Relationships Between Seismic Data and Facies Proportions in Meandering Fluvial Systems

Albouy, Emily¹, Laure Philippe², Marie-Christine Cacas¹, Simon Lopez³, Caroline Joseph¹ (1) Institut Français du Pétrole, Rueil Malmaison Cedex, France (2) ENSG Nancy, Nancy, France (3) Ecole des Mines de Paris, Fontainebleau, France

Basin modelling using algorithms based on the physical laws governing sediment trans­port can provide the general vertical evolution of lithological proportions and depositional slope on grid sizes which are generally kilometric in scale. Deterministic or geostatistical modelling can provide a more detailed, downscaled picture of specific depositional environ­ments respecting the vertical and horizontal proportions provided by more regional model­ling and by well or outcrop control.

Stratigraphic simulations of meandering fluvial environments have been carried out in order to reproduce different depositional architectures: using a mixed deterministic-strati-graphic approach, geological processes described by the equations of fluvial hydrodynam­ics (meander migration, erosional downcutting) are simulated with a random component governing catastrophic events (avulsion, levee breaching…)

The lithological 3D grids which have been simulated, representing constrasting fluvial architectures and lithological compositions, have been used to create synthetic seismic data by 1D convolution. The quality of the correlation between classical seismic attributes such as amplitude and energy and sand body proportion is shown to depend both on the thick­ness of the observed interval and on the presence of certain secondary facies. Some textur­al attributes based on statistical crossplots describing the spatial correlation of seismic amplitude over a sliding window can greatly improve the success of prediction of sandbody location.

The comparison of the seismic attributes of the blocks generated by stratigraphic simu­lations with attributes calculated from real seismic data can allow a quantitative evaluation of the relative success of stratigraphic models in recreating sedimentary architecture in an given depositional environment.