The Delphi research consortium at the Delft University of Technology (‘TU Delft’) in The Netherlands has been very successful in developing new preprocessing, imaging and inversion algorithms for the geo-energy industry in the areas of seismic reservoir exploration and seismic reservoir monitoring (see Delphi Consortium).

Some of the Delphi algorithms have become industry standards, like the surface-related multiple elimination (SRME) algorithm for 2D and 3D marine seismic data. In the last decade, a lot of emphasis has also been put on illumination of targets in complex 3D subsurface structures by incoherent sources (‘blended acquisition’) for both land and marine, and the accompanying deblending algorithms. One of the observations is that multiple scattering can be viewed as ‘natural blending’, meaning that via multiple reflections the subsurface is illuminated many more times and multiples might even illuminate areas that are not reached by primary reflections.

Taking all this into account, it means that multiple scattering provides great opportunities in imaging the subsurface beyond the capabilities of the traditional primary-only assumption that is embedded in most imaging processes. Because acquisition geometries will always be limited due to natural obstacles, logistic and economic reasons, the complementary illumination from the multiples becomes very useful and will reinforce the primary reflections. In fact, what used to be considered noise becomes valuable signal and without any additional costs.

Within the Delphi consortium these considerations led to the concepts of the Full Wavefield Migration (FWM), where surface and interbed multiples are contributing to the final image and their usual cross-talk is removed. Furthermore, erroneous or uncertain propagation velocities can be included in the inversion process, which give rise to the Joint Migration Inversion (JMI) algorithm. These concepts have led to the establishment of a fully automatic procedure in which the reflectivity image and the underlying propagation velocity model are updated while matching all events in the data: primaries, surface and internal multiples!

Based on the research concepts mentioned above, a new technology company was founded, Delphi Studio for Imaging (DS4I), offering Delphi solutions to the industry on a commercial basis. DS4I focuses on the application and demonstration of the enhanced new Delphi algorithms on strategic seismic datasets. The main objective is to provide high-resolution seismic solutions to describe and characterize the subsurface. In the last few years, DS4I has developed a proprietary implementation of the JMI algorithm that can also handle datasets with sparse geometry acquisition, like 3D VSP and 3D OBN seismic data. Other areas of expertise are deblending, seismic data interpolation and primary/multiple estimation. The mission of DS4I is to accelerate innovation by providing the early application of new algorithms to seismic field data or making such algorithms better accessible to the industry.