Perpustakaan Digital ITB

The purpose of history matching is to obtain a reservoir model with the smallest error between simulation data and historical data. In conducting historical reservoir model matching, there are several processes for adjusting reservoir parameters, such as aquifer strength, transmissibility, rock area, relative permeability, static properties, etc. Generally, this process is done manually and requires a long processing time. To solve this problem, there is a history matching process which is assisted by an automated strategy for history matching. In determining the uncertainty parameter, sensitivity analysis is carried out to detect the reservoir parameter that has the most influence on reservoir performance. This sensitivity process is carried out using the Latin Hypercube algorithm which will assist in identifying the parameters that are influential and less influential in the history matching process. Then the Particle swarm Optimization Algorithm is used in the optimization process to minimize the value of the difference between historical data and calculations. There are 12 uncertainty variables used in the history matching process, where the parameters that have a significant influence include permeability, relative permeability in certain regions, and aquifer thickness. Probabilistic Forecasts are carried out on models that have been matched with estimates for 25 years. This process is intended to obtain optimal oil production results with maximum RF values. Maximum oil production results are obtained by using work over scenarios with infill wells and re-in scenario. The result for this process is successfully generated a forecast model with an RF value of 38.17%