Identifikasi Parameter Signifikan Dalam Penentuan Prioritas Penanganan Heatloss Pada Proses Cyclic Steam Stimulation Menggunakan Artificial Neural Network

Flonia, Monalisa (2021) Identifikasi Parameter Signifikan Dalam Penentuan Prioritas Penanganan Heatloss Pada Proses Cyclic Steam Stimulation Menggunakan Artificial Neural Network. Other thesis, Universitas Islam Riau.

Text 173210527.pdf - Submitted Version Download (3MB)

Abstract

Heatloss is one of the significant challenges in steam injection. However, both steam injection and oil production processes inevitably involve heat loss due to the temperature difference between the fluid and the surrounding formation. Therefore, it is necessary to do research on the factors that affect heat loss from fluid to formation. Steam is an efficient medium for heating the subsurface layer and the liquid reservoir contained therein, because much of the energy available in the steam is in the form of latent heat which is released at a constant temperature as it condenses and is connected to the relatively cold subsurface. In this study observed several parameters that affect heat loss in the wellbore and reservoir, namely injection parameters such as steam injection pressure, steam temperature, steam quality, steam injection rate, soaking time. Wellbore parameters such as inside and outside of outer tubing, casing, injection well depth. Then a research will be carried out using one of the methods from artificial intelligence (AI) called artificial neural networks (ANN) to be able to study the performance parameters and classify the parameters that most influence the occurrence of heat loss (heat loss) faster by doing train and test. Therefore, this study will focus on the application of ANN using python software in classifying heatloss which has previously been carried out on CMG (Computer Modeling Group) software. By using 525 sample data, the results are classified as very good and optimal using 20 hidden layer nodes with an RMSE value of 0.118 for training data and 0.178 for testing data, while the MAPE value for training data is 0.708 and testing data is 0.988. While the Coefficient of Determination (R²) in the training data and testing data is 0.999 and 0.998 so that it can be classified as having high accuracy results because it is close to a value of 1. While the most influencing parameters, respectively, are outside of radius tubing, inside of radius tubing, injection temperature, soaking time, inside of radius casing, injection pressure, depth, injection rate, outside of radius casing and steam quality.

Actions (login required)

View Item