Delphi technology can be applied to the data of Delphi sponsors by making bilateral arrangements (‘experimental processing services’). The Delphi team has developed new algorithms in the topics of deblending, interpolation beyond aliasing, primary-multiple separation (SRME+), automatic velocity estimation + full wavefield migration (JMI) for both surface and VSP data, and broadband full waveform inversion for reservoir characterization (FWI-res). On request, all these algorithms will be applied to the data of our sponsors in the Studio.
The objective of Delphi is to develop new concepts in (robotized) seismic acquisition and preprocessing, full wavefield migration (FWM), joint migration-inversion (JMI) and full waveform inversion (FWI-res) to decrease the uncertainties in subsurface models. So far, algorithms have been developed on the topics of acquisition design for coherent and incoherent shooting, deblending, near-surface preprocessing, data-driven primary-multiple separation (SRME+), automatic velocity estimation combined with full wavefield migration (JMI) for surface and VSP data, and broadband full waveform inversion for reservoir characterization (FWI-res). Recently, also EM modeling and EM inversion is included. In Delphi, new concepts are tested on our synthetic and / or field data (‘proof of principle’), but execution of processing jobs – how useful they may be for giving practical experience to our students – is not aimed for. In the consortium we prefer to keep the focus on developing and testing new concepts.
In the last few years, however, we obtain an increasing number of requests to apply Delphi algorithms to the data of our sponsors. This is particularly the case for complex situations in land and marine, where current technology does not give satisfactory answers. Actually, many of the new sponsors inquire whether Delphi can provide specialized processing services. It makes the membership more attractive. Looking at global innovation trends, we see an increasing interaction between the development and deployment of new concepts. Practical experience shows that this way of working significantly accelerates the innovation path, decreasing the time between the birth of an idea and the start of commercial use. Looking also at the need of applying the Delphi algorithms to a variety of field data at an early stage, we have started to establish a company Delphi Studio for Imaging that is specialized in processing field data with Delphi technology. We aim at a cyclic interaction between the Delphi Consortium and the Delphi Studio (see Figure above). With such an interaction, we can ap- ply Delphi concepts to field data without directly disturbing the research projects. In addition, we can feedback the experience in the Studio to the research teams. In this way we expect to better connect science with business. Figure above shows the principle. We have started with providing services in the following topics:
1. Deblending and interpolation beyond aliasing
2. Using multiples in migration for improved images
3. Automatic velocity estimation as part of full wavefield migration
4. Estimating elastic parameters in the reservoir by full waveform inversion.
The Studio resides on the campus of the Delft University.