Global Optimization for Parameter Estimation of Oil Reservoirs

Monday, April 22, 2002 - 1:30pm - 2:30pm
Keller 3-180
Susana Gomez (National Autonomous University of Mexico (UNAM))
In order to forecast the production, it is necessary to characterize the reservoir to find the properties of the porous media. An ill-posed inverse problem has to be solved.

If the inverse problem is solved through the minimization of a least squares objective function, the ill-posedness may imply multiple solutions. Thus the need to use a global optimization method able to find several solutions with good match to the data, that produce alternative scenarios of production. These solutions give also information about the uncertainty of the model parameters.

Here we will present a new implementation of the Tunneling Methods both sequential and parallel, and will give numerical results when used to characterize a synthetic and a real oil reservoirs (History Matching).