| dc.description.abstract | The development of fine grid geolgocial models has attracted great attention in the past
decades. Meanwhile, the need for reliable upscaling and coarsening techniques is
continuing. Besides the computational efficiency, upscaling can also offer other
advantages. The desire for the assessment of risk and uncertainty in reservoir
performance is another key issue that is attracting the researchers. Predictions are
necessarily of a statistical character because uncertainty is involved in almost all the
aspects of the reservoir characterization. Significantly upscaled models are desired when
the full assessment of project risk and uncertainty are to be accomplished. The problem of
upgridding fine scale models into the coarsened ones is still an attractive and challenging
topic demanding much more effort in the reservoir simulation field.
We proposed a modified static coarsening algorithm that has better performance without
introducing extra computation cost. This algorithm combines adjacent layers based on
static calculations such that the heterogeneity measure of a defined static property is
minimized within the layers. In addition, the geological model coarsening will also rely
on preserving geological marker information. This combination of static calculation and
geological information enables this algorithm to generate models more closely to the true
ones. The power and utility of our approaches have been demonstrated using both
synthetic and field examples.
To assist the optimal coarsening procedures, we developed and implemented a GUI
(Graphical User Interface), named MARS. We focused on building up a C++ based user
interface which enables users to handle access the upgridding simulation visually. This
MARS software package is a general purpose GUI for applications that make use of
graphs as an underlying data model. MARS, which allows user to create simulation cases,
import and modify data, and generate graphical geological figures, is developed to
facilitate the operation of this coarsening procedures and the interpretation of the results
obtained by this model. The user of MARS will be graphically guided through the entire
process of creating coarsening simulations. | en |
