir modelling using reservoir simulation software (CMG) STARS. By using 300 data with a ratio of 80% dataset for training and 20% dataset for testing. The R2 value of the training data is 0.592 and the testing data is 0.210. For the random forest method, the R2 value for testing data is 0.860 and for training data is 0.321. It can be concluded in this study that the use of Random Forest is more accurate than using ANN seen from the R2 value.

Actions (login required)

View Item