Process-Based and Stochastic Modelling of a Fluvial Meandering System: Example of the Loranca Miocene Succession (Spain)

Isabelle Cojan¹, Cédric Beaudelot², Francois Geffroy¹, Sébastien Laratte², Didier Renard¹, Christophe Rigollet², and Jacques Rivoirard^1
1Ecole des Mines de Paris, Fontainebleau, France
²Gaz de France, Paris, France

Process-based models provide a better insight into the geometry of the sedimentary bodies and the 3D architecture in heterogeneous reservoirs. They also enable us to test the impact or the interactions of the various controls on the sedimentary succession. In the following, we present the potential of the model developed for fluvial meandering systems to choose/evaluate the choice of the fluvial parameters through a case study. The model is based on the physical equations that rule the channelized flow and takes into account the variability of the processes by a stochastic component. The model runs with a limited number of parameters and produces various architectures (net to gross, connectivity) by varying sorne parameters such as the frequency or intensity of overbank events, the spatial and temporal distribution of avulsions ...

The meandering fluvial deposits of the Tortola Formation (Loranca basin) correspond to a 130 m thick succession, outcropping along more than three kilometres long cliffs. Input parameters for the modelling are based on field measurements (pointbar height and width, mud plug thickness, paleosol maturity stage, mean distance in-between paleosols, crevasse splay thickness and lateral extension,…) combined with stratigraphic (mean aggradation rate) and bibliographic (slope) data.

Channel parameters did not change significantly throughout the succession while large variations in sandbody connectivity were observed. These are taken care of by changes in frequency of avulsions and aggradation rate (ie frequency of floods). Internal structure of the channel deposits are well reproduced by the simulations including channel lags, complex point bars and mud and sand plug. The modelling also honors data at three given sections providing a full 3D block that is fully documented with facies.

AAPG Search and Discovery Article #90078©2008 AAPG Annual Convention, San Antonio, Texas