Muzdalifah, Sarah (2023) Optimasi Alkaline-surfactant-polymer Flooding Di Reservoir Karbonat Dengan Metode Bayesian Network. Other thesis, Fakultas Teknik Perminyakan.

Text 183210797.pdf - Submitted Version Restricted to Registered users only Download (3MB) | Request a copy

Abstract

Alkaline-surfactant-polymer (ASP) flooding is a Chemical Enhanced Oil Recovery (CEOR) method which is proven to be effective in increasing oil production. However, an optimal injection strategy is still needed so that the use of ASP Flooding injection has a high success rate. ASP Flooding is suitable to be applied to carbonate reservoirs because carbonate reservoirs are oil-wet in nature where oil tends to stick to the reservoir rock. The injection fluid components in this study were alkaline Na2CO3 (sodium carbonate), ABS surfactant (alkylbenzene sulfonate) and HPAM polymer (hydrolize polyacrylamide). This research was conducted to get the best results in the form of an optimal scenario of the parameters used in the utilization of ASP injection. This research begins with obtaining sensitivity data in the form of a dataset, then proceed with data processing through a jupyter notebook on machine learning python with the Bayesian Network algorithm. The steps taken to determine the optimum value of each parameter are by conducting training and testing data scenarios with data ratios of 70:30, 80:20 and 90:10 with the Bayesian Network algorithm. From the research results obtained optimization on the parameter Bottom Hole Pressure (BHP) in the Training 80% data scenario and Testing 20% data with an R2 value of 0.980 with an MSE of 1.08432 is the model with the highest optimization value. Whereas in the parameters of surfactant concentration, salinity and adsorption, optimal results were not obtained because there was no distribution of data from the Recovery Factor value.

Actions (login required)

View Item