Perpustakaan Digital ITB

Waterflooding is a widely used technique in the petroleum industry to enhance oil recovery by injecting water into reservoir formations. Accurate prediction of waterflood performance is crucial for effective reservoir management. This study focuses on developing a predictive modeling approach using growth curves and curve fitting techniques to estimate the waterflood key performance indicators in the TM Field such as water cut and cumulative oil production. The 4W, 4Y4, and Liu’s growth curve models are implemented and compared to evaluate their effectiveness in capturing reservoir behavior and predicting waterflood performance. The growth curve models are optimized using Levenberg-Marquardt, trust region reflective, and dogbox methods in curve fitting. The accuracy and reliability of the predictive models are assessed using metrics such as root mean square error (RMSE) and coefficient of determination (R-squared). The predictive model developed in Python demonstrated its capability to forecast various performance indicators, such as cumulative oil production and water cut. The model's predictions were compared to the simulation forecast results, and the accuracy and reliability of the model were assessed. The results showed that each growth curve model yields promising predictions, with the majority of RMSE values below 0.09 and R-squared values above 0.89. A sensitivity analysis is performed to observe the precision of the predictive model by varying the time data interval employed as the input for the testing data. However, a comparative analysis reveals that each growth curve model has its own advantages, so the most suitable growth curve model for implementation can vary depending on the actual field conditions. This study contributes to the understanding of growth curve implementation and curve fitting techniques in waterflood performance prediction, providing valuable insights for reservoir management and decision-making processes in the TM Field.