Tamara, Fairuz Deby (2023) Penentuan Skin Formation Reservoir Pada Buildup Test Dengan Menggunakan Svm Algoritma Pada Sumur X. Other thesis, Universitas Islam Riau.

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

Abstract

Well testing is generally used to determine the conditions at the time the well was produced. So far, formation damage in wells is usually carried out using a method known as a pressure buildup test. Analysis of the results of pressure buildup tests generally uses a technique called a horner plot which has been around for quite a long time. So a new technique is needed that can determine the skin factor value. This research aims to be able to predict the skin formation value from the buildup test using the support vector machine method. In this study, researchers built a simulation reservoir model using the Computer Modeling Group (CMG)-IMEX. Then carried out 500 experimental datasets with six parameters as well test input data, namely Initial pressure, Production rate, Thickness, Factor Volume Formation, Permeability and Viscosity using the Design of Experiment (DoE) method with Skin formation as the response parameter. To make it easier to determine skin formation, it is done using the Support Vector Machine Algorithm method with Python Programming Language for modeling, as well as hyper parameter tuning. Of the 500 DoEs that have been prepared, data sharing is carried out with a ratio of 80% for training and 20% for testing data. From the simulation results carried out by the Support Vector Machine Algorithm, a predictive model was obtained with Mean Square Error (MSE) and Mean Absolute Error (MAE) values which were close to 0 and for R2 training and testing with values of 0.990 and 0.991 respectively. In this research, we explain the application of machine learning in determining reservoir parameters and the predictive model Support Vector Machine Algorithm can be used as a reference and evaluate the performance of the Buildup test in predicting Skin formation values with an accuracy level of 0.991 and in a short time without using reservoir simulation which can take a lot of time.

Actions (login required)

View Item