Mapping faults in 3D seismic data: why the method matters  

3D seismic reflection imagery is the most widely used resource for interpreting the geometric structure of faults in the subsurface. Yet these endeavours carry uncertainties, the significance of which are rarely discussed. We explore how the application of different workflows yield different interpretations of a single high-quality 3D seismic image-set. We describe and apply five mapping workflows, based on 2D derivations of imagery to map an array of small-scale faults. Some workflows use vertical profiles, a conventional approach, others use plan views. We also vary the amount of under-sampling. Stacking the fault maps derived from the five workflows into a heat map shows broadly similar trends and distributions of faults irrespective of the workflow deployed. However, juxtaposition mapping reveals differences in fault length and network pattern (linkage and segmentation) arising from the different workflows. We quantify the total fault areas and distribution of fault lengths for each workflow – revealing significant differences in these statistics. Our results show that mapping strategies impact the interpretation of fault geometry, their network patterns and derived statistics. This understanding is critical for assessing and risking fault interpretations – deploying multiple workflows can reveal inherent uncertainty in structural interpretation of 3D seismic imagery.