Advanced seismic techniques such as EEI inversion, simultaneous inversion and so called facies based inversion are now routinely adopted for quantitative reservoir characterization of rock properties. Rock physics provides the crucial link between the quantitative geophysical measurements derived from inversion and geological parameters. However, in order to ensure that rock physics parameters, as encountered in well data, are accurately determined from spatially equivalent seismic data requires specialist processing beyond conventional image orientated workflows. Seismic data conditioning is specifically designed to provide a calibrated and conditioned AVO signal. In this regard reservoir specific processing routines are developed to mitigate multiple, random and coherent noise, over or under-corrected seismic velocities, frequency distortions and other undesirable effects. Synthetics generated at well locations are crucial for quality control at each step of the seismic data conditioning workflow. This paper demonstrates the advantages of performing seismic data conditioning, on a regional (multi-survey) North Sea seismic dataset, prior to reservoir characterization using a facies based inversion.
To take full advantage of AVO (amplitude versus offset) inversion techniques seismic data needs to be carefully conditioned so that any variation in amplitude is solely the result of reflection-coefficient changes and not seismic processing artefacts (Chopra and Castagna 2014). AVO gradient in particular is highly influenced by far offset amplitudes, yet provides a key control on the output of Vp/Vs and thereby significantly influences fluid and lithology prediction. Data conditioning routines are therefore focussed on the preservation of the AVO gradient, with each seismic dataset and each specific reservoir interval typically requiring a unique optimisation approach. Over-corrected and under-corrected seismic velocities, incorrect multiple removal parameters,
improper migration and poor angle stack range selection can all cause significant distortion of AVO signatures on pre-stack seismic gathers. However, with a targeted seismic data conditioning (SDC) workflow these undesirable effects can be minimised and corrected allowing noise to be subtracted without impacting primary AVO signals. Poststack noise that has not been removed by pre-stack SDC techniques must also be subtracted by routines that do not damage AVO signatures. In this paper, we summarise the specific SDC steps required for marine seismic data and
demonstrate them on a regional, multi-survey, North Sea, seismic dataset via the application of facies based inversion (Kemper and Gunning 2014).
