The Role of Faults in Turbidite Compartmentalization at the Bed-scale

Manzocchi, Tom¹, John J. Walsh², Mark Tomasso³, Julian Strand², Conrad Childs², Peter Haughton⁴ (1) University College Dublin, Dublin, 4, Ireland (2) University College Dublin, Dublin 4, Ireland (3) The University of Texas at Austin, Austin, TX (4) University College Dublin, Dublin, Ireland

Sandstone connectivity has been measured in a comprehensive suite of faulted and unfaulted bed-scale turbidite models assuming permeable sands and impermeable shales and shale smears. Sedimentologically the sheet-like geometries are characterized by differences in net: gross ratio, amalgamation ratio and the shape, size and orientation distributions of the beds. Fault-system variables include fault frequency, length, maximum throw and orientation distributions. Shale smears have been modeled using either cut-off values of deterministic fault-surface proxy-properties (e.g. SGR, SSF, CSP), or by explicitly modeling stochastic smears associated with each shale horizon on the fault-surfaces.

Factor analysis is used to establish sensitivities to connectivity and to connectivity changes in representative models as a function of model parameters or, more usefully, dimensionless ratios between sedimentological and fault-related parameters. If the size of faults (maximum throw and length) are a few times smaller to a few time larger than the bed sizes (thickness and length), we find that most variables considered can be influential on the connectivity of the sequences, with particular variables being more or less important depending on the values of other fault-related or sedimentological variables. In general, however, systems at these scales are rarely more compartmentalized in three dimensions than their unfaulted counterparts, irrespective of whether and how fault rock properties are modeled. Faults that are large in relation to the beds can, however, have a significant positive or negative impact on connectivity depending on the characteristics of the shale smears.