TX 75083-3836 U.S.A., fax 01-972-952-9435.
AbstractThe use of computer models allows the forecast of the behavior of oil fields and the optimization of the production process. Usually, the simulation is hindered by the size of the field, the number of wells, the complexity of the system, and the amount of data necessary to guarantee the reliability of the model. This work addresses the reduction of computational effort, maintaining the level of precision of more complex simulations, when the interest is just in a certain area of the field, for example, in the forecast of the behavior of mature fields. There are several ways to accomplish fast simulations of this type of problem. Two techniques were studied in this work: local refinement and the simulation with open boundaries. For the evaluation of local refinement technique several different grids were tested. For the analysis of the simulation with open boundaries, the original field was decomposed in smaller units and the flows through the boundaries were represented by virtual wells. Models with variable number of virtual wells in each boundary were tested. Using the developed methodology it was possible to reduce significantly the time of simulation maintaining a reasonable precision.