The methodology for 3D fault seal analysis in siliciclastic sequences is now well established. Using a structurally robust framework model, fault-rock phyllosilicate content can be estimated from a combination of fault displacement and stratigraphic architecture, and can then be used in combination with faulting/burial history to predict present-day seal parameters such as threshold capillary pressure and permeability. Sensitivities can be investigated, for example varying the fault displacement to model uncertainty in seismic mapping or sub-resolution structure (e.g. normal drag). Extending this methodology to previous geological time steps requires careful restoration of the displacement pattern on the fault, a technique usually referred to as fault displacement backstripping. Different approaches are required depending upon the fault style. ‘Growth faults’, i.e. those where sediment thickening across the fault at the free surface dominates the fault displacement gradient, are backstripped by removing horizon offset down-dip on the fault surface. By contrast, for post-depositional faults it may be more appropriate to remove uniform increments of displacement to restore the horizon offsets to previous time-steps in the fault propagation history. For faults that have been inverted, it is possible to restore them to their pre-inversion displacement pattern for input to fault seal analysis. Applying fault displacement backstripping to a number of examples, we make the general observation that on any given fault, fault seal properties were likely to have been more heterogeneous in the geological past, with greater distinction between sealing and non-sealing parts of the fault surface. Considerable complexity is added to this process when faults are linked together at branch lines. Firstly, fault displacements on each of the linked faults are likely to have been locally subparallel to the branch line rather than being dip-slip. Secondly, the relative timing of activity on the different fault surfaces can lead to various ways of partitioning the fault displacements through time. In turn this leads to significant differences in the restored offsets. Failure to consider them correctly (geologically and analytically) yields incorrect predictions of reservoir connectivity and fault seal at earlier geological times.
AAPG Datapages/Search and Discovery Article #90291 ©2017 AAPG Annual Convention and Exhibition, Houston, Texas, April 2-5, 2017