Introducing Automation in the Interpretation of 3D Seismic Data

X. Ji¹ and Y. Luo ¹

Abstract

This paper demonstrates that some aspects of 3D seismic interpretation can be performed automatically by computers. Two essential interpretation procedures are picking of horizons and faults. We will show that all horizons can be picked automatically via an inversion algorithm without any human intervention. Moreover, we introduce a new method which processes 3D seismic volumes and generates images of fault likelihood and corresponding fault-plane. Based on these fault attributes, fault surfaces can be tracked automatically. Instead of picking horizons based on amplitudes of seismic cubes, local dips of seismic events are computed and inverted for 3D curved time surfaces; each of these surfaces corresponds to a specific geological deposition time. The inversion will yield thousands of such geological time horizons, unlike the few picked horizons available from manual seismic interpretation. In practice, we have made this inversion algorithm applicable to seismic cubes of any size by making use of a parallel iterative solver. In addition to the inversion of horizons, we have employed a novel edge detection technology to calculate fault-related attributes, including the fault likelihood and fault-plane orientations. We will demonstrate that the new technology can reduce the noise resulting from conventional coherency-based methods. These enhancements are achieved by scanning among all possible fault angles and generating the most likely fault attributes. These tools enable 3D seismic interpretation to be completed in a much reduced cycle time. It also allows interpreters to have a quick overview of the depositional history and fault networks. With this type of automation, human bias is also minimized and thus reduction of potential errors when interpreting seismic volumes is readily achieved.

AAPG Search and Discovery Article #90188 ©GEO-2014, 11th Middle East Geosciences Conference and Exhibition, 10-12 March 2014, Manama, Bahrain