Prediksi Recovery Factor Injeksi Surfactant Huff and Puff Menggunakan Artificial Neural Networks Pada Sumur Minyak Dengan Kadar Paraffin Tinggi

Umam, A Habib Al (2020) Prediksi Recovery Factor Injeksi Surfactant Huff and Puff Menggunakan Artificial Neural Networks Pada Sumur Minyak Dengan Kadar Paraffin Tinggi. Other thesis, Universitas Islam Riau.

[img]
Preview
Text
153210340.pdf - Submitted Version

Download (1MB) | Preview

Abstract

Improved Oil Recovery (IOR) is a renewed Enhanced Oil Recovery (EOR) to increase oil recovery and inhibit a drastic production reduction from oil wells with high viscosity. One of IOR methods to be used in this is surfactant soaking method, which is injecting surfactants as a chemical and conducting a soaking or huff and puff on a well. Using 7 parameters, surfactant concentration, surfactant volume, surfactant mole weight, soaking time, number of cycles, injection rate and production rate, research will be carried out to predict the recovery factor (RF) value using the artificial neural networks (ANN) method. This ANN method is a deep learning method from input data and produces output data. By using 500 data with a ratio of 80% of the results of the RF calculation model from the CMG software for training and 20% of the results of the model for testing. In order to get optimal prediction results from RF using the ANN method, trial and error will be used on determining the number of hidden layer nodes. The optimal and stable hidden layer nodes are obtained at nodes 11 with value of MAE 0.293 and MAPE 1.42%, then the MSE and RMSE values in the training data are 0.1260; 0.3550 and the testing data are 0.3485; 0.5903. Other statistical analysis values such as R, R2, adjusted R2 are 0.9742; 0.9491; 0.9489 for training data and 0.9528; 0.9077; 0.9068 for testing data. The conclusion from this research using ANN to predict RF using 11 nodes of hidden layer is proven successful and very good.

Item Type: Thesis (Other)
Uncontrolled Keywords: IOR, surfactant, huff and puff, paraffin, artificial neural networks
Subjects: T Technology > T Technology (General)
Divisions: > Teknik Perminyakan
> Teknik Perminyakan
Depositing User: Mia
Date Deposited: 14 Mar 2022 04:45
Last Modified: 14 Mar 2022 04:45
URI: http://repository.uir.ac.id/id/eprint/8641

Actions (login required)

View Item View Item