nisms and reservoir heterogeneity properly on a numerical mesh. In addition, the numerical dispersion and diffusion An adaptive local mesh refinement algorithm originally developed for unsteady gas dynamics by M. J. Berger is extended to properties of conventional low-order methods can domiincompressible flow in porous media. Multilevel iteration and donate the physical dispersion terms in the models. Third, pomain decomposition methods are introduced to accommodate the rous media flow often involves important information on elliptic/parabolic aspects of the flow equations. The algorithm is many physical scales. Representing physically meaningful applied to a two-phase polymer flooding model consisting of a data on various length and time scales efficiently during a system of nonlinear hyperbolic mass conservation equations coupled to an elliptic pressure equation. While the various numerical numerical simulation is a formidable task.
methods used have been presented previously, our emphasis is on
To provide useful information for the development of their consistent combination within the adaptive mesh refinement recovery processes, field-scale simulations may need to framework to treat important problems in porous media flow. To resolve fine-scale localized flow behavior. Usually this achieve efficient, easily maintainable code, we have exploited the means that the computational mesh must be sufficiently features of object-oriented programming for the overall program structure and data management. Examples of algorithmic perforfine to resolve the length scales of important transient and mance and computational results are provided. แฎ 1997 Academic Press static features. Consider that a single petroleum reservoir may be hundreds of meters thick and tens of kilometers in diameter and involve hundreds of wells. Also, fluid models
The U.S. Government's right to retain a nonexclusive royalty-free pense needed to obtain a desired level of resolution in a license in and to the copyright covering this paper, for governmental variety of numerical simulation problems [5, 13, 14, 27, 43, purposes, is acknowledged.