The presence of multiples, especially interbed multiples in the reservoir zone, pose an interpretation risk; hence, ambiguity in the extraction of meaningful reservoir properties. In this paper, we present a case study from the Culzean project high-density ocean-bottom cable (HDOBC) survey demonstrating our multiple attenuation approach for predicting and subtracting surface and interbed multiples.

The main differentiators for this study are the use of curvelet transforms for the surface multiple model subtraction and a layer-based method for interbed multiple prediction. The curvelet transform facilitated the separation of conflicting dips in the reservoir zone; hence, a more effective deterministic water layer demultiple and 3D surface-related multiple elimination SRME model subtraction. As the legacy processing and initial analysis of the Culzean data emphasised the fact that interbed multiples have limited velocity discrimination from the primaries in the reservoir zone, conventional methods based on velocity, dip or periodicity are not capable of targeting these multiples. However, with a geophysical constraint on the generator information derived from the borehole data available in the area, the data-driven technique of layer-based interbed multiple prediction provided good results.