Creating a 3D image from 2D data using structurally conformable interpolation: a case study from the Beagle Sub-basin, NSW, Australia

Abstract

2D seismic data still provides key information for companies evaluating new permits on offer or entering new basins. However, working on multi-vintage 2D data can be time-consuming for several reasons, including getting correct navigation, variability of physical parameters like amplitude, time and phase between different vintages, and then interpreting the 2D data itself, which often results in gridding artefacts. In a step change to the use of traditional 2D data, TGS has developed a methodology called ‘structurally conformable interpolation’ – also known as 2Dcubed. It is created using input data from available 2D migrated stacks and velocities from available vintages. The workflow includes survey matching of different vintages, data-driven geological model building to interpolate large distances between existing data, and a 3D post-stack migration to minimise the 2D migration artefacts. The merging of these datasets successfully creates a 3D migrated image from legacy 2D data, offering better structure and continuity while increasing confidence in its interpretation. Interpretation of a 3D volume is much more efficient than when using 2D data and is free from 2D artefacts. With this methodology TGS has completed a project covering a 40000km2 area in the Beagle Sub-basin, north-west Western Australia, using existing 2D data from over 42 different vintages. The resulting output ‘Beagle Cube’ interpolated 3D volume has been interpreted for major regional trends and structures. The results are very consistent with the original 2D data, but with better definition of major structures. Another study comparing the interpretation between the interpolated 3D volume and the real open-file 3D shows excellent preservation of the structural picture within the interpolated 3D volume, not at the same level as real 3D, but it gives greater confidence in the regional interpretation conducted within areas that do not have 3D coverage. This paper will address how the interpolation methodology works stage by stage, the results of the final product and how it assists in performing regional interpretation in a quick timeframe.