Pivot, Frédérik; Forge, Aurèle
TOTAL, Pau, France.
Modern geomodels include more information coming from seismic data, because techniques like seismic reservoir characterization or fine strati-seismic interpretation allows to extract precise and relevant quantitative information from seismic cubes. However, building a geomodel is always a matter of balance : even if seismic data quality is good, showing subtle details, final grid compiles only the main features controlling the fluid flow, discarding what is considered secondary by the integrated team joining geologists, geophysicists and reservoir engineers. But mistakes always exist in these choices. Moreover, continuous work on seismic data always bring new information after the geomodel building began: edited horizons, new faults, new attributes, new time-depth conversion.
Once the geomodel is built, an important phase of dynamic calibration begins for matching production history and well tests. We think that this phase must also include what we call the "back loop " on seismic data. Indeed, goal is not only to calibrate dynamically the model but also to verify that the structure and the static reservoir properties are really compatible with the evolving detailed seismic interpretation and the advanced lithoseismic attributes.
In this paper, we present several key technologies that allow to check the reservoir grid in combination with seismic and interpretation data, and, if mistakes are identified, to correct them without re-building the grid from scratch:
These technologies are illustrated on various exemples, including turbiditic clastic and carbonates fields.
AAPG Search and Discovery Article #90155©2012 AAPG International Conference & Exhibition, Singapore, 16-19 September 2012