Integration of Fault Location Uncertainty in Time to Depth Conversion

Abstract

Summary Faults are objects interpreted from seismic data that are often difficult to characterize. Not only the type of a fault can appear different depending on the domain the modeler is working on (time or depth), but accurate shape mapping often extends below the level of coherent seismic data. This proves especially impactful when we consider that simple errors in the fault-network characterization can render a subsurface model useless in terms of its flow properties and provide erroneous values for volumes of reserves. Nevertheless, in absence of proper technology, depth conversion workflows do not account for fault uncertainties, instead relying on a single deterministic interpretation. This paper presents a new methodology that considers positional fault uncertainty in time-to-depth conversion workflows. By considering different realizations of a fault, the target surface is dynamically adapted to this new position and can be used to update further volumetric computations. The proposed methodology can be fully automated and results in a more complete exploration of the uncertainty space regarding fault interpretation.